These drugs work in part by slowing the stomach. That is the mechanism, not a side effect, and it is why a smaller meal feels like enough. The question this page is about is what happens at the far end of that mechanism, when the stomach slows past the point of usefulness and into a diagnosis with a name, an imaging threshold and a treatment pathway of its own — and it is a genuinely different question from the ordinary nausea of a dose increase, which the titration page covers and which resolves on its own in most people.
What the published cases show
A systematic review published in 2026 searched PubMed, Embase, Scopus and Web of Science from inception to October 2025 for symptomatic gastroparesis associated with these drugs, and found twelve case reports describing thirteen patients. [1] Most were women, most had type 2 diabetes, and cases occurred in people without diabetes too. Semaglutide was the agent implicated most often, followed by liraglutide, dulaglutide and exenatide — which is roughly the order of how widely each is prescribed, so the ranking tells you less than it appears to.
Symptom onset ranged from hours after a dose to several months into treatment, and it commonly followed either a dose escalation or a restart after a break in therapy — the same two moments that a titration schedule is built around. Reported symptoms were nausea, vomiting, abdominal pain, bloating, early satiety and an inability to keep food down. Imaging or endoscopy often showed a distended stomach still holding its contents, with no mechanical blockage to explain it.
How the diagnosis is actually made
Gastroparesis means delayed gastric emptying with no outlet obstruction, and the standard test is gastric emptying scintigraphy: a labeled meal, then a measurement of how much is still in the stomach four hours later. [2] More than 10% retained, in someone with the symptoms and without a blockage on endoscopy or a CT scan, meets the 2022 American College of Gastroenterology and 2025 American Gastroenterological Association definition. Severity is banded off the same measurement — 10% to 15% is mild, 16% to 35% moderate, and above 35% severe. A carbon-13 breath test is also FDA-approved for the diagnosis.
The condition itself is uncommon. A 2018 claims analysis put the prevalence of documented gastroparesis at 21.5 per 100,000 people, and it runs two to four times more often in women than men. In a US epidemiological study of 82.6 million patients, the causes broke down as type 2 diabetes 51.7%, postsurgical 15%, medication-induced 11.8%, idiopathic 11.3%, type 1 diabetes 5.7% and other 4.5%. Medication-induced is the category these drugs fall into, alongside opioids, cannabis and anticholinergics.
What happened to the reported patients
Management in the published cases was mostly supportive — fluids, antiemetics, and letting the stomach recover. Discontinuing the drug resolved the symptoms in every patient reported. That is the most useful sentence in the review, and it is also the sentence most easily over-read: thirteen people all got better after stopping, which is encouraging and is not the same as a demonstrated reversibility rate. What it does establish is that stopping is the intervention, and that the outcomes published so far have been good.
The general treatment ladder is unchanged by the cause. First-line for mild disease is a small-particle diet low in fat and indigestible fiber, plus antiemetics. Moderate adds prokinetics such as metoclopramide or erythromycin. Severe may need liquid or jejunal feeding, and refractory cases are sometimes treated with a pyloric myotomy or an implanted gastric stimulator. None of that is specific to a GLP-1, and almost none of the published cases needed any of it.
What this means before ordering
The practical content here is narrow and worth stating plainly. Dose increases and restarts are the moments to watch, vomiting that stops you eating is not the ordinary nausea of week three, and the drug is prescribed — which means there is a clinician to call. Whether the seller you are buying from makes that easy is a separate question, and most of them publish very little about it. A discontinuation that goes through a support inbox is a slower discontinuation than one that goes through a prescriber.
There is also a cost consequence that nobody prices. An emptying study, an endoscopy and an unplanned stop all sit outside the monthly figure, and outside the trial protocols that produced the efficacy numbers. Rare events are still expensive when they land on you.