{ "schemaVersion": "1.0.0", "datasetVersion": "1.0.0", "reviewedAt": "2026-08-18", "title": "LifeScore compounded GLP-1 evidence audit", "boundary": "Editorial evidence audit of published sources. Not medical, dosing, sourcing or purchase advice, not a safety verdict on any product, and not proof that an untested claim is false.", "licence": { "licence": "CC-BY-4.0", "licensor": "lifescore", "covers": "The selection of sources, the gate scheme, the coding decisions, and the original summary text written by lifescore Edge.", "doesNotCover": "The cited publications themselves, which remain with their own rights holders and are referenced by DOI or PMID rather than reproduced; and the underlying facts, which are not subject to copyright.", "note": "The licence grants only such rights as the licensor holds. Nothing here asserts ownership of the audited literature." }, "auditorIndependence": { "auditor": "lifescore Edge", "publisher": "lifescore", "declared": [ "Edge does not currently carry advertising, sponsorships, affiliate links or paid placements.", "The people and organizations covered do not approve articles before publication.", "None of the audited parties commissioned, funded, reviewed or saw this audit before publication.", "lifescore sells no GLP-1 product, compounded or approved, and holds no position on either side of the comparison." ], "limits": [ "This audit was not externally peer reviewed.", "The auditor is the publication itself; there is no institutional affiliation and no third-party editorial board.", "Authorship is organizational rather than individual, so the coding cannot be attributed to a named expert with a public track record.", "Independence from a commercial interest is not a claim of expertise, and it does not make a coding decision correct." ], "standardsUrl": "https://lifesco.re/edge/methodology/" }, "method": { "question": "Do compounded GLP-1 products have the same human evidence as the approved ones?", "selection": { "question": "Do compounded GLP-1 products have the same human evidence as the approved ones?", "databases": [ "PubMed via the NCBI E-utilities API", "Europe PMC" ], "searchedOn": "2026-08-18", "query": "compounded semaglutide OR compounded tirzepatide OR compounded GLP-1", "recordsReturned": 106, "recordsWithCompoundInTitle": 32, "unitOfAudit": "one published source", "inclusion": "A source enters the audit when it bears on whether a compounded GLP-1 product has the human evidence of the approved one: it administered such a product to a person, examined its composition, aggregated adverse-event reports about it, or states a professional body's position on it.", "exclusion": "Sources that carry the search terms without bearing on that question are not coded. Exclusions with a stated reason are listed in the dataset's `excluded` field; sources surfaced but not read in full are listed separately in `screenedNotRead` and carry no gate coding.", "readingDepth": "Each audited record was read from the indexed PubMed record or the publisher's own article page, including conflict-of-interest and funding statements, rather than from a search snippet.", "limits": [ "The 106-record screen is stated, not reproduced: the screened record list was not captured, so this dataset cannot demonstrate that only two studies administered a compounded product.", "The 106 records were never counted for composition studies, so no claim is made about how many exist.", "Coding of a source's independence follows the gate dictionary and describes disclosed ties; it is not a finding of misconduct and not a legal judgment.", "This is an editorially selected set of six sources, not a systematic review and not a literature census." ] }, "unitOfAudit": "one published source", "gates": { "compoundedArticle": "Was the article actually