GLP-1 Medications and Surgery: What the Anesthesia Guidance Actually Says

If you take a GLP-1 medication and you have a procedure coming up, the advice you find online may well be out of date, because the official guidance reversed itself in the space of sixteen months. In June 2023 the American Society of Anesthesiologists told clinicians to hold weekly GLP-1 receptor agonists for a full week before any elective procedure, and by October 2024 a guidance document signed by five organizations had replaced that blanket instruction with individualized risk assessment and advised most patients to carry on taking their medication.

That reversal was not really an admission of error, since it is the ordinary result of collecting outcome data where only surrogate data had existed before. The original concern made perfect mechanistic sense, because GLP-1 receptor agonists slow gastric emptying, food left in the stomach during anesthesia can be aspirated into the lungs, and aspiration pneumonitis is a serious complication. What happened next is that the studies confirmed the first half of that chain convincingly and then failed to confirm the second half at all.

What follows covers what the research measured, why residual stomach contents and aspiration events came apart, and what the current guidance asks your care teams to do instead. This is educational content about published clinical guidance rather than medical advice, and any decision about your medication before a procedure belongs to the clinicians involved in your case.

The concern is pulmonary aspiration, meaning stomach contents travelling up the esophagus and down into the airway while you are sedated or under general anesthesia and have lost the reflexes that would normally prevent exactly that.

Standard preoperative fasting rules exist to make sure your stomach is empty when anesthesia begins, and the usual instruction allows no solid food for six to eight hours and no clear liquids for two hours beforehand. Those intervals were worked out from normal gastric physiology in people not taking any drug that alters it, which is precisely where GLP-1 receptor agonists complicate the picture. Slowed gastric emptying is not an unfortunate side effect of semaglutide or tirzepatide, because it is part of how the medication produces satiety and blunts the glucose rise after a meal, so the reasonable worry was that a fasting window calibrated for a normal stomach might leave a GLP-1 user still holding food.

The delay is real but moderate, and how large it looks depends heavily on the instrument used to measure it.

Hiramoto and colleagues addressed this in a systematic review and meta analysis published in the American Journal of Gastroenterology in 2024. Pooling the scintigraphy studies, which track a radiolabelled solid meal as it leaves the stomach, they found a gastric emptying half time of 138.4 minutes in people taking a GLP-1 agonist compared with 95.0 minutes on placebo, giving a pooled mean difference of 36.0 minutes with a confidence interval running from 17.0 to 55.0.

The measurement modality mattered enormously here. Ten studies covering 411 patients used the indirect acetaminophen absorption test and found no significant delay whatsoever, so the same drug class produced opposite conclusions depending on which instrument was pointed at the stomach. A delay of roughly 36 minutes in the solid emptying half time is genuinely there, but it sits a long way from the hours of stasis that the earliest perioperative discussions implied.

The June 2023 American Society of Anesthesiologists guidance gave a single rule with no risk stratification at all, which was to hold daily dosed GLP-1 receptor agonists on the day of the procedure and to hold weekly dosed formulations one full week beforehand.

That instruction applied irrespective of why you were taking the medication, covering type 2 diabetes and weight management prescriptions equally, and irrespective of the procedure. The document also flagged that holding it for longer than one dosing interval raised glycemic control questions for anyone managing diabetes. It described itself as consensus based rather than evidence based, the standard label for guidance written because clinicians needed something workable before the relevant trials had been run.

The guidance changed because the outcome studies arrived and did not support a blanket hold, while the cost of holding everyone turned out to be higher than anyone had anticipated.

The October 2024 multisociety clinical practice guidance, led by Kindel and colleagues and published across Clinical Gastroenterology and Hepatology and Surgical Endoscopy, replaced the rule with a process. Its central recommendation is that perioperative GLP-1 use should rest on shared decision making between you, the procedural team, the anesthesia team and whoever prescribes the medication, balancing your metabolic need against your individual risk.

Two practical problems drove that shift. Holding a weekly GLP-1 is not a free action, because for anyone using it for glycemic control it means a week of worse glucose control immediately before a surgical stress event, and for anyone still climbing through a titration schedule it can mean restarting escalation and repeating the nausea phase. Procedures were also being cancelled on the morning they were due, frequently for patients who had fasted exactly as instructed and had nothing in the stomach at all.

Residual gastric content goes up substantially while measured aspiration does not go up detectably, and reconciling those two findings is the central puzzle of this entire literature.

On the residual content side, Sen and colleagues published a cross sectional study in JAMA Surgery in 2024 that examined more than 120 patients scheduled for elective procedures who had all followed standard fasting instructions. Preprocedural gastric ultrasonography found increased residual gastric content in 56 percent of GLP-1 receptor agonist users against 19 percent of nonusers, and after adjustment for confounders GLP-1 use was associated with a 30.5 percent higher prevalence. Pooled across eighteen studies and roughly 165,000 patients the association holds up consistently, with reported odds ratios for elevated residual content in the range of 6.

On the outcome side, a 2025 meta analysis by Ho and colleagues in the Journal of the Endocrine Society pooled seven studies covering 541,408 patients, of whom 91,829 were taking a GLP-1. For elective surgery, across three studies and 386,973 patients, the combined risk ratio for aspiration was 1.00 with a confidence interval of 0.76 to 1.30, and for endoscopic procedures, across four studies and 154,435 patients, it came to 1.10 with an interval of 0.95 to 1.27, so neither result reached statistical significance.

What makes that absence convincing rather than merely underpowered is the sensitivity analysis, because the same methodology applied to opioid use returned a risk ratio of 2.68 with an interval of 1.89 to 3.81. The analysis was clearly capable of seeing a real aspiration signal, and it simply did not see one for GLP-1 medications.

Several explanations for the gap are plausible and none is settled, since retained gastric volume may never reach the threshold that actually produces aspiration, anesthesia teams who know about the risk may already be modifying their technique, and aspiration is rare enough that even half a million patients could leave a small absolute increase undetectable. The honest summary is that the surrogate marker moved while the outcome did not, and the current guidance weights the outcome.

The 2024 guidance names specific risk factors instead of treating every GLP-1 user as a single undifferentiated group, and those distinctions turn out to matter a great deal in practice.

The escalation distinction is the one people miss most often. Gastrointestinal effects of GLP-1 receptor agonists are at their most pronounced while the dose is still being increased and they attenuate considerably once you settle at a stable dose, so someone three weeks into a titration and someone eighteen months into maintenance are not carrying the same risk even on an identical milligram amount. The guidance also directs that every patient be assessed on the day of the procedure for symptoms of delayed gastric emptying, which is a far more useful instruction than a calendar rule applied weeks in advance.

It recommends a three part approach applied when delayed gastric emptying is genuinely suspected, rather than a precaution applied to everyone who happens to take the medication.

Diet comes first, and where there is clinical concern about delayed emptying the guidance recommends a liquid only diet for at least 24 hours before the procedure on top of the usual fasting protocol, since liquids leave the stomach far faster than solids and this addresses the specific problem without requiring anyone to stop treatment. Assessment comes second, in the form of point of care gastric ultrasound that can visualize retained contents directly and tell a full stomach from an empty one at the bedside, although the guidance is candid that availability remains limited by institutional resources, operator variability and credentialing requirements. Technique comes third, because if retained contents are suspected or confirmed the guidance suggests considering rapid sequence induction with tracheal intubation to secure the airway quickly, weighed against the alternative of simply cancelling and rescheduling the procedure.

There are three specific places where this evidence stops, and they are worth naming clearly rather than glossing over.

The outcome data are almost entirely retrospective, and large real world database studies detect common events well while capturing rare ones poorly, which matters here because aspiration is rare. No randomized trial has ever assigned patients to hold or continue a GLP-1 before surgery, so the whole outcome literature rests on observation rather than allocation.

The residual content studies used ultrasound and endoscopic visualization at single time points, which tells you that volume was present without telling you whether it would ever have been aspirated, so the link between the surrogate and the outcome remains inferred rather than measured.

Newer and more potent agents are also underrepresented, since most pooled data are weighted toward semaglutide and the earlier agonists. Tirzepatide, retatrutide and the coming generation of multireceptor agonists have not accumulated comparable perioperative datasets, and it would be unwise to assume the gastric emptying profile transfers unchanged across compounds with different receptor activity.

The mechanism behind the original worry was correct, because GLP-1 receptor agonists do slow gastric emptying and people taking them arrive at procedures with more in the stomach than the fasting rules predict. What did not survive contact with the data was the assumption that more stomach contents would translate into more aspiration, since across 541,408 patients it did not, and the same analysis picked up a clear signal for opioids, so that absence was not a failure to look properly.

The guidance that replaced the 2023 rule is more demanding rather than more relaxed, because assessing your symptoms on the morning of a procedure requires real clinical judgment where counting back seven days on a calendar required none at all. It is also more honest about its own footing, since the five societies who signed it labelled it guidance rather than an evidence based guideline, which is the right label for a synthesis of retrospective data on a rare outcome.

If you take a GLP-1 medication and you have a procedure on the calendar, the practical takeaway is not a rule you can apply on your own but a conversation you should start early with whoever prescribes it, whoever performs the procedure and the anesthesia team, all of whom now work from a framework that accounts for your dose phase, formulation, symptoms and other conditions rather than treating every patient identically.

None of this is medical advice, and nothing here should be used to start, stop or adjust any medication on your own.