Research evidence
Compounded GLP-1 Safety: The Published Evidence
The four verified studies behind every safety claim on this platform, with what each establishes and what it cannot.
Four independent sources document a specific mechanism of harm in compounded GLP-1 supply: the multi-dose vial reintroduces a measurement step that an approved pen removes, producing ten-fold dosing errors. Poison centre contacts rose from roughly 1,000–1,500 a year before mid-2021 to more than 8,000 in 2023. No study compares health outcomes between compounded and approved preparations, so no comparative risk rate is published here.
Key findings
- Poison centre GLP-1 contacts rose from ~1,000–1,500/year pre-2021 to more than 8,000 in 2023 (n=10,033 analysed).
- Semaglutide's share of GLP-1 exposures rose from 24.6% to 64.2% after the 2021 weight-management approval.
- Wrong-dose events were the most frequent scenario by raw count (n=5,171).
- Compounded vials contributed to ten-fold overdoses through unit confusion — documented in a case series and the national dataset.
- No study compares health outcomes between compounded and approved preparations; no comparative risk rate exists.
| Verified studies | 4 |
|---|---|
| Largest dataset | 10,033 GLP-1 exposures, National Poison Data System 2012–2023 |
| Peak annual poison centre calls | More than 8,000 (2023) |
| Documented error magnitude | Ten-fold overdose from multi-dose vials |
| Outcome comparison studies | No Public Evidence Found none published |
| Evidence verified | 2026-07-23 |
What does the evidence actually show about compounded GLP-1 safety?
The honest answer is narrower than either side of this debate usually claims, and it is worth stating precisely because the precision is the finding.
No study has compared health outcomes between patients receiving compounded and approved GLP-1 preparations. That comparison has not been run, so anyone claiming compounded products are categorically more dangerous — or equally safe — is going beyond the evidence.
What the evidence does establish, consistently and across four independent sources, is a specific mechanism of harm: the multi-dose vial reintroduces a measurement step that an approved pen removes, and that step is producing ten-fold overdoses at measurable volume.
The four sources this platform relies on
| Study | Design | Scale | Identifier |
|---|---|---|---|
| GLP-1 receptor agonist exposures reported to US poison cen | Retrospective analysis of the National Poison | 10,033 exposures | Journal of Medical Toxicology 2026;22:275–285 · doi: |
| Administration errors of compounded semaglutide reported t | Case series | 3 patients | Journal of the American Pharmacists Association 2023 |
| Safety analysis of compounded GLP-1 receptor agonists: a p | Retrospective disproportionality analysis of t | 81,078 GLP-1 reports; 707 involving compounded products | Expert Opinion on Drug Safety 2025 · doi:10.1080/147 |
| FDA alert: dosing errors associated with compounded inject | Regulatory safety communication | — | US Food and Drug Administration 2024 |
GLP-1 receptor agonist exposures reported to US poison centers, 2012–2023
Journal of Medical Toxicology 2026;22:275–285 · doi:10.1007/s13181-026-01121-z · Retrospective analysis of the National Poison Data System · 10,033 exposures
Why this study matters here. This is the largest dataset quantifying the dosing-error problem, and it isolates the vial-versus-pen distinction as a mechanism rather than an inference.
What it found
- Poison centre calls involving GLP-1 receptor agonists rose from roughly 1,000–1,500 per year before mid-2021 to more than 8,000 in 2023.
- Semaglutide's share of GLP-1 exposures rose from 24.6% to 64.2% after the June 2021 weight-management approval.
- Unintentional therapeutic errors predominated. Wrong-dose events were the single most frequent scenario by raw count (n=5,171), followed by wrong-time errors (n=1,160).
- Patient demographics shifted younger (mean age 57.0 to 51.6) and more female (68.9% to 78.2%).
- Compounded semaglutide and tirzepatide in vials contributed to 10-fold overdoses through confusion between units, millilitres and milligrams.
- Poison centres noted emerging exposures to retatrutide bought online as a research chemical.
What it cannot establish
- Poison centre contact is the endpoint; the study does not track hospitalisation, death, or longer-term outcomes.
- It cannot establish how many patients made similar errors without contacting a poison centre.
- Exposure reports are voluntary and are not a population denominator.
Administration errors of compounded semaglutide reported to a poison control center: case series
Journal of the American Pharmacists Association 2023 · PMID 37392810 · Case series · 3 patients
Why this study matters here. It documents the mechanism directly: vial plus non-matched syringe plus absent counselling produced ten-fold errors.
What it found
- Three adverse drug events followed incorrect administration of semaglutide obtained from compounding pharmacies and an aesthetic spa.
- Two patients self-administered 10-fold dosing errors.
- All three experienced notable nausea, vomiting and abdominal pain, with most symptoms lasting days.
- One patient reported receiving a vial with syringes for self-administration and no pharmacist counselling.
- The authors state that compounded semaglutide vials lack the safety features of prefilled manufactured pens and permit large overdoses.
What it cannot establish
- Three cases cannot establish incidence or comparative risk.
- Case series are subject to reporting and selection bias.
- Findings describe specific dispensing practices, not all compounded supply.
Safety analysis of compounded GLP-1 receptor agonists: a pharmacovigilance study using FAERS
Expert Opinion on Drug Safety 2025 · doi:10.1080/14740338.2025.2499670 · PMID 40285721 · Retrospective disproportionality analysis of the FDA Adverse Event Reporting System, 2018–2024 · 81,078 GLP-1 reports; 707 involving compounded products
Why this study matters here. It is the only published attempt to compare compounded against non-compounded GLP-1 safety signals in a national database.
What it found
- Of 81,078 GLP-1 receptor agonist reports in FAERS between 2018 and 2024, 707 involved compounded products.
- The analysis examined adverse events, medication errors and product quality issues across liraglutide, semaglutide and tirzepatide.
- Reporting odds ratios were calculated with logistic-regression adjustment.
What it cannot establish
- FAERS is a spontaneous reporting system: reports are voluntary, unverified, and carry no denominator.
- Disproportionality signals indicate reporting patterns, not causation or incidence.
- Compounded products are likely under-identified in reports, so 707 is a floor rather than a count.
FDA alert: dosing errors associated with compounded injectable semaglutide products
US Food and Drug Administration 2024 · Regulatory safety communication
Why this study matters here. It is the agency's own statement of the problem, and it names the two distinct failure points: patient measurement and prescriber calculation.
What it found
- FDA received reports of adverse events from overdoses caused by dosing errors with compounded semaglutide.
- Errors arose both from patients measuring and administering incorrect doses and from health care providers miscalculating doses.
- Reported adverse events included nausea, vomiting, abdominal pain, fainting, headache, migraine, dehydration, acute pancreatitis and gallstones.
- In some reports patients administered five to twenty times the intended dose.
What it cannot establish
- A safety communication reports received signals; it is not an epidemiological study.
- It does not quantify incidence or compare against approved-product presentations.
What the four sources agree on
Read together, they converge on one mechanism and diverge on almost everything else. The convergence is on presentation: a vial requiring the patient to draw a dose, combined with syringes not calibrated for the product and inconsistent units, produces errors of a specific magnitude — ten-fold — that approved pens structurally cannot.
The divergence is on scale. The poison-centre data quantifies contacts, not outcomes. The case series documents mechanism in three patients. FAERS identifies 707 compounded reports without a denominator. None of these supports an incidence rate, and this platform does not publish one.
The practical conclusion is therefore about a design choice rather than about compounding as such. A compounded preparation supplied in a single-dose presentation with matched syringes and clear counselling removes most of the documented failure mode. One supplied as a multi-dose vial with mismatched syringes and no counselling reproduces it.
What this means
- A specific failure mode — ten-fold dosing error from multi-dose vials — is documented across four independent sources.
- Poison centre contacts involving GLP-1 agonists rose more than five-fold after the 2021 weight-management approval.
- FDA has issued its own safety communication naming both patient measurement and prescriber calculation as failure points.
What this does not mean
- That compounded preparations cause worse health outcomes than approved products — no study has compared outcomes.
- That an incidence or comparative risk rate can be derived from these sources; none carries a denominator.
- That all compounded supply shares the same risk profile, which depends on presentation and counselling.
- No randomised or cohort study compares outcomes between compounded and approved GLP-1 preparations.
- Poison centre and FAERS data are voluntary reporting systems without population denominators.
- The case series describes three patients and cannot establish incidence.
- Reported figures describe the period each study covers and may not reflect current practice.
Study records — 4 sources
- Journal of Medical Toxicology 2026;22:275-285checked 2026-07-23
- Poison control case series (PubMed)checked 2026-07-23
- FAERS pharmacovigilance analysis (PubMed)checked 2026-07-23
- FDA alert on compounded semaglutide dosing errorschecked 2026-07-23
Related records in this section
Section hub: Regulatory standards · Methodology · Status definitions · Right to respond
Frequently asked questions
Are compounded GLP-1 preparations more dangerous than approved ones?
No study has compared health outcomes between them, so that question is unanswered. What is documented is a specific failure mode — ten-fold dosing errors from multi-dose vials — that approved pen presentations structurally prevent.
How many people have been harmed?
No source supports an incidence figure. Poison centre data counts contacts, FAERS counts voluntary reports, and neither has a population denominator.
What is the ten-fold error?
Drawing ten times the intended dose from a multi-dose vial, usually through confusion between units, millilitres and milligrams on a syringe not calibrated for the product.
Does this apply to all compounded preparations?
The documented risk attaches to presentation rather than to compounding as such. A single-dose presentation with matched syringes and counselling removes most of the failure mode.
Has FDA said anything directly?
Yes. FDA issued a safety communication on dosing errors with compounded injectable semaglutide, naming both patient measurement and prescriber miscalculation as failure points.
Sources
- Journal of Medical Toxicology 2026;22:275-285
- Poison control case series (PubMed 37392810)
- FAERS pharmacovigilance analysis (PubMed 40285721)
- FDA alert on compounded semaglutide dosing errors
Update history
| Date | Change |
|---|---|
| 2026-07-23 | Record published at current snapshot. |
Dates change only for substantive updates. Entities may submit a correction or response through the right-to-respond process.