Ozempic Adverse Events: What the FDA FAERS Data Shows
Analysis of semaglutide (Ozempic, Wegovy) adverse event reports from the FDA FAERS database. Real numbers, not headlines.

Semaglutide reports are climbing fast
Semaglutide, marketed as Ozempic (diabetes) and Wegovy (weight loss), has become one of the most-reported drugs in the FDA Adverse Event Reporting System (FAERS). The growth tracks with prescribing volume: more patients taking it means more reports filed.
That does not automatically mean the drug is more dangerous than alternatives. FAERS is a voluntary reporting system. More media attention on a drug leads to higher reporting rates, regardless of actual risk.
How FAERS data works
The FDA Adverse Event Reporting System collects voluntary reports from healthcare professionals, consumers, and manufacturers. Manufacturers are required to report; everyone else is voluntary.
Key limitations to understand before analyzing this data:
- Reports do not prove causation
- Duplicate reports exist (same event reported by patient and doctor)
- No denominator (you cannot calculate incidence rates from FAERS alone)
- Reporting rates vary by drug popularity and media coverage
Top reported reactions for semaglutide
Based on openFDA FAERS queries through Q2 2026, the most frequently reported adverse reactions for semaglutide include:
| Reaction | Approximate Report Count | % of Total |
|---|---|---|
| Nausea | ~12,400 | 18% |
| Vomiting | ~7,200 | 10% |
| Diarrhea | ~5,100 | 7% |
| Abdominal pain | ~4,300 | 6% |
| Headache | ~3,800 | 5% |
| Pancreatitis | ~2,100 | 3% |
| Gastroparesis | ~1,400 | 2% |
GI side effects dominate. This aligns with the drug's mechanism of action (GLP-1 receptor agonist that slows gastric emptying).
Querying semaglutide data from openFDA
The openFDA drug adverse events endpoint accepts queries by generic name:
curl "https://api.fda.gov/drug/event.json?search=patient.drug.openfda.generic_name:\"semaglutide\"&count=patient.reaction.reactionmeddrapt.exact&limit=20"
This returns a counted list of all reported reactions, sorted by frequency. The count parameter is what makes this useful for analysis rather than just browsing individual reports.
Serious vs. non-serious reports
FAERS classifies reports as "serious" if they involve death, hospitalization, disability, life-threatening situations, or congenital anomalies.
For semaglutide, roughly 35-40% of reports are classified as serious. That sounds high, but it is comparable to other widely-prescribed injectable medications. The "serious" flag does not mean the drug caused the outcome.
The pancreatitis signal
Pancreatitis has been a focus of safety discussions for the entire GLP-1 receptor agonist class since exenatide (Byetta) first raised concerns in 2007. The FDA added pancreatitis warnings to semaglutide labeling.
In FAERS data, pancreatitis reports for semaglutide number around 2,100 through mid-2026. Whether this represents a genuine elevated risk above background rates requires controlled epidemiological studies, not FAERS signal detection alone.
Comparing semaglutide to other GLP-1 drugs
Other GLP-1 receptor agonists in the FAERS database for context:
| Drug | Total Reports (all time) | Serious % |
|---|---|---|
| Semaglutide (Ozempic/Wegovy) | ~69,000 | ~37% |
| Liraglutide (Victoza/Saxenda) | ~45,000 | ~42% |
| Dulaglutide (Trulicity) | ~31,000 | ~39% |
| Tirzepatide (Mounjaro/Zepbound) | ~28,000 | ~34% |
Semaglutide's higher absolute count reflects its larger patient population, not necessarily a worse safety profile.
How to interpret this data responsibly
Three rules for working with FAERS data:
-
Never calculate incidence rates from FAERS alone. Without knowing total patients exposed, a report count is meaningless for risk estimation.
-
Compare within drug classes, not across them. A diabetes drug and a chemotherapy agent have completely different expected adverse event profiles.
-
Look at proportional reporting ratios (PRR). The ratio of a specific reaction for one drug versus all other drugs in the database is more informative than raw counts.
FAQ
Is Ozempic dangerous based on FAERS data? FAERS data cannot answer that question. It identifies safety signals for further investigation. The prescribing label already reflects known risks based on clinical trials.
Why are there so many reports for semaglutide? High prescribing volume plus intense media coverage equals high reporting rates. This is expected for blockbuster drugs.
How often is FAERS updated? The FDA updates the openFDA FAERS endpoint quarterly. The raw FAERS data files are also available for download quarterly.
Can I build alerts on new semaglutide reports? The quarterly update cycle means real-time alerting is not possible with the public API. New data appears in batches.
Published on 2026-08-12 · 4 min read
← Back to all articles