GLP-1s and Surgery: What the Aspiration Evidence Actually Shows

Tell them. That is the whole headline.

If you are on a GLP-1 and you have anything booked that involves sedation or a general anaesthetic — surgery, a colonoscopy, an endoscopy, a dental procedure under sedation, egg retrieval — the anaesthesia team needs to know, and they need to know before the day.

That is the actionable part, and it is at the top because plenty of women have not been asked. Weight-loss medication does not always make it onto a pre-op form, and if the form does not ask, nobody finds out.

The rest of this page is why it matters, and why the honest answer is more interesting than either version you will have seen.

Two questions, two different answers

Almost all the coverage of this collapses two separate things. Keep them apart and the evidence makes sense.

Question one: does a GLP-1 leave food in your stomach even when you have fasted properly? Yes, dramatically more often.

Question two: does that translate into more people actually inhaling stomach contents under anaesthetic? Not in the data so far.

A 2025 systematic review and meta-analysis in Anaesthesia screened 9,010 studies and pooled 28 observational ones, answering both¹:

That is an odds ratio, and it is worth not over-translating: because leftover stomach content is common rather than rare, an odds ratio of 5.96 overstates how much more often it happens. The JAMA Surgery ultrasound study below puts it on the risk scale directly — an adjusted prevalence ratio of 2.48 (95% CI 1.23–4.97), or about 30% higher in absolute prevalence³. So: substantially more often, not six times as often. And no detectable increase in the event itself, across 471 aspiration cases.

Both findings were rated low certainty by the GRADE framework, which matters in both directions — it is not strong evidence of harm, and it is not strong evidence of safety.

What "food still in there" actually looks like

The most striking single study used gastric ultrasound on the morning of surgery. Researchers recruited 107 patients who had taken once-weekly semaglutide within 10 days of their procedure, and 113 who had not, all fasted to current guidelines².

Increased residual gastric content: 43 of 107 (40%) on semaglutide, against 3 of 113 (3%) without. In propensity-weighted analysis the odds ratio was 36.97 (95% CI 16.54–99.32).

Read that again: two in five, after doing the fasting correctly.

And the finding the authors set out to test: stopping the drug for 10 days or less was not enough to normalise it. Fasting to the letter of the guidelines did not empty these stomachs.

Two other factors came out independently associated with leftover content in the same analysis, and one is worth knowing on a page written for women: male sex (OR 2.28, 95% CI 1.29–4.06), and age (OR 0.95 per year, 95% CI 0.93–0.98)². Being female was associated with lower odds, not higher. That does not change the need to tell your anaesthetist, but it is a small piece of reassurance in a subject with very little of it.

A separate cross-sectional study in JAMA Surgery used the same ultrasound method in 124 patients following standard pre-procedure fasting, and also examined how the duration of drug interruption related to residual contents³.

So why is aspiration not going up?

This is the genuinely interesting part, and there is no settled answer. Three plausible pieces:

Aspiration is rare to begin with. Even 471 events across 185,414 patients is a low rate. Detecting a modest increase in a rare outcome takes enormous numbers, and these were observational studies, not trials.

Anaesthetists are already adapting. Once a patient is flagged as being on a GLP-1, the team can change technique — gastric ultrasound before induction, a rapid-sequence approach, treating the stomach as full regardless of fasting time. A risk that gets managed stops producing events. That is a success, not an absence of risk, and it depends entirely on them knowing.

The studies are observational. They reflect real practice, including all the precautions taken in it.

Which is why "aspiration is not increased" is not a reason to stay quiet. The most likely explanation for the reassuring number is the precaution.

What this means for you, practically

• Say it out loud at every stage — booking, pre-op assessment, and again to the anaesthetist on the day. Do not assume it transferred between them.
• Say the name of the drug and your last dose date. "I'm on a weight-loss injection" is not enough; the interval is the thing they need.
• Ask specifically what they want you to do about the dose. Whether and when to hold it is their call, not something to decide from an article — including this one.
• Do not stop it on your own initiative before checking, particularly if you also have diabetes, where an unplanned interruption has its own consequences.
• Ask whether they use gastric ultrasound. Some centres do, and it turns the question from guesswork into a look.
• Mention it for sedation too, not just full anaesthesia — colonoscopy and dental sedation count.

A multisociety clinical practice guidance document on the perioperative use of these drugs now exists and is worth your clinician knowing about⁴. What it recommends for your dose, your procedure and your history is a conversation for them to have with you.

The symptom that changes the conversation

If you are vomiting, feeling persistently full, or bringing up food you ate many hours earlier in the days before a procedure, tell them that specifically. Those are signs your stomach is emptying slowly right now, and they are more informative than any fasting rule.

Delayed gastric emptying is the established mechanism behind all of the above, and it is dose-related and variable between people⁵.

Where this sits on the evidence scale

Well-quantified, but formally low certainty — and I am using the reviewers' grade, not a friendlier one. Both pooled outcomes were rated low certainty by GRADE, and the withholding result very low. What is unusually good here is the quantity rather than the grade: twenty-eight observational studies pooled, a consistent and very large effect on residual gastric contents, a null result on aspiration across 185,414 patients, and formal GRADE assessment of both. Nobody is guessing.

The limitation is the one the reviewers named themselves: it is all observational and rated low certainty. There are no randomised trials here and there probably will not be, because randomising people to an aspiration risk is not something an ethics committee will approve.

One more pooled result from the same review belongs here, because it is the most actionable thing in it and because leaving it out would let the paragraph above sound bleaker than the evidence is. Across five studies, 1,706 patients and 208 cases, withholding at least one dose before a procedure was associated with lower odds of residual gastric contents — odds ratio 0.51 (95% CI 0.33–0.81), at very low certainty¹.

So the honest summary is that whether to withhold has a signal in its favour; how long does not. The 10-day finding says a short interruption is not automatically enough, and nobody has established what is.

What would sharpen it: a study establishing what interruption length actually normalises gastric contents, and per-molecule and per-dose data rather than pooling across drugs. Until someone answers it, "ask your anaesthetist" is not a dodge. It is the honest state of the field.