A 2026 cohort study asked whether taking a GLP-1 drug in the month before breast cancer surgery changes what happens afterwards [1]. It covered 750 patients at one tertiary center between January 2024 and August 2025 (n=750), of whom 209 had active GLP-1 use verified against pharmacy records rather than self-report. The reported answer is no association, and that answer needs reading carefully — as it does across the rest of the perioperative evidence.
After adjustment, surgical site infection came out at an adjusted odds ratio of 1.28 (95% CI 0.81–2.03). Seroma requiring aspiration was 1.22 (95% CI 0.83–1.79). Flap necrosis of grade II or worse was 1.41 (95% CI 0.83–2.40). None of those is statistically significant, because each interval crosses 1.
Every point estimate also sits above 1, and every upper bound is substantial. This is a study that did not detect a difference, not a study that showed there is none. With 209 exposed patients, an increase of the size those upper bounds allow would be invisible here. A reader looking for reassurance gets some; a reader looking for proof of safety does not.
The exposed patients also differed before anyone operated: higher body mass index, more diabetes. The authors adjusted for that and applied inverse probability weighting, which is the right response and not a cure — the groups were unlike in ways that are measurable, and possibly in ways that are not. The same caution applies to the joint work in knee replacement.
None of this touches the reason anesthetists ask about these drugs, which is delayed gastric emptying and aspiration risk at induction rather than wound healing weeks later. Those are separate questions with separate evidence, and this study addressed the second.
The practical point is unchanged by the statistics: tell the surgical team you are taking one. That is the same advice the shoulder surgery work reaches by a different route, and it does not depend on which way these three odds ratios eventually settle. Sellers rarely raise it, which is part of the wider gap catalogued in what sellers do not publish.