Holding a GLP-1 Before Endoscopy or Surgery: What the Trial Found

These drugs slow the stomach deliberately, which is a problem when someone is about to sedate you. In 2026 a randomized trial finally tested what to do about it: patients who held a single dose before an upper endoscopy had clinically significant retained stomach contents 3.1% of the time, against 25.0% for those who continued.^[1] The trial was stopped early because the gap crossed its own safety boundary. But the most useful thing it found was about the prep, not the drug — and almost nobody quotes it.

Why a slowed stomach matters here specifically

Sedation removes the reflexes that keep stomach contents out of your airway. Anesthetic fasting rules exist to make sure the stomach is empty by the time that happens, and they were written for stomachs emptying at a normal rate. A drug whose whole mechanism includes slowing gastric emptying can leave food behind after a fast that would ordinarily be sufficient.

The consequences run from inconvenient to serious: an examination the endoscopist cannot complete, a procedure abandoned partway, an unplanned breathing tube, or aspiration.

What the randomized trial found

OCULUS ran at two US referral centers, randomizing adults on a stable GLP-1 or GLP-1/GIP dose to hold one dose before an upper endoscopy or to continue.^[1] Its endpoint was a composite of the things that actually disrupt a procedure: contents preventing examination, premature termination or intubation, or an aspiration event needing extra monitoring or admission.

The finding almost nobody quotes

In the subgroup having an upper endoscopy plus a colonoscopy — 25 patients, who were therefore on clear liquids the day before as standard bowel prep — not one patient had clinically significant retained contents. Not in the continue arm, not in the hold arm.^[1]

Nobody on clear liquids the day before had a problem, whether they held the drug or not.

The trial's own conclusion draws that out: clear liquids the day prior may mitigate the risk regardless of GLP-1 or GIP use.^[1] If that holds up in larger numbers, the practical answer to this whole question may turn out to be a diet instruction rather than a drug hold — which matters, because holding a dose has its own costs for someone whose blood sugar depends on it.

Retained food is not the same as aspiration

This distinction is where most coverage goes wrong. A meta-analysis of 23 studies covering 262,018 patients found GLP-1 users had substantially more retained gastric contents — an odds ratio of 4.54 (95% CI 3.30–6.24) — and more procedures terminated early.^[2]

And it found no significant difference in aspiration pneumonia.^[2] That is the outcome people are frightened of, and across a quarter of a million patients it did not separate.

Both facts are real and they support different actions. Retained contents four and a half times as often is a good reason to change how procedures are prepared for. It is not evidence that people on these drugs are being harmed at scale, and a page that reports the first number without the second is telling you half of it.

A second, independent meta-analysis published in Anaesthesia reached the same split from a separate set of studies, which is the strongest form this evidence can currently take.^[4] Two teams, different literature searches, same dissociation.

An odds ratio of 1.04 with an interval from 0.87 to 1.25 is about as flat as a result gets. Six times the retained food, and no detectable change in the harm that retained food is feared for.

Does skipping a dose help?

This is the question patients actually ask, and the same review provides the only pooled answer available. Across five studies and 1,706 patients, withholding at least one dose before a procedure was associated with lower odds of residual gastric contents — an odds ratio of 0.51 (95% CI 0.33–0.81).^[4]

What the guidance says, and why it keeps changing

The American Gastroenterological Association issued a Rapid Clinical Practice Update on this, explicitly framed as timely guidance on a fast-moving question rather than a settled standard.^[3] Anesthetic and gastroenterological societies in several countries have since published their own, and they do not agree in every detail.

That disagreement is not incompetence; it is what an evidence base looks like while its first randomized trial is still being absorbed. The practical consequence for you is that the instruction you get depends on which society your proceduralist follows.