GLP-1 receptor agonists in FAERS: adverse-event reporting by the numbers

The FDA Adverse Event Reporting System (FAERS) holds 396,294 individual safety reports that name at least one GLP-1 receptor agonist — semaglutide, tirzepatide, liraglutide, dulaglutide, exenatide, or lixisenatide — in any role, filed between 2004 and early 2026. Tirzepatide alone appears in 137,408 reports despite entering the US market in May 2022, making it the fastest-accumulating active substance in the modern FAERS record. The single most common reaction term across the whole class is not pancreatitis or gastroparesis; it is nausea, at 51,649 reports. And the class's most underappreciated reporting category is the medication-error cluster — incorrect dose, extra dose, accidental underdose, and dose omission — which together appears in 91,806 reports, 23.2% of the class total.

This article is a class-level descriptive read of GLP-1 reporting in FAERS, written for pharmacovigilance, medical-affairs, market-access, and specialty-pharmacy teams who encounter FAERS numbers in safety briefings, payer dossiers, and litigation monitoring. Every figure here is computed from the public openFDA FAERS extract (export dated 2026-06-08, 20.3 million total reports). It is not a disproportionality signal analysis, it is not a safety claim, and it is not incidence data. FAERS is a spontaneous-reporting system: it captures what was reported, never what was caused, and it has no exposure denominator. Those caveats are not a footnote — they are the analytic frame, and the reporting-mix distortions they produce are explained in detail below. For the database-wide aggregate picture, see the companion FAERS analysis of 20 million reports; this article stays on the GLP-1 class.

Methodology, in one paragraph

A report is counted for a substance when the substance's generic name appears anywhere in that report's drug list — as a suspect, concomitant, or interacting drug. Because a single report can name several drugs, the per-substance totals sum to more than the 396,294 class-level reports, and a report naming both semaglutide and metformin counts once for the class but once each for those two substances. A "serious" report is any report the FDA coded as serious; a report is counted toward death when either the seriousness field carries a death flag or the outcome field carries a Fatal outcome. Reactions are coded to MedDRA Preferred Terms; the openFDA extract carries mixed casing for some terms, so reaction counts in the tables below are case-normalized to avoid splitting the same concept across "Nausea" and "NAUSEA." The "any role" counts reported here are intentionally inclusive — they count a report whenever the generic name appears in the drug list — so they will exceed suspect-only or differently normalized counts for the same agent cited elsewhere; the companion database-wide analysis ranks substances on a different counting basis, and its single-substance figures are not directly comparable to the any-role totals here.

The reporting-volume picture, by drug

Two columns deserve immediate emphasis, because both are commonly misread.

Volume does not rank risk. Tirzepatide leads on raw count because it is the newest, most rapidly prescribed GLP-1 RA in a market that grew explosively after the 2021 SELECT cardiovascular result and the 2022–2024 obesity approvals. Reporting volume tracks prescribing volume and public attention far more than it tracks inherent hazard. A drug with ten times the prescriptions and ten times the media coverage will accumulate roughly ten times the reports even at identical true risk.

Serious % does not rank risk either. The serious-fraction differences across these six substances look dramatic — tirzepatide 16.1% versus liraglutide 47.4% — and they are real as descriptions of the reporting mix, but they are not comparisons of inherent danger. Liraglutide and exenatide have been on the market since 2010 and 2005, respectively; their reports skew older, skew toward a type 2 diabetes population with more comorbidity, and include a long tail of historical pancreatitis and thyroid litigation filings. Tirzepatide's report flood is dominated by consumers describing titration errors and transient gastrointestinal symptoms during the first weeks of therapy, which mechanically depresses the serious fraction. Read the serious % column as a description of who is reporting what, not as a head-to-head safety verdict. The only defensible way to compare hazard across these drugs is a disproportionality analysis with a defined comparator and a 2×2 contingency statistic, not a raw-percentage eyeball test.

The growth curve

The class went from 8,880 reports in 2015 to 92,324 in 2025 — a 10.4-fold increase in a decade — and the composition inverted. In 2015 the class was liraglutide and dulaglutide; by 2025 tirzepatide alone accounted for 66% of class reports. The 2026 figure is partial (early-year filings) and should not be annualized. Three forces drive the curve: prescribing volume (which multiplied after the obesity and cardiovascular-risk-reduction indications), consumer awareness (which channels reports through the voluntary MedWatch path rather than manufacturer submission), and litigation activity (which deposits large batches of attorney-filed reports in specific quarters).

