The most frequently reported problem with tirzepatide is not nausea. It is people administering the wrong dose — 28,115 reports, more than double the next entry. Add the related categories and roughly 41,000 of tirzepatide's reports to the FDA describe a dosing mistake rather than a reaction to the drug.
That does not appear on any product label, because labels list what a drug does to you. This page is about something else: what people actually filed with the FDA. The two lists are not the same, and the difference is where the useful information is.
Read this before the numbers
These figures come from the FDA Adverse Event Reporting System, FAERS, queried through the openFDA API on 3 August 2026. It is a genuinely valuable dataset and it is also the single most misread one in medicine. Four things have to be understood or every number below will mislead you.
- There is no denominator. FAERS records reports, not rates. It does not know how many people took the drug, so a count can never tell you how likely anything is.
- A report is not a finding. Anyone can file one — patients, doctors, lawyers, manufacturers — and filing one does not establish that the drug caused the event. Reports are a signal to investigate, not a conclusion.
- Attention inflates counts. A drug in the news, in litigation, or newly launched attracts more reports than an older one used by the same number of people. Reporting behaviour is part of what these numbers measure.
- Bigger is not worse. The clearest proof is below, and it is worth sitting with before reading any comparison table anywhere on this subject.
Why the biggest number is the least alarming
Set the totals against the serious outcomes:
| Molecule | Total reports | Deaths reported | Hospitalisations |
|---|---|---|---|
| Tirzepatide | 140,435 | 775 | 8,670 |
| Dulaglutide | 86,770 | 1,632 | 10,332 |
| Semaglutide | 63,441 | 1,183 | 13,099 |
| Liraglutide | 50,219 | 1,841 | 11,291 |
Tirzepatide has the most reports of any GLP-1 and the fewest deaths. Liraglutide has the fewest reports and the most. If report volume measured danger, that table would be impossible.
What it actually reflects: liraglutide has been on the market since 2010 and is used heavily in older patients with established type 2 diabetes, a population in which deaths occur for many reasons. Tirzepatide is recent, heavily prescribed for weight management in a generally healthier population, and — as the next section shows — its report pile is dominated by something that is not a medical event at all.
This is why a headline counting FAERS reports tells you almost nothing, and why we are showing the serious-outcome columns alongside rather than the totals alone.
The dosing-error finding
Sort each molecule's reports by whether they describe a dosing mistake rather than a bodily reaction:
| Molecule | Total reports | Dosing-error reports | Share |
|---|---|---|---|
| Tirzepatide | 140,435 | 41,277 | 29% |
| Dulaglutide | 86,770 | 14,758 | 17% |
| Semaglutide | 63,441 | 4,703 | 7% |
| Liraglutide | 50,219 | 0 | 0% |
Nearly a third of everything reported about tirzepatide is someone taking it wrong. For liraglutide the figure is zero.
Broken down, tirzepatide's dosing reports are:
- 28,115 — incorrect dose administered
- 9,000 — extra dose administered
- 4,607 — accidental underdose
- 4,162 — dose omission
- 2,023 — accidental overdose
- 1,487 — inappropriate schedule of administration
Why tirzepatide and not liraglutide
We cannot prove the mechanism from FAERS, so this is a reading rather than a finding — but the pattern lines up with how each drug is packaged.
Liraglutide is a daily pen with a dial. You turn it to the number and inject. It has been in use since 2010, the routine is daily so it becomes habit quickly, and there is no measuring step. Zero dosing-error reports.
Tirzepatide is supplied in more than one format. There is a single-dose autoinjector, and there is the multi-dose KwikPen, and there are vials — which, with compounded tirzepatide, mean drawing a dose yourself into a syringe. Each format is a chance to select or measure incorrectly, and a weekly schedule is easier to lose track of than a daily one. Dulaglutide sits in between at 17%.
The practical implication is unglamorous and worth acting on: if you are on tirzepatide, the device you are handed matters as much as the dose written on the prescription. Confirm which format you have, confirm the dose that format delivers, and confirm the day. The pricing difference between the KwikPen and the single-dose Pen is covered in our GLP-1 cost index; the clinical point here is that they are not interchangeable in handling.
What the reactions themselves show
Filtering out the administration reports, the most-reported bodily effects for semaglutide are the ones the label predicts, in roughly the order patients describe:
- Nausea — 9,513
- Vomiting — 6,161
- Diarrhoea — 5,718
- Decreased appetite — 4,352
- Constipation — 4,141
- Headache — 2,816
- Delayed gastric emptying — 2,796
- Fatigue — 2,773
Two observations worth making. The gastrointestinal cluster dominates completely, which matches both the trials and what people describe in practice — the mechanics of why are in our guide to the GLP-1 side effects timeline. And delayed gastric emptying appears explicitly, at nearly 2,800 reports. That is the mechanism behind the nausea rather than a separate problem, and it is the reason anaesthetists now ask about GLP-1 use before sedation.
