A meta-analysis of over 5 million patients found a statistically significant 2.3-fold increased risk of colorectal cancer in GLP-1 receptor agonist users, but the risk was not significant when compared directly to other diabetes drugs.
RR 2.31 — but context mattersGLP-1 users showed 2.3-fold higher colorectal cancer risk vs non-users, but no significant increase when compared to other diabetes drug users — suggesting confounding by disease rather than drug effect
What the researchers found
Seven retrospective cohort studies involving 5,066,681 patients were analyzed. The pooled analysis found a significantly increased colorectal cancer risk in GLP-1 RA users compared to the reference population (RR 2.31; 95% CI: 1.82-2.93; I² = 36%; p < 0.0001).
However, when GLP-1 RA users were compared specifically to users of other diabetes drugs, the colorectal cancer incidence was not significantly different (OR 1.73; 95% CI: 0.21-14.18; p = 0.61; I² = 100%). The extremely wide confidence interval and 100% heterogeneity in this comparison indicate the data is highly inconsistent. Quality assessment showed low-to-moderate risk of bias across studies.
Why it matters
With tens of millions of people now taking GLP-1 drugs, even a small increased cancer risk would have enormous public health implications. This meta-analysis raises an important safety signal that demands further investigation. However, the nuanced finding — that the risk disappears when comparing to other diabetes drugs — suggests the association may be confounded by obesity and diabetes themselves, which are well-established colorectal cancer risk factors. This is exactly the kind of data that regulatory agencies and clinicians need to make informed prescribing decisions.
How the study worked
Systematic review and meta-analysis following PRISMA guidelines. Researchers searched PubMed, Embase, Cochrane Library, Web of Science, and ClinicalTrials.gov from inception to December 2024. Seven retrospective cohort studies analyzing GLP-1 RA effects on colorectal cancer risk in type 2 diabetes patients were included. Study quality was assessed using the Newcastle-Ottawa Scale. Random-effects models were used for pooled analysis, with heterogeneity evaluated by I² statistic.
What this study cannot tell us
All 7 included studies were retrospective cohort studies, which cannot prove causation and are susceptible to confounding. Obesity and diabetes — the conditions requiring GLP-1 treatment — are themselves major colorectal cancer risk factors, creating inherent confounding. The comparison to other drugs showed 100% heterogeneity, indicating highly inconsistent results across studies. The duration of GLP-1 RA exposure and specific drugs used were not consistently reported. No dose-response analysis was possible. The data cannot distinguish whether GLP-1 drugs increase cancer risk or whether detection bias (more medical monitoring in treated patients) explains the findings.
How to read the evidence
This is a meta-analysis of retrospective cohort studies — a strong evidence synthesis method, but limited by the observational nature of the underlying data. The enormous sample size (5+ million) provides statistical power, but confounding by indication (patients take GLP-1 drugs because they have obesity/diabetes, which independently increase CRC risk) cannot be fully controlled.
When this study was published
Published in 2025 with data through December 2024, this is a very current safety analysis reflecting the latest pharmacovigilance data on GLP-1 drugs and cancer risk.
The bigger picture
GLP-1 drugs are the fastest-growing drug class in history, making safety surveillance critically important. This meta-analysis joins a complex landscape of cancer risk signals: some studies have flagged concerns about thyroid cancer (based on rodent data), while others have suggested GLP-1 drugs may protect against certain cancers through anti-inflammatory mechanisms. The colorectal cancer signal here is concerning but likely confounded. Long-term pharmacovigilance data from the millions now taking these drugs will ultimately clarify whether there is a true causal risk.
Questions still open
- Does the apparent increased colorectal cancer risk reflect a true drug effect or confounding by obesity and diabetes, which are themselves CRC risk factors?
- Would prospective clinical trials with longer follow-up confirm or refute this association?
- Should patients with existing colorectal cancer risk factors receive enhanced screening when prescribed GLP-1 drugs?
Common questions
Should I be worried about colorectal cancer if I'm taking a GLP-1 drug?
Why would there be an apparent increased cancer risk that isn't actually caused by the drug?
Read the original research
Association between GLP-1 receptor agonists as a class and colorectal cancer risk: a meta-analysis of retrospective cohort studies.
BMC gastroenterology, 25(1), 614
Citation
Zhong, Ying; Wu, Tingting; Khan, Najeeb Ullah. (2025). Association between GLP-1 receptor agonists as a class and colorectal cancer risk: a meta-analysis of retrospective cohort studies.. BMC gastroenterology, 25(1), 614. https://doi.org/10.1186/s12876-025-04211-4