If you take a GLP-1 drug and have an operation booked, this is the study to know about. Across 119 UK hospitals and 47,039 patients, regurgitation or stomach contents entering the lungs happened to about 1 in 71 people taking a GLP-1 drug, against about 1 in 802 of everyone else. The events clustered at the point where patients were waking up. The work was published in Anaesthesia on 16 September 2026 and is the largest prospective look yet at a risk that drug labels already warn about.
What the study found
The GLIMPSE investigators screened adults having elective or emergency procedures under the care of an anaesthetist and asked whether they had used a GLP-1 receptor agonist in the previous three months. Of 47,039 patients, 1,348 had — 2.9%, or roughly 1 in 36 people on an operating list. That figure alone is part of the finding: this is no longer a rare situation for an anaesthetic team to handle.
Pulmonary aspiration or regurgitation occurred in 19 of those 1,348 patients, or 1.41% (95% CI 0.85% to 2.19%). Among the 45,691 patients not taking the drugs it occurred in 57, or 0.12% (95% CI 0.09% to 0.16%). That is an odds ratio of 11.39 (95% CI 6.47 to 20.05, p < 0.001). Put in absolute terms, the difference is about 1.3 extra events per 100 patients — roughly 13 per 1,000. The authors also found that how these patients were managed varied widely from hospital to hospital: whether the drug was stopped beforehand, and which anaesthetic and airway techniques were chosen, differed between sites. Teams were clearly trying to reduce the risk, and the events still happened, most often during emergence from anaesthesia rather than at the start.
How it fits what was already known
The mechanism is not in dispute. GLP-1 drugs slow gastric emptying, which is part of how they curb appetite, and food left in the stomach is the raw material for an aspiration event. That reasoning is already written into the products: liraglutide labelling carries a Warnings and Precautions section on pulmonary aspiration during general anesthesia or deep sedation, and the same warning language has been added across this drug class. Earlier coverage of GLP-1 drugs on this site reported an adolescent study in which most patients still had stomach contents after a long fast. What GLIMPSE adds is the outcome rather than the mechanism: not how many stomachs were full, but how many people actually aspirated.
It also puts a size on a risk that has been argued about since anaesthetic societies started issuing advice on these drugs. An eleven-fold odds ratio sounds alarming; 1.41% is the number to hold onto alongside it. Both are true, and they describe the same thing from different ends.
What is not yet known
This is an observational cohort, not a randomised trial, and the comparison reported is a crude one. People taking GLP-1 drugs differ from people who are not in ways that independently raise aspiration risk — obesity, diabetes, gastro-oesophageal reflux, the type of surgery they are having. The abstract does not report an adjusted estimate, so some unknown share of the eleven-fold figure belongs to those differences rather than to the drug. The event count is also small in absolute terms: 19 events drive the entire GLP-1 estimate, which is why the confidence interval runs from roughly 6-fold to 20-fold. The study cannot say which management strategy works, because it observed practice rather than testing it, and it describes UK practice in one period. None of that changes the direction of the finding. What the study documents plainly is that practice varied — including whether the drug was stopped beforehand — and that no strategy observed here was tested against another.