The reaction profile

The fifteen most frequent MedDRA Preferred Terms across all GLP-1 reports, case-normalized:

The headline observation is that the GLP-1 reaction profile is dominated by known, on-label pharmacology — nausea, vomiting, diarrhoea, decreased appetite, weight decreased — plus a large operational layer of dosing-error terms and injection-site terms. "Off label use" and "Drug ineffective" together (33,000 reports) reflect the access-driven use of GLP-1 RAs for cosmetic and off-indication weight loss during the shortage era, not a safety signal.

The medication-error layer

Grouping the dosing-error terms — incorrect dose, extra dose, accidental underdose, dose omission, wrong technique, inappropriate schedule, and overdose — yields 91,806 reports, or 23.2% of the class. This is the single largest thematic category in GLP-1 FAERS reporting, larger than the entire gastrointestinal cluster excluding nausea. Tirzepatide carries the heaviest medication-error load: "Incorrect dose administered" is its number-one reaction term at 28,068 reports, followed by "Extra dose administered" at 8,990 and "Accidental underdose" at 4,606. The pattern tracks the monthly titration schedule (2.5/5/7.5/10/12.5/15 mg), the pen-injection learning curve, and the 2022–2025 period in which compounded and non-FDA-approved versions entered the supply chain. FDA's own compounded-product adverse-event tally — 605 reports for compounded semaglutide and 545 for compounded tirzepatide as of July 31, 2025 — sits inside this larger error picture.

The gastrointestinal and renal signals that reached the label

Several of these clusters map directly onto recent FDA labeling action. In October 2025 the agency strengthened the Ozempic and Wegovy labels to add or amend ileus, intestinal obstruction, severe constipation including fecal impaction, and acute kidney injury, and to state that Ozempic is not recommended in patients with severe gastroparesis — the same gastroparesis and ileus terms that appear here as 4,430 and 3,013 reports. In November 2024 FDA added a class-wide warning about pulmonary aspiration of residual gastric contents during general anesthesia or deep sedation. And on January 13, 2026, FDA requested removal of the suicidal ideation and behavior warning from the Saxenda, Wegovy, and Zepbound labels after a comprehensive review found no increased risk; the 1,698 suicidal-ideation reports in this extract are part of the evidentiary record FDA weighed, and their existence in FAERS is not, on its own, evidence of causation — which is precisely the conclusion FDA reached.

The pancreatitis cluster (11,514 reports) is the class's largest serious-event category by frequency and is concentrated in the older agents — liraglutide carries 2,296 pancreatitis reports and exenatide historically carries the highest pancreatitis reporting fraction in published GLP-1 pharmacovigilance studies. Published disproportionality analyses have reached mixed conclusions: a 102,257-participant network meta-analysis found a neutral pooled relative ratio of 0.96 (95% CI 0.31–3.00) for GLP-1 RAs and acute pancreatitis, while Cleveland Clinic data place tirzepatide pancreatitis rates at roughly 0.3–0.4% across doses. FAERS counts cannot adjudicate that question; they identify it for surveillance.

Who is reporting, and from where

The 79.5% consumer share is the single most important interpretive fact in this dataset. By reporter type, the GLP-1 class is 79.5% consumer-reported — far above the roughly 45.6% consumer share across the FAERS database as a whole. That reporter-type share is distinct from the submission channel: across the database, most reports reach FDA through manufacturer submission rather than through a direct MedWatch filing, because manufacturers are legally required to forward reports they become aware of, so a consumer-originated report is still typically transmitted by the manufacturer. What is unusual about the GLP-1 class is not the channel but the reporter mix — direct-to-consumer advertising, social-media awareness, patient-advocacy reporting tools, and active litigation solicitation have made consumers the originators of GLP-1 reports at a rate no other drug class matches. Consumer-originated reports are systematically less complete than manufacturer-filed ICSRs — they more often lack dose, lot, concomitant medications, and medical history — and they skew toward events consumers can self-identify (nausea, injection-site pain, dosing confusion) and away from events that require clinical workup (subclinical pancreatitis, renal function decline).

Geographically, 73.0% of GLP-1 reports originate in the United States, well above the database-wide US concentration of roughly 57%. Among reports carrying a recorded sex, 64.2% list female (233,337 of 363,923 sex-coded reports), consistent with the predominantly female, obesity-indication prescribing pattern in the post-2022 period. The indication field names a diabetes-related term in 38.1% of reports and an obesity or weight term in 13.0%; the remainder carry no indication, a non-metabolic indication, or an indication recorded for a concomitant drug.

What this means for pharmacovigilance and access teams

Treat FAERS GLP-1 counts as a monitoring layer, not an evidence layer. The numbers above tell you where FDA's postmarket surveillance attention is concentrated and which reaction clusters are most likely to drive the next labeling supplement, advisory committee, or risk communication. They do not tell you that tirzepatide is safer than liraglutide, or that semaglutide causes gastroparesis at any particular rate. Any internal slide that ranks these drugs by raw report count or by serious % without the denominator caveat is misleading.

For signal detection, use disproportionality and cite the peer-reviewed work. A reporting odds ratio or empirical Bayesian geometric mean against a defined comparator is the minimum defensible signal statistic, and even then a signal is a hypothesis for further study, not a finding of causation. Published GLP-1 FAERS disproportionality work — including tirzepatide-versus-semaglutide acute kidney injury analyses (one 2022–2025 study reported an AKI reporting rate of 0.47% for tirzepatide versus 1.07% for semaglutide, ROR 0.44, 95% CI 0.38–0.50) and class-wide gastrointestinal disproportionality studies — is the right place to start.

Expect the medication-error layer to drive payer and FDA action separately from the pharmacology. The 23.2% dosing-error share is a reminder that GLP-1 risk management is partly a product-design and channel problem (titration schedules, pen ergonomics, cold-chain, compounded-product provenance) rather than purely an adverse-reaction problem. Access teams building REMS, hub, or specialty-pharmacy programs should treat the error cluster as the actionable operational signal.

Use the label, not the report count, as the coverage anchor. When a payer policy cites a GLP-1 safety concern, the defensible source is the current FDA labeling — the October 2025 gastroparesis, ileus, and acute kidney injury language; the November 2024 pulmonary aspiration warning; and the January 2026 removal of the suicidal-ideation language — not a FAERS tally. FAERS tells you what is being watched; the label tells you what FDA has concluded. For the mechanics of reading an individual FAERS report and its MedDRA coding, seriousness criteria, and reporter fields, see the companion FAERS report reading workflow.

Sources

• US FDA, FDA Adverse Event Reporting System (FAERS) / openFDA drug adverse event data, public extract (20,328,575 reports, export dated 2026-06-08). Aggregates computed by PharmaDossier from the openFDA FAERS full extract. https://open.fda.gov/data/faers/
• US FDA, FAERS Public Dashboard and FDA Adverse Event Monitoring System. https://www.fda.gov/drugs/surveillance-post-drug-approval-activities/fda-adverse-event-monitoring-system-aems
• US FDA, "FDA Requests Removal of Suicidal Behavior and Ideation Warning from GLP-1 RA Medications," January 13, 2026. https://www.fda.gov/drugs/drug-safety-communications/fda-requests-removal-suicidal-behavior-and-ideation-warning-glucagon-peptide-1-receptor-agonist-glp
• US FDA, "FDA's Concerns with Unapproved GLP-1 Drugs Used for Weight Loss" (compounded-product adverse-event counts as of July 31, 2025). https://www.fda.gov/drugs/drug-alerts-and-statements/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss
• US FDA, Medscape / FDA safety-related labeling change: pulmonary aspiration warning for GLP-1 receptor agonists, November 5, 2024. https://www.medscape.com/viewarticle/fda-updates-glp-1-label-pulmonary-aspiration-warning-2024a1000k84
• Defective Drug Site, "GLP-1 Regulatory Updates (2026): FDA Label Changes, EMA Signals & Safety Warnings," October 2025 Ozempic/Wegovy gastroparesis, ileus, acute kidney injury label changes. https://defectivedrugsite.com/glp-1-drugs/lawsuits/regulatory-updates
• Liu L et al., "Adverse events administering GLP-1 receptor agonists: a cross-sectional study" (FAERS, 2015–2024), 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC12927500
• "Gastrointestinal Safety Assessment of GLP-1 Receptor Agonists in the US: A Real-World Adverse Events Analysis from the FAERS Database." https://pmc.ncbi.nlm.nih.gov/articles/PMC11675942
• "Comparative Renal Safety of Tirzepatide and Semaglutide: An FDA Adverse Event Reporting System (FAERS) Disproportionality Study," J Clin Med, 2025. https://www.mdpi.com/2077-0383/14/21/7678
• Cleveland Clinic, "Reconciling GLP-1s and Pancreatitis" (tirzepatide pancreatitis rates; network meta-analysis RR 0.96). https://consultqd.clevelandclinic.org/glp1-agonists-and-pancreatitis