At a planned endoscopy, the studies found no increase. A meta-analysis of twelve studies and 210,216 patients found pulmonary aspiration in 0.16% of GLP-1 users and 0.12% of non-users [4]. The odds ratio was 1.23, and it was not statistically significant. Emergency care is the exception. After emergency stroke thrombectomy, aspiration pneumonia occurred in 26.8% of GLP-1 users against 10.3% of non-users [1].
The reason for the concern is known. These drugs slow gastric emptying. Food can remain in the stomach after a normal fast, as what to tell the anesthesiologist sets out. Food left in the stomach is not aspiration. Aspiration means stomach contents reach the lungs. The first is more common on these drugs. The second is what the studies below count.
Planned endoscopy: no increase found
Three analyses reach the same result. Alkabbani and colleagues compared 43,365 adults with type 2 diabetes in two US insurance databases [2]. Everyone had used a GLP-1 or an SGLT2 inhibitor within 30 days of an upper endoscopy. Aspiration ran 4.15 per 1,000 on a GLP-1 and 4.26 per 1,000 on an SGLT2 inhibitor. The risk ratio was 0.98 (95% CI 0.73 to 1.31).
Velji-Ibrahim and colleagues matched 46,948 patients with diabetes or obesity [3]. Post-endoscopy aspiration pneumonitis had a hazard ratio of 0.92 (95% CI 0.54 to 1.56). Pneumonia had a hazard ratio of 1.01 (0.83 to 1.24). Neither differed between groups.
The meta-analysis adds a warning inside its null result [4]. Its case reports include three patients who aspirated despite stopping the drug more than six days earlier. All three had fasted for over eight hours. The Zepbound label carries a similar warning, covered in how long Zepbound stays in the body.
Emergency stroke treatment: a higher rate
A stroke gives no notice. Nobody fasts, and nobody holds the drug. Sardana and colleagues reviewed 740 consecutive patients having thrombectomy at one center [1]. Of these, 41 were on a GLP-1 at arrival.
The authors state the rule first. Thrombectomy should not be withheld based on GLP-1 exposure alone. Their findings support targeted aspiration-prevention steps instead. An untreated large-vessel stroke is catastrophic. A pneumonia is treatable.
Aspiration pneumonia within 30 days occurred in 26.8% of GLP-1 users and 10.3% of non-users. The absolute difference was 16.5 percentage points. The matched adjusted odds ratio was 3.25 (95% CI 1.43 to 7.40). Twenty-six point eight percent of 41 is eleven pneumonias. Every estimate rests on those eleven events, which is why the interval is wide. The same caution applies to the overdose comparison.
Two outcomes did not differ. Symptomatic brain bleeding occurred in 4.9% against 9.1%, with an odds ratio of 0.32 (95% CI 0.04 to 2.53). Function at 90 days was also similar, with a common odds ratio of 1.14 (0.58 to 2.25). A complication rose. Recovery did not measurably suffer.
How the two settings fit together
The mechanism is the same in both settings. The circumstances differ. Before a planned procedure, the patient fasts and the team knows the drug is on board. In an emergency, neither is true. The motility evidence is in the gastroparesis guide. How other operations fare is in the surgery outcomes guide.