Explore this article's sources with AI
Follow PeptideMethods on Google
Semaglutide vs Tirzepatide in FAERS
A 2026 FAERS analysis found different reporting patterns for semaglutide and tirzepatide, without establishing comparative risk or causality.
Why we wrote this. Spontaneous-report comparisons are easily mistaken for risk estimates. This readout keeps that boundary visible.
In this article (5 sections)
A 2026 analysis of US spontaneous safety reports found different reporting patterns for semaglutide and tirzepatide. Semaglutide had higher reporting odds for several gastrointestinal and kidney events, while alopecia was relatively more often reported with tirzepatide[1]. These are reporting signals, not comparative rates of harm and not proof that either medicine caused an event.
The study used the FDA Adverse Event Reporting System, known as FAERS. This database collects reports from clinicians and patients as well as manufacturers. FDA has since transitioned to the Adverse Event Monitoring System, but it continues to describe spontaneous reports as one input to post-market surveillance rather than a standalone measure of incidence or causality[2].
How the comparison was built
The researcher examined FAERS submissions from January 2022 through June 2026. After duplicate reports were removed, predefined exposure rules identified 60,423 reports listing semaglutide alone and 143,126 listing tirzepatide alone. Another 853 listed both medicines without another suspect active ingredient[1]. A larger report count does not mean a medicine has a higher event rate because the number of exposed patients is unknown.
The main statistic was the reporting odds ratio, or ROR. It compares how often a particular event appears among reports for one medicine with how often it appears among reports for the comparator. The analysis also used adjusted logistic regression and false-discovery-rate correction. A latency analysis examined the interval between reported treatment start and event onset[1]. That distinction matters.
An ROR above 1 means the event occupied a larger share of one drug's reports. It does not mean the event was that many times more likely among users. FAERS lacks a reliable denominator for total users, and reports can be affected by publicity or length of time on the market. Differences in patient populations may also shape the comparison[2].
Semaglutide had higher odds for selected reports
Compared with tirzepatide, semaglutide had higher reporting odds for impaired gastric emptying, with an ROR of 6.39. The ROR was 4.56 for intestinal obstruction or ileus and 2.51 for acute kidney injury. Pancreatitis had an ROR of 2.18, while vomiting had an ROR of 2.11[1]. These figures compare the composition of submitted reports, not event frequency per treated patient.
The largest quoted signal was optic ischaemic neuropathy, with an ROR of 15.77 for semaglutide versus tirzepatide[1]. A large disproportionality statistic can arise from a small event count, selective reporting, or differential attention. The abstract does not provide the underlying count for this event, so its absolute scale cannot be assessed from the indexed record alone.
Alopecia moved in the other direction. Its ROR was 0.74 when semaglutide was compared with tirzepatide, meaning alopecia made up a relatively larger share of tirzepatide reports[1]. That result still cannot establish how commonly hair loss occurs in either user population or whether weight change and other factors explain some reports.
Hospitalization and death were report outcomes
After adjustment, semaglutide reports had 2.80 times the reporting odds of hospitalization and 1.21 times the reporting odds of death compared with tirzepatide reports[1]. These outcomes describe submitted case reports. They are not randomized comparisons and do not establish excess mortality or hospitalization risk among all users.
Adjustment can account only for variables that are captured and modeled. It cannot recover a missing exposure denominator. It also cannot fully correct stimulated reporting, underreporting, missing clinical details, or confounding by why each medicine was prescribed. Differences in approval timing and use patterns can affect which reports enter the database.
FDA uses spontaneous-report systems to detect possible safety concerns that may need further evaluation. The agency states that a report does not mean a medicine caused the event. Reports can be incomplete, duplicated despite cleaning efforts, or influenced by outside information[2]. Signal detection is therefore a starting point for investigation.
The erectile-dysfunction finding was exploratory
In a male-only exploratory analysis, erectile dysfunction appeared in 92 of 16,673 semaglutide reports and 46 of 27,321 tirzepatide reports. The reported ROR was 3.29, with a 95% confidence interval from 2.31 to 4.69 after false-discovery-rate correction[1].
The paper explicitly calls this result hypothesis-generating[1]. The report denominators are not user denominators, and the analysis cannot determine baseline sexual function or relevant comorbidities for every user. A population-based design with verified exposure and a suitable comparison group would be needed to estimate risk.
What this FAERS study cannot answer
This analysis cannot tell a reader which medicine is safer overall. It cannot calculate the chance that a user will experience a named event. Nor can it show that switching medicines would reduce risk. Those questions need designs that measure total exposed populations and control clinical differences between groups.
The defensible reading is narrower: semaglutide and tirzepatide generated different patterns within submitted reports through June 2026. Those patterns identify questions for active surveillance or population studies, while the spontaneous-report data themselves do not establish comparative clinical risk[1].
Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.
Frequently asked
Did this study prove semaglutide is riskier than tirzepatide?
No. It compared the proportions of events in spontaneous reports. It did not measure event rates among all users and cannot establish causality.
What is a reporting odds ratio?
It measures whether an event occupies a larger share of reports for one drug than for a comparator. It is a signal-detection statistic, not a relative risk.
Which events were reported relatively more with semaglutide?
The analysis found higher reporting odds for selected gastrointestinal events, acute kidney injury, pancreatitis, vomiting, and optic ischaemic neuropathy. Absolute risks were not estimated.
What did the analysis say about erectile dysfunction?
A male-only exploratory analysis found a reporting signal for semaglutide versus tirzepatide. The paper called it hypothesis-generating and said population-based confirmation is required.
Sources
No revisions yet. First published .