GLP-1 drugs and cancer risk: what the evidence actually shows

What did @dr.yousufzafar actually say?

An oncologist made a reasonable, hedged case that GLP-1 receptor agonists may modestly reduce the risk of certain cancers, particularly those linked to obesity or hormonal activity. He pointed to three potential mechanisms: weight loss, improved insulin control, and reduced chronic inflammation. He also said this is early science, mostly observational, and that these drugs are not cancer treatments.

That's a more careful framing than most TikTok health content. He didn't claim GLP-1 drugs cure cancer, and he explicitly said "there's no evidence that it can help somebody who already has cancer." He also flagged that the evidence is stronger in women than in men, and that long-term data simply doesn't exist yet. For a 90-second video, that's a lot of appropriate caveats packed in.

Does the science back this up?

Yes, mostly, but the picture is more complicated than the video suggests. The observational data is real, the effect sizes are modest, and the mechanistic explanations are plausible. But we're still far from understanding causation.

A 2024 study published in JAMA Network Open (Shi et al.) analyzing over 1.6 million patients found that semaglutide was associated with a reduced risk of 10 out of 13 obesity-related cancers compared to non-GLP-1 diabetes medications. That's a headline number, but it comes with a major asterisk: people who lose weight and stick with a medication regimen are systematically different from those who don't, and that confounding is hard to fully control for in observational work.

On the mechanistic side, the insulin-cancer connection is well-established. Hyperinsulinemia promotes tumor growth through IGF-1 signaling pathways (Pollak, 2008, Nature Reviews Cancer). The inflammation angle also has real support. Chronic low-grade inflammation, the kind associated with visceral obesity, is a known driver of carcinogenesis. Whether GLP-1 drugs reduce inflammation directly or primarily through weight loss is still being worked out.

What did they get wrong (or right)?

He got the broad strokes right. The three mechanisms he named are the same ones researchers actually discuss. He got the epistemic humility right too, which is rarer than it should be.

Where the video falls slightly short is in presenting weight loss, insulin control, and inflammation as three parallel explanations when they're actually deeply intertwined. Better insulin control is partly a consequence of weight loss. Lower inflammation is partly a consequence of both. These aren't independent levers, and presenting them as a clean three-item list implies more mechanistic clarity than the field actually has.

He also said "the evidence is stronger in women than in men" without explaining why. That's worth expanding. Obesity-related cancers include several hormone-dependent cancers, such as postmenopausal breast cancer and endometrial cancer, that simply have no male equivalent. The sex difference in the data probably reflects cancer type distribution more than a true biological sex effect of the drug itself. That's a meaningful distinction that got glossed over.

What should you actually know?

This is promising early-stage science. It does not mean GLP-1 drugs are cancer prevention tools you should add to your regimen for that reason alone.

The studies showing reduced cancer risk are observational, meaning they track what happens in the real world rather than randomly assigning people to treatment. That design can't prove the drug caused the benefit. People prescribed GLP-1 drugs may have better healthcare access, more consistent follow-up, and more health-motivated behaviors across the board. A 2023 review in The Lancet Diabetes and Endocrinology (Lega and Lipscombe) emphasized exactly this limitation, noting that confounding by indication remains a serious problem in this literature.

The one high-quality randomized trial data we have, from cardiovascular outcomes trials like LEADER and SUSTAIN-6, were not designed to study cancer and had follow-up periods too short to draw meaningful cancer conclusions. The randomized evidence on cancer specifically does not yet exist at the scale needed.

If you're on a GLP-1 drug for weight management or type 2 diabetes and you've heard this research, it's reasonable to find it interesting. It is not a reason to start one of these medications if you wouldn't otherwise be a candidate. And it is not, under any interpretation, a substitute for cancer screening.