One study that has received a lot of attention coning out the American Society of Anesthesiologists’ ANESTHESIOLOGY 2026 meeting was “Impact of GLP-1 Receptor Agonists on Postoperative Major Adverse Cardiovascular Events and Infection: A Cohort Study.”
“The use of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) has increased since their approval for type 2 diabetes mellitus (T2DM) and weight loss,” the study abstract states. “Beyond glycemic control, these agents can reduce the incidence of major adverse cardiovascular events (MACE). However, their effects in the perioperative setting remain uncertain, where hemodynamic fluctuations, altered physiology, and surgical stress may modify their risk-benefit profile. We evaluated the association between perioperative GLP-1 RA use and postoperative MACE and infection.”
The retrospective cohort study used the EPIC COSMOS database—a centralized, de-identified EHR data warehouse created by aggregating data from more than 1,700 hospitals and 40,000 clinics that use the Epic EHR. It included adults 18 years and up undergoing elective surgeries lasting less than an hour with 24 or more hours of postoperative hospitalization from January 2016 to December 2024.
The study’s outcomes of interest were MACE—defined as a composite outcome including myocardial infarction, myocardial injury, stroke, and all-cause mortality—and infection, including surgical site infections, UTI, pneumonia, and sepsis. It included 2,028,601 patients, of whom 145,610 (7%) received GLP-1 therapy. After propensity score matching, MACE occurred in 1,140 of 114,249 GLP-1 users (1%) and 3,908 of 360,168 (1.1%) of non-users, while postoperative infections occurred in 5,652 of 114,249 (4.9%) of users and 19,039 of 360,168 (5.3%) of non-users.
The study concluded that “perioperative GLP-1 use was associated with reduced odds of major adverse cardiovascular events and postoperative infections in elective noncardiac surgery. These findings suggest the perioperative benefit of GLP-1 use. Prospective studies are required to examine protocols for GLP-1 usage in the perioperative period.”
The study was performed by clinical research fellow Josephine Tolan, MD, and her colleagues at the Center for Outcomes Research, Department of Anesthesiology, Critical Care and Pain Medicine, McGovern Medical School at UTHealth Houston. Periop Leader Network recently spoke with Dr. Tolan about its findings.
Let’s put your study in context. What’s the current perioperative standard of care for handling patients who use GLP-1 medications, and how could or should your study’s results potentially change that?
Dr. Tolan: There’s a lot of interest in this topic because it’s actively evolving as we learn more about these medications and how they affect patients in different ways. The current guidelines for anesthesia according to ASA are for an individualized approach for management of these medications perioperatively. It depends on a lot of things: the medication they’re taking, how recently they’ve had a dosage increase or been started on the medication, the surgery they’re having, the kind of anesthesia they’ll be going under, the position they’ll be in during the surgery.
All of these things can change their risks of the big concern around GLPs, which is aspiration. When you take GLP medications, it slows your gastric emptying time. Your stomach is fuller for longer after a meal, which can mean the risk of having some food in your stomach after the normal eight-hour fasting window. That’s the big concern and the big risk around taking GLPs. The problem is, these medications have a long half-life, and finding the exact window of how long ahead of time they should be paused is difficult. In emergent cases, that’s not always possible. So we’re still doing a lot of research to understand what best practices should be and what other options we have besides just holding the medication to address that concern of fasting.
With all of that in mind, your study puts that to the side to focus on cardiovascular risks.
Josephine Tolan, MD, Clinical Research Fellow, Center for Outcomes Research, Department of Anesthesiology, Critical Care and Pain Medicine, McGovern Medical School, UTHealth Houston
Right. One of the things we think about is, yes, there’s a risk for this medication, but as we’re putting new practices in place and figuring out the best fasting window and all of that, we need to remember people are taking this medication for a reason and that it’s beneficial in a lot of ways as well. A lot of new research has been done in the basic science space that looks into more of the mechanism of action of these medications. Prospective studies in the general population—not perioperative population, but just general people taking GLPs—are seeing benefits in things other than the FDA-approved benefits of metabolic benefit and weight loss. We’re also seeing in the general population studies showing improvement of cardiovascular disease and things like that, and some basic science looking at the mechanisms of how that’s happening.
That raises the question of well, could there be benefits, and how will that change our decisions when it comes to whether or not these patients should hold a medication or not? One of the things we were interested in was that we have seen in the general population there might be this benefit of cardiovascular outcomes. Is that going to be seen in a perioperative setting where you have other risks for cardiovascular disease besides the baseline due to the stress of the operation and the anesthesia themselves?
The interesting finding was we actually did see that protective effect might be maintained, but this was in isolation of those other risks. So our findings are not saying everyone having surgery should be taking a GLP-1, not at all. It’s just providing insight that this benefit might be seen, and we can take that into account when we design future prospective studies, looking at outcomes for patients on GLP-1s, as well as trying to design that ideal protocol for fasting.
Our bigger question was: Should we be taking into account potential benefits of this medication in our fasting? So, if a patient is already on this medication for other medical reasons, is it more beneficial to continue the medication and find another way to address the risk of aspiration other than holding the medication entirely—like a longer fasting window or a clear liquid diet for 24 hours or some other option instead, rather than holding the medication for a period of a couple weeks beforehand? Because then we might negate this benefit.
Did you distinguish between patients who were taking GLP ones for obesity and patients who were taking it for diabetes? Did you distinguish among different types of GLP-1s? How granular did you get?
This was a retrospective study, so we were using a database to look backwards in time. The big limitation to that is we don’t always have all of the data we would have in an ideal world. You have to deal with the data that’s put into this database in real time by doctors who don’t know the study you’ll be doing in the future, so we unfortunately didn’t have some of that information.
I think the biggest gap we had was that we didn’t know which protocol the surgeon used for fasting. We just knew the patients were taking this medication. We didn’t know if they had it held or continued the medication up to the point of surgery. For some patients we knew which medications they were taking, and for others we didn’t. So we weren’t able to do an in-depth analysis on the specific granular medication differences.
What we were able to gather and say definitively is that these patients having non-cardiac surgeries were prescribed and taking GLP-1 medications prior to their surgery and had that exposure. Then we were able to follow them afterwards for a period of time to see which of those potential markers of cardiovascular risk we saw. We were able to gather information about type of surgery being performed, age, race, all these things you try to control for, and were able to try and make our exposure group and control group similar with some of our statistical analysis.
This is why you can’t take retrospective studies and apply them to clinical practice directly. They’re mostly there to help give you information of some correlations that exist so you can then design studies that delve deeper into the cause and effect behind those correlations. Because of all the limitations that existed, I don't think our study can definitively say, “these GLP-1 medications have this effect in the perioperative period.” It can say, “We see that patients taking GLP-1s have improved outcomes, and we’re interested in seeing more about why, and looking at how that can impact our clinical care moving forward.”
This was a pretty good effect, with 11% odds reduction. Moving forward, I’m interested in looking at a prospective design study where we’re able to gather more of that information we didn’t have in the database so we can get more granular in the question we’re asking, and having more control over those variables to know for sure what’s causing the outcome you’re seeing. That would be the study that would inform practice.
What reaction did you get when you presented your study?
Presenting it to the other researchers there, the big thing they were excited about was the database and how large it was. It’s a fairly new database that has millions and millions of patient records, and a lot of information in it. A lot of people were really excited about the potential for the types of retrospective studies you can do with this database.
Retrospective studies, while they aren’t authoritative, serve an important role in the research process by allowing us to fairly cheaply and quickly look at a question and determine if enough is there to support a more expensive and more informative prospective trial. For a lot of people, the first step to looking at a question is going to be a retrospective study to show whether or not there is an effect that’s worth looking into. A lot of people did have that reaction that “this is really interesting, but I want to know more about the actual clinical outcome this could have, particularly since your question didn’t really look at the risk of aspiration. We’re excited about this type of study design, but more needs to be done and looked into to find how this could actually affect our practice.” Studies like this can be used to write grants that help fund the studies that are more expensive to do.
With that in mind, how far along are we on the knowledge curve when it comes to the impact of GLP-1s in a perioperative sense? How much do we not know as compared to what we know
I think there’s a lot more known in the general population than the perioperative population for these medications, but these populations can be very different because of the risks you see in a perioperative setting. It’s important to look at this population in isolation. You can make some inferences based on what we’re seeing in the general population.
This study was really driven by us seeing some studies that showed that the general population does have improvement in these outcomes. I think it’s important to study it on its own. In that regard, there’s not a ton out there yet. I know a lot of people are actively working on it, though. The majority of the focus has been on the fasting part of it, and we're still pretty early in doing those studies as well.
These medications are so new, and research takes a while to do properly. So there’s still a lot to learn.
What should perioperative leaders take away from your study?
I think a big takeaway is to remember that there are benefits to these medications. A lot of people I know in the perioperative space are very concerned about the fasting side of it. You don’t want to cause an aspiration. That’s a really big complication, and I don’t want to minimize that at all. But I think we are still early in our understanding of how these medications could be beneficial, so keep an eye on the literature, see what we learn, and keep an eye on the guidelines.
The initial guidelines a couple years ago had been “hold these medications for at least a week for all patients.” That has now shifted to “approach this on a patient-to-patient basis—look for these particular factors that might make a patient more at risk for aspiration, and then make a patient-to-patient decision on it.” That’s a sign to me that people are starting to get more open to the idea that there might be a benefit to this, and maybe we don’t understand the fasting as well as we want to yet.
The big message is to keep an eye on the ASA. They’re really good at updating those guidelines once robust data is available. Be aware of data that’s coming out and use that to guide your practice.
What’s next for you? Wil you be performing additional studies?
Our group right now is doing a randomized control trial of fasting protocols for GLP-1 medications. It’s designed to be in patients getting endoscopies who are taking GLP-1 medications randomized to either holding the medication for seven days, holding the medication for seven days doing a normal eight-hour fast, or continuing to take the medication but doing a 24-hour clear liquid diet instead. Prior to their procedure, we do an ultrasound of their stomach and see if we can quantify how full it is.
We’re still pretty early in that. We don’t have any data to share for that, but that’s the study we decided to design after seeing some of this data coming out from our retrospective, which showed us there is utility to examining what fasting protocols would work to allow patients to continue taking their medication without a pause.
The problem with randomized control trials is they take a long time to do. I need to recruit a lot of patients. So it’ll be a couple years yet, probably, but we will happily share that once we have it.
Any final thoughts?
I really appreciate the interest in the study. I was really surprised; I did not expect so much media attention post-meeting from this, and it’s cool to have so much interest. I think some people are aiming for the headline of, “GLP-1 is a miracle, it protects you.” I’m hesitant to say that, or that our study should definitively change practice. But I am excited for where it can lead and how it will lead research directions to hopefully eventually change practice and allow us to provide the best care for our patients.
