Gastroparesis — delayed gastric emptying that can cause nausea, vomiting, early satiety, bloating, and in severe cases bezoar formation or malnutrition — is one of the most-discussed adverse-event signals associated with semaglutide, tirzepatide, and the broader GLP-1 class. The mechanistic link is straightforward: delayed gastric emptying is part of how GLP-1 receptor agonists produce satiety and reduce post-prandial glucose excursions. The clinical question is what portion of that pharmacology tips into pathology, in whom, at what dose, and for how long.
Three data sources inform the 2026 answer: FDA FAERS (passive adverse-event reports), a small but growing peer-reviewed case-series and observational literature, and FDA label history. None of them alone is sufficient; together they describe a signal whose magnitude is real but whose population incidence is lower than raw FAERS counts suggest. This piece walks through each.
What FAERS can and cannot tell you
The FDA Adverse Event Monitoring System (AEMS, formerly FAERS) is a passive spontaneous-reporting system. Clinicians, patients, and manufacturers submit reports of suspected drug-related adverse events. The system is valuable for early signal detection — the 2023 label updates for semaglutide were driven in part by FAERS pattern recognition — but it is structurally not an incidence-measurement tool.
FAERS has no denominator. A raw count of gastroparesis reports associated with semaglutide does not, on its own, tell you the rate per million patient-years of exposure, because total prescription volume is not part of the FAERS dataset. It also suffers from ascertainment bias: high-profile adverse events and litigation activity both inflate reporting. Reports tied to GLP-1 drugs have risen sharply since 2022, paralleling the explosion in prescribing volume. How much of the report growth reflects true incidence growth, and how much reflects reporting-behavior growth, is genuinely unsettled.
That is why serious pharmacovigilance analysis pairs FAERS with a second data source — typically insurance-claims or EHR-based cohort data — to attach a denominator.
The peer-reviewed evidence: Sodhi et al. and what followed
Sodhi and colleagues published the most widely cited cohort analysis in JAMA in October 2023 (JAMA 2023;330(18):1795-1797; PMID 37796527). Using a large insurance-claims dataset and restricting to patients using GLP-1 drugs for weight loss (not diabetes), the authors compared GLP-1 users against bupropion-naltrexone users — a contemporary non-GLP-1 weight-loss comparator.
Key findings: semaglutide and liraglutide users had significantly elevated risks of biliary disease, pancreatitis, bowel obstruction, and gastroparesis compared with bupropion-naltrexone users. The gastroparesis hazard ratio was ~3.67 (95% CI 1.15–11.90). Absolute incidence was low in both groups — this is not a common adverse event — but the relative risk was material.
Follow-on cohort work has nuanced the signal. A 2024 multi-center comparison of tirzepatide vs semaglutide found the tirzepatide cohort had ~64% lower risk of de novo gastroparesis in non-diabetic obesity patients, which is consistent with tirzepatide's dual-agonist pharmacology modulating gastric emptying differently than semaglutide. None of this overturns the Sodhi finding; it refines it.
FDA label history
The semaglutide (Ozempic) label has been updated multiple times since 2023 to reflect post-approval adverse-event reporting:
- September 2023: ileus (intestinal paralysis) added to the warnings section. - November 2024: pulmonary aspiration during general anesthesia or deep sedation added, tied to delayed gastric emptying increasing aspiration risk under sedation. - January 2025: severe pancreatitis and acute kidney injury warnings updated.
The DailyMed prescribing information for semaglutide explicitly states the product is not recommended in patients with severe gastroparesis. That language — "not recommended in severe gastroparesis" rather than "contraindicated" — is load-bearing: it reflects clinical judgment that pre-existing gastroparesis is a relative contraindication, not an absolute one.
Practical framing
For patients and clinicians in 2026, the honest read is this. Gastroparesis is a recognized adverse-event signal of GLP-1 receptor agonists, documented in peer-reviewed cohort studies, acknowledged in FDA label language, and consistent with the known pharmacology of the class. Absolute incidence in treatment-naïve obesity populations is low — measured in events per thousand person-years, not percent — but relative risk compared to non-GLP-1 weight-loss interventions is elevated.
The populations at highest risk based on current evidence are patients with pre-existing gastroparesis (explicitly flagged on the label), patients with symptomatic diabetic gastroparesis, and patients experiencing severe or prolonged dose-escalation GI symptoms who continue escalating regardless. The mitigation is conservative titration, attention to anesthesia planning (the aspiration-risk question is the newest label concern), and willingness to down-titrate or discontinue rather than push through intractable symptoms.
PeptaHub's GLP-1 pillar covers the class pharmacology; our peptide-safety reference walks through the general framework for weighing trial-population risk against individual context; and the semaglutide vs tirzepatide comparison covers the head-to-head adverse-event differences.