GLP-1 Medications and Obesity-Related Cancer: What a 359,595-Person Study Found

What the 359,595-person study found

The clearest signal in this research is also a limited one: GLP-1 medications were linked to fewer obesity-related cancers in some comparisons, but the pattern changed depending on which diabetes drug people took and how advanced their diabetes was. That’s an intriguing clue, not evidence that GLP-1s prevent cancer. The distinction matters because a link between treatment and outcomes in medical records can reflect differences among patients, not just an effect of the medication.

The analysis examined electronic health records for 359,595 adults with type 2 diabetes treated at a Texas medical center from January 2010 through July 2025. None had a history of obesity-related cancer when they began treatment. Researchers compared people starting GLP-1 medications with people starting metformin, insulin, SGLT2 inhibitors, DPP-4 inhibitors, or sulfonylureas, while accounting for diabetes severity and existing heart or kidney complications. They tracked new cancer cases over three and five years. That broad comparison gives the findings more context than a single drug-versus-drug match, though it can’t eliminate every difference between the groups.

At three years, people starting GLP-1s had obesity-related cancer rates about 1.25 percentage points lower than those starting insulin and 1.14 points lower than those starting sulfonylureas. The differences persisted at five years. But GLP-1 users had similar rates to people taking metformin, SGLT2 inhibitors, or DPP-4 inhibitors. Among patients with the most advanced diabetes—especially those with serious heart or kidney complications—SGLT2 inhibitors were associated with the lowest cancer rates. The pattern doesn’t point to one simple protective effect; it shifts with the comparison drug and patients’ health.

Why the comparisons matter

It’s tempting to read the insulin comparison as a drug effect. That conclusion would be premature. This was an observational study: researchers followed treatment choices made in routine care rather than randomly assigning medications. Insulin is often prescribed when diabetes is harder to control, so people starting it may already face different health risks. SGLT2 inhibitors are often chosen with heart and kidney health in mind, which can also mean their users start out with different health profiles. Adjusting for recorded factors helps make comparisons fairer, but it can’t guarantee that the groups were alike in every relevant way.

Here’s the thing: the choice of comparison changes what the result can tell us. A lower cancer rate among GLP-1 users than among insulin users may reflect differences in treatment, differences in patients, or both. Similar rates compared with several other drug groups also complicate a broad claim that GLP-1 medications reduce cancer risk. This is more significant than it looks: the study’s most useful contribution may be the pattern of associations across treatment groups, not a headline claim about prevention.

The longer-term results have another limit. Many GLP-1 users had only recently started treatment, leaving fewer long-term users in the five-year analysis and making those estimates less certain. The data doesn’t fully explain what drives the associations—or whether GLP-1 treatment itself plays a role. A longer follow-up could help clarify whether the observed differences persist, but this study can’t answer that question on its own.

What this means for treatment

If you take a GLP-1 or are considering one, this study alone isn’t a reason to change your medication. Treatment decisions should reflect your needs for blood sugar control and heart and kidney care; a possible cancer benefit hasn’t been established. Talk with your health care provider before switching or stopping a drug, and keep up with routine checkups and recommended cancer screenings regardless of which diabetes medication you use. The result is a research signal, not a new treatment instruction.

The study gives researchers a lead to investigate. For now, the established benefits of GLP-1 medications remain the sounder basis for prescribing them. The cancer question is still open.

What researchers should examine next

Future work can test whether the associations hold as more patients accumulate longer treatment histories. It can also examine how the findings vary across levels of diabetes severity and existing complications, since those differences appear relevant here. Studies designed to better separate medication effects from treatment choice would help address the central uncertainty. No single analysis can settle that question quickly.

And yet, the current evidence has a practical use: it helps define what researchers need to investigate, while giving patients a reason to be cautious about sweeping claims. If you're working in this space, the takeaway is that treatment comparisons need to account for why a drug was prescribed in the first place (and this is the part most people overlook). The association is worth studying. Proof of prevention? Not yet.