The signal nobody labels
One entry in semaglutide's top ten is not a symptom at all: off-label use, 6,280 reports, with a further 2,906 filed as product use in an unapproved indication. Tirzepatide carries 8,291 off-label reports.
That is roughly 9,000 reports for semaglutide describing the drug being used outside its approved indication — most obviously Ozempic, which is approved for type 2 diabetes, being taken for weight loss.
It is a reporting category rather than a harm, and it does not mean those patients came to any. What it does is put a number on something usually discussed anecdotally, and it explains why the brand written on the prescription matters so much for coverage. That distinction is set out in the FDA-approved GLP-1 medications list.
How to use any of this
- Do not compare drugs on report counts. The table above shows why. Without exposure data the comparison is meaningless, and the direction it points is often backwards.
- Do treat the dosing-error signal as actionable. It is not about the molecule; it is about handling, and handling is something you control.
- Do not read a reported event as a caused event. FAERS is a place to look for signals, not a record of what a drug did.
- Do mention GLP-1 use before any procedure with sedation. The delayed gastric emptying reports are the reason this became standard practice.
- Report your own experience through the FDA's MedWatch programme if something happens. These datasets only work because people file.
Counts were queried from the FDA Adverse Event Reporting System through the openFDA API on 3 August 2026 and will change as reports accumulate. FAERS reports are voluntary and unverified, contain duplicates, and do not establish that a drug caused the event described; the database has no information about how many people took each medication, so no rate can be calculated from it. This is general information, not medical advice. Do not change a dose or stop a medication based on anything here — talk to your prescriber.
Scientific References
3 sources- 1
U.S. Food and Drug Administration
FDA Adverse Event Reporting System (FAERS)
FDA · 2026
- 2
openFDA
Drug Adverse Event API
U.S. Food and Drug Administration · 2026
- 3
U.S. Food and Drug Administration
MedWatch: The FDA Safety Information and Adverse Event Reporting Program
FDA · 2026
References open in a new tab. Content is reviewed against peer-reviewed literature as part of our editorial policy.
About the author
Editor, Modern Weight Science
Claudiu Gheorghe is the editor of Modern Weight Science. He is not a physician. His role is to synthesize peer-reviewed studies, clinical-trial data, and FDA prescribing information into clear, plain-language explanations, and to make sure every factual claim on the site traces back to a cited source. Any decision about starting, changing, or stopping a medication belongs with a licensed clinician who knows your history.
Every claim is checked against peer-reviewed research through our review process and fact-checking policy.
Frequently Asked Questions
What is the most reported problem with tirzepatide?
Incorrect dose administered, with 28,115 reports in the FDA's adverse event database — more than double the next entry, which is nausea. Adding related categories, about 41,000 of tirzepatide's reports describe a dosing mistake rather than a reaction to the drug.
Does a higher number of FDA reports mean a drug is more dangerous?
No, and this class demonstrates it clearly. Tirzepatide has the most reports of any GLP-1 (140,435) and the fewest reported deaths (775). Liraglutide has the fewest reports (50,219) and the most deaths (1,841). FAERS has no denominator, so counts cannot be turned into rates, and reporting is influenced by how new a drug is and how much attention it receives.
Why do tirzepatide users make more dosing errors?
It cannot be proven from the data, but the pattern tracks packaging. Liraglutide is a daily dial-a-dose pen with no measuring step and has zero dosing-error reports. Tirzepatide comes as a single-dose autoinjector, a multi-dose KwikPen, and vials, and weekly schedules are easier to lose track of than daily ones. Dulaglutide sits between them at 17%.
What side effects do people actually report for semaglutide?
Nausea (9,513), vomiting (6,161), diarrhoea (5,718), decreased appetite (4,352), constipation (4,141), headache (2,816), delayed gastric emptying (2,796) and fatigue (2,773). The gastrointestinal cluster dominates, which matches both the trial data and patient accounts.
What does off-label use mean in these reports?
It is a reporting category, not a harm. Semaglutide carries 6,280 off-label use reports plus 2,906 for use in an unapproved indication, and tirzepatide 8,291. The most common case is a diabetes-labelled brand being used for weight loss. It does not mean those patients were harmed.
Should I tell my anaesthetist I take a GLP-1?
Yes. Delayed gastric emptying appears explicitly in the reports, at nearly 2,800 for semaglutide, and it is the reason asking about GLP-1 use before sedation became standard practice. The stomach may not be empty when fasting rules assume it is.
Continue learning
Where to read next
Not medical advice. This guide is for general education only. GLP-1 medications, dosing, and treatment suitability are decisions for you and a licensed clinician who knows your full medical history.