Compounded GLP-1 Research: the Trial That Does Not Exist (2026)

A pharmacovigilance analysis of adverse-event reports naming compounded products specifically, compared against the brands. The compounded reports clustered around medication errors, accidental overdose, confusion over dosing, and reactions at the injection site. ⚠ Note what that pattern suggests: the signal is concentrated in how the product is delivered — vials and syringes rather than pens — rather than in the molecule. Spontaneous reports have no denominator and cannot establish cause, so this identifies where to look rather than what is true.

PMID 40285721 ↗DOI 10.1080/14740338.2025.2499670 ↗

The most consequential analytical paper here. Mass-spectrometry analysis of mass-compounded tirzepatide-with-B12 products from multiple US suppliers found an impurity nobody had characterized before, present across many lots — not the active peptide, and not one of the known degradation products either. Its toxicology is unknown and its potential to provoke an immune response is unstudied. This does not prove harm. It does establish that something is in these vials that is not in the approved drug and that nobody has characterized.

PMID 42010938 ↗DOI 10.1080/14740338.2026.2663185 ↗

A comparison of compounded and follow-on preparations against the reference products using chromatography, mass spectrometry and aggregation assays, finding measurable differences in oligomer content, particulate load and impurity profile. ★ The authors are careful in a way worth copying: they explicitly decline to infer clinical inequivalence, noting that bioavailability, efficacy and immunogenicity all require dedicated comparative trials. What they show is that the manufacturing gap is real and measurable at molecular level.

PMID 39379664 ↗DOI 10.1007/s11095-024-03771-6 ↗

The European counterpart, drawn from the EMA's pharmacovigilance database and isolating reports where the product was flagged as counterfeit rather than genuine. Reports of low blood sugar, of dosing errors, and of severe gastrointestinal events all ran high against authentic semaglutide. ⚠ Counterfeit and compounded are not the same category, and conflating them is a common error — but this is one of only two analyses that manage to separate non-genuine supply from the authentic background at all.

PMID 42137313 ↗DOI 10.3389/fphar.2026.1805842 ↗

The foundational clinical paper on how people actually get hurt, and it is not about chemistry. Patients given multi-dose vials drew their doses in insulin units instead of milligrams, producing tenfold and twentyfold overdoses, with several needing emergency care for protracted vomiting and dehydration. ★ This is the hazard that distinguishes a compounded vial from a prefilled pen, and it is entirely a delivery-format problem — the same molecule in a pen does not produce it.

PMID 37392810 ↗DOI 10.1016/j.japh.2023.06.017 ↗

The population-level version of the case series above, combining adverse-event, poison-control and adjacent surveillance data across the years compounded distribution expanded. Accidental overdose reports rose sharply in both absolute and proportional terms, concentrated in multi-dose-vial dispensing rather than prefilled pens. It cannot isolate cause, but it demonstrates that the pattern one poison center documented is national rather than local.

PMID 39552465 ↗DOI 10.1080/14740338.2024.2430306 ↗

The American Diabetes Association's formal position, and the most authoritative US specialty guidance on this question. Its conclusion: reach for a compounded product only when a documented shortage makes the approved drug genuinely unavailable, and only through a state-licensed pharmacy filling a prescription written for that patient. It catalogs the documented signals — dosing errors, salt-form substitution, impurity findings — alongside the complete absence of comparative efficacy data.

PMID 39620926 ↗DOI 10.2337/dci24-0091 ↗

A practical review written for the clinician who gets asked about this in a ten-minute appointment. It covers the shortage framework, the difference between the two categories of compounding pharmacy, the absence of pre-market equivalence testing, the salt-form and impurity concerns, and what a patient should be told before a prescription is written. Pairs naturally with the ADA statement — one sets the position, this one operationalizes it.

PMID 40966636 ↗DOI 10.37765/ajmc.2025.89787 ↗

A cross-sectional audit of the direct-to-consumer sellers themselves, cataloging pricing, dosing instructions, oversight claims, salt-form disclosure and pharmacy licensure across the sites reaching one state's residents. ★ This is the paper closest to what we do here, and its finding matches ours: disclosure of compounded status, salt form and sourcing is inconsistent and frequently incomplete. No bioequivalence trial can capture that, and it determines what a buyer actually receives.

PMID 39776466 ↗DOI 10.1080/20523211.2024.2441220 ↗

A regulatory analysis using the semaglutide and tirzepatide shortages as a case study, contrasting the two compounding pathways and walking through the shortage-resolution timeline and the cease-compounding deadlines that followed. Useful precisely because the legal position moved so far in eighteen months: a prescription that was lawful when written may not be lawful to refill, and this is the clearest reference for why.

PMID 42143782 ↗