"Stomach paralysis" is the everyday name for gastroparesis, a condition in which the stomach empties too slowly without anything physically blocking it. The Ozempic label does not list it as a side effect. What the label does say is that semaglutide slows early stomach emptying, that Ozempic "is not recommended in patients with severe gastroparesis", and that ileus and intestinal obstruction have been reported since the drug went on sale. Between those three statements sits most of the argument about whether Ozempic causes the condition.
Every label figure below was read on 18 November 2026 from DailyMed: Ozempic injection (version 20, effective 1 June 2026), Wegovy (version 19, effective 18 June 2026) and Zepbound (version 40, effective 28 August 2026). Study figures come from the published papers, cited by PubMed number. Report counts come from FDA's adverse event database through openFDA, updated 28 October 2026.
What the label says, and what it does not
The label has four relevant passages, in four different places.
- Mechanism (section 12). Semaglutide's glucose-lowering effect "also involves a minor delay in gastric emptying in the early postprandial phase" — the period just after a meal.
- Warning (section 5.7). Severe gastrointestinal reactions were reported in 0.4% of patients on 0.5 mg and 0.8% on 1 mg, against 0% on placebo, and Ozempic "is not recommended in patients with severe gastroparesis". The Wegovy and Zepbound labels carry the same sentence.
- Postmarketing (section 6.2). Ileus, intestinal obstruction, and severe constipation including faecal impaction are listed among reactions reported after approval. The label adds its standard caveat that such reports come from a population of uncertain size and cannot reliably establish frequency or cause.
- Patient information. Before starting, people are asked to tell their prescriber about "severe problems with your stomach, such as slowed emptying of your stomach (gastroparesis)".
So the label treats gastroparesis as a reason for caution in people who already have it, not as a recorded effect of the drug. Gastroparesis does not appear in the adverse reaction tables or the postmarketing list.
There is also a sentence in the same label that is rarely quoted. Directly after its interaction studies, it reports that "in a separate study, no apparent effect on the rate of gastric emptying was observed with semaglutide 2.4 mg". Read beside the mechanism section, the two lines fit one pattern: the slowing is concentrated early after a meal and early in treatment, and fades with continued dosing. The Zepbound label states that pattern outright for tirzepatide — a marker of stomach emptying peaked 55% lower after the first 5 mg dose and showed "no meaningful impact" by week 6. That fading is why the risk is usually described as clustering around starting and dose increases, and also why a measurement taken at steady state can look normal.
What the studies counted
| Study | Population | What was counted | Result |
|---|---|---|---|
| Sodhi 2023, JAMA, PMID 37796527 | US claims (PharMetrics Plus), 2006–2020; obesity code, no diabetes | Gastroparesis = diagnosis code or a promotility prescription | Semaglutide 4 events / 613 users (9.1 per 1,000 person-years); liraglutide 66 / 4,144 (7.3); bupropion-naltrexone 3 / 654 (3.1). Adjusted HR 3.67 (1.15–11.90) |
| Alkabbani 2026, Diabetes Obes Metab, PMID 42244136 | 313,342 matched pairs, type 2 diabetes, US commercial claims; median 5.2 months on treatment | Composite: severe constipation, gastroparesis, obstruction | 1.02 vs 0.75 per 100 person-years against SGLT-2 inhibitors; HR 1.37 (1.30–1.45), raised for each component |
| Crisafulli 2026, Ann Intern Med, PMID 41183330 | Type 2 diabetes, 2019–2024; 65,238 / 20,893 / 46,620 matched pairs | Composite incl. pancreatitis, biliary disease, obstruction, gastroparesis, severe constipation | Semaglutide vs dulaglutide HR 0.96 (0.87–1.06); tirzepatide vs semaglutide 1.07 (0.90–1.26) |
| Olubodun 2026, PLoS One, PMID 42594084 | Systematic review of published cases to October 2025 | Symptomatic gastroparesis reported with a GLP-1 drug | 12 case reports, 13 patients; semaglutide most often named; resolution after stopping in all |
| Vlaeminck 2026, Anaesthesia, PMID 41631344 | 44 semaglutide users before elective surgery, 44 matched controls | Stomach contents on ultrasound after fasting | "Full stomach" 21/43 (49%) vs 8/44 (18%); solid contents 18/43 vs 3/44 |
Three things in that table are easy to miss.
The headline study rests on seven events across two of its arms. The 3.67 hazard ratio compares GLP-1 users as a group with bupropion-naltrexone users, but the semaglutide arm had 4 gastroparesis events and the comparison arm 3. The wide confidence interval — from 1.15 to 11.90 — is what seven events look like. The study also counted a prescription for a promotility drug as gastroparesis, which is a broader net than a confirmed diagnosis.
The large study found a real but small difference. With 313,342 matched pairs, the excess on GLP-1 based drugs was 0.27 events per 100 person-years — roughly one extra serious motility event for every 370 people treated for a year, on the study's composite measure. That figure is our arithmetic from the published rate difference, not a number the authors state.
Between molecules, the best comparison found nothing. In the Annals cohort, semaglutide, dulaglutide and tirzepatide had statistically indistinguishable rates on a composite that included gastroparesis.
The surgical ultrasound study measures something different — food left in the stomach after the usual pre-operative fast — and it matters for anaesthesia rather than for chronic symptoms. Its patients had already had one semaglutide dose withheld. What the labels say about surgery is set out on GLP-1s and surgery.
What FDA's report database shows, and why it misleads
FDA's adverse event database contains no reaction term spelled "gastroparesis": a search for it returns no reports for any of these drugs. Reports describing the condition sit under the broader coded term "impaired gastric emptying".
| Molecule | Reports naming it | "Impaired gastric emptying" | Share |
|---|---|---|---|
| Semaglutide | 78,023 | 4,175 | 5.4% |
| Dulaglutide | 89,369 | 2,003 | 2.2% |
| Liraglutide | 51,303 | 655 | 1.3% |
| Tirzepatide | 181,068 | 1,873 | 1.0% |
Semaglutide's share is five times tirzepatide's. The Annals cohort, which measured actual patients rather than reports, found the two drugs' rates indistinguishable. The likeliest explanation for the gap is reporting, not risk: a report is a voluntary submission, its volume rises with media attention and litigation, and semaglutide was the drug at the centre of the 2023 coverage and of most legal claims. The table is useful as a record of what has been reported, and of how reports are coded. It is not a measure of how often the condition occurs.
The court record
Gastroparesis is the main injury alleged in the federal gastrointestinal docket. On 15 August 2025 the court ruled that a plaintiff claiming drug-induced gastroparesis must have had a gastric emptying study at the time of diagnosis, excluding expert evidence that symptoms and their timing were enough. The ruling decided what counts as proof of the diagnosis; it did not decide whether the drugs cause it. The full record is on the GLP-1 lawsuit page.
Does it go away?
The only evidence is the published cases. In the 13 patients gathered by the 2026 systematic review, symptoms began anywhere from hours to several months after treatment started, often after a dose increase or an inappropriate restart, and resolved after the drug was stopped in every case. That is encouraging and it is weak: unresolved cases are less likely to be written up, and no prospective study has followed stomach emptying after stopping. The label is silent on recovery.
What is and is not established
Established: semaglutide slows early stomach emptying; the label advises against its use in severe gastroparesis; large studies find more serious gut-motility events on GLP-1 drugs than on comparison drugs, at absolute rates around 1 per 100 person-years; and the best head-to-head comparison finds no difference between semaglutide, dulaglutide and tirzepatide.
Not established: how often semaglutide produces confirmed gastroparesis in someone who did not have it, whether it ever persists after stopping, and whether the risk differs by dose. Much of what is published counts a broader set of events, or a looser definition, than a confirmed gastric emptying study.
The nausea, vomiting and constipation behind most of these reports are covered on GLP-1 nausea and GLP-1 constipation, and the rest of Ozempic's label record on Ozempic side effects. Our broader read of the safety record is on is GLP-1 safe.
This page reports what the label, studies and records say and is not medical advice. Vomiting that does not stop, inability to keep food or fluids down, or severe or persistent abdominal pain on Ozempic are reasons to contact a prescriber promptly, as our editorial standards set out.
