GLP-1s and Surgery: When to Stop Ozempic Before Anesthesia, and Why the Rule Changed

If you take a GLP-1 drug and have a procedure coming up, the single most important step is to tell your surgeon and anesthesiologist, including whether you are on a GLP-1/GIP combination like tirzepatide. After that, the guidance has two layers that can look contradictory. The 2023 rule was to hold the drug: skip a once-weekly injection the week before, and skip a daily dose the day of. The 2024 multisociety guidance softened that to a risk-based approach, in which most people can keep taking the drug and instead follow a 24-hour liquid diet before the procedure.

The reason both versions are still circulating is that the field changed its mind quickly, and patient handouts have not all caught up. The decision now depends on which drug, what dose, whether you have gut symptoms, and whether you take it for diabetes or weight. This piece walks through what the drugs do, what the data show about actual risk, and how to handle the days before surgery.

In June 2023 the American Society of Anesthesiologists issued the first guidance on the question, and it was a blanket instruction: hold daily GLP-1s on the day of any procedure and weekly ones for a week beforehand, regardless of whether the drug was for diabetes or weight loss. The worry was delayed gastric emptying leaving food in the stomach that could be regurgitated and inhaled under anesthesia. The society was candid that the evidence was thin, limited mostly to case reports.

In October 2024 a five-society coalition, the ASA together with the American Gastroenterological Association, the bariatric surgeons (ASMBS), the perioperative-obesity group (ISPCOP), and the GI endoscopy society (SAGES), reversed the default. Their guidance said most patients can continue the drug, and that the higher-risk minority should use a 24-hour liquid diet rather than simply stop. It went further, warning that stopping the drug solely because a patient takes it for obesity could reflect bias and should be avoided. Through 2025, consensus statements from US, UK, and Australia and New Zealand groups converged on the same idea: continue the drug, modify the fast.

GLP-1 drugs slow gastric emptying. That is part of how they curb appetite, and it is also the root of the surgical concern: after the standard overnight fast, a GLP-1 user is more likely to still have food or fluid in the stomach. The data on this point are consistent and strong. In an endoscopy study at Cedars-Sinai, retained food showed up in 17.4% of GLP-1 users versus none of the controls. Meta-analyses pooling tens of thousands of patients put the odds of retained gastric contents roughly three to five times higher in GLP-1 users, with aborted upper-endoscopy procedures about five times more common.

The first randomized trial, OCULUS, published in JAMA Internal Medicine in 2026, made the causal link directly: among patients having upper endoscopy, 25% of those who continued their GLP-1 had a clinically significant amount of fluid left in the stomach, versus 3.1% of those who held it. The trial was stopped early because the difference was so clear. So the question is not whether these drugs leave more in the stomach. They do. The question is whether that translates into actual harm.

Here the evidence gets weaker and more mixed, which is the part most coverage skips. Aspiration, where stomach contents are inhaled into the lungs, is the outcome that actually matters, and it stays rare. The two largest surgical studies, covering 366,476 and 392,065 patients, found no statistically significant increase in aspiration pneumonia among GLP-1 users (adjusted odds ratios of 0.78 and 1.07). Most endoscopy meta-analyses found the same, with aspiration rates hovering around 0.1 to 0.2% and confidence intervals crossing one.

The signal is not entirely absent. One large Gastroenterology cohort of 963,184 endoscopy patients did find a modest but real increase, with aspiration pneumonia in 0.83% of GLP-1 users versus 0.63% of non-users. And case reports document aspiration even in patients who stopped the drug for more than six days and fasted overnight. The honest summary is that retained stomach contents are common, but progression to actual aspiration is uncommon and serious, and the true magnitude is still uncertain. That gap, between a clear surrogate signal and a faint outcome signal, is exactly why the societies pulled back from blanket stopping.

If you and your team decide to hold the drug, the timing depends on how long it lingers. Weekly injectables (semaglutide as Ozempic or Wegovy, tirzepatide as Mounjaro or Zepbound, dulaglutide) are typically held for at least a week. Daily and short-acting drugs (oral semaglutide, liraglutide) are usually held the day of or 24 to 48 hours ahead.

The one-week figure is contested. A Mayo Clinic Proceedings review noted that gastric retention only returned to normal after a two-week pause, and an orthopedic study presented at the 2025 AAOS meeting recommended stopping 14 days before total joint replacement. The AGA, meanwhile, has acknowledged it is uncertain whether holding a single weekly dose meaningfully restores normal gut motility at all. So even within the hold camp, one week may be too short, while a Houston Methodist analysis found solid-food emptying was delayed only about 36 minutes and pegged aspiration risk near three in 100,000, arguing the seven-day hold is unnecessary for many. The lack of a clean answer is real, not a gap in this summary.

Why you take the drug matters for the decision. For someone using a GLP-1 to manage type 2 diabetes, stopping it for a week can meaningfully worsen blood-sugar control around surgery, and in patients with cardiovascular disease the perioperative consequences can extend to blood pressure and fluid balance. That is the core reason the 2024 guidance warned against withholding the drug solely because the indication is obesity, without a specific risk factor.

The practical version: a person on tirzepatide purely for weight loss has more room to pause it safely than a person on semaglutide for diabetes, whose glucose may swing if the drug stops. Diabetes patients who do follow a liquid-diet prep are often told to choose sugar-free or low-sugar clear liquids so the fast does not destabilize their blood sugar. These are calls to make with the prescriber, not from a package insert.

Upper endoscopy is where retained food matters most, because the scope enters the stomach directly and food there can both obscure the view and raise aspiration risk. For these procedures, teams often add a clear-liquid diet for 24 to 48 hours, and they reschedule if a patient reports nausea, vomiting, or a feeling of fullness that suggests food is still sitting in the stomach.

Colonoscopy is more forgiving. The bowel preparation is already a liquid-only purge, so the stomach tends to clear anyway, and holding the drug solely to reduce retained contents is often unnecessary. Where a patient cannot stop the drug, a two-day prep helps ensure the colon is clear. The upshot is that the same drug warrants more caution before an upper endoscopy than before a colonoscopy.

Forgetting to hold the drug is not an automatic cancellation. The standard response is to treat the patient as having a full stomach and adjust the anesthetic: a rapid-sequence induction, head elevation, and related precautions that lower the chance of regurgitation. Many centers can also run a quick point-of-care gastric ultrasound on the day of surgery to see whether the stomach is actually empty; if it is, the procedure proceeds, and if it is not, the team uses full-stomach precautions or, for elective cases, may delay.

The same logic covers emergencies. When surgery cannot wait and the drug cannot be stopped in time, the anesthesiologist simply assumes a full stomach and takes those precautions during the operation. The point of disclosing the drug is to let the team choose the right approach, not to get the procedure called off.

The practical checklist is short. Tell your surgeon and anesthesiologist you take a GLP-1, and name the specific drug and dose. Ask, well before the date, whether they want you to hold it or continue with a liquid-diet prep, since both are now legitimate paths and the answer depends on your drug, your dose, and your reason for taking it. If you are in the dose-escalation phase or having gut symptoms like nausea or vomiting, raise that, because active symptoms are the strongest reason to delay an elective procedure.

The common mistakes run in both directions. Stopping a diabetes GLP-1 for too long can cost you glucose control for a fixed surrogate concern; not disclosing the drug at all takes the liquid diet, the ultrasound, and the modified anesthetic off the table. The field has moved from a single blanket rule to a conversation, and the patient's job is to start that conversation early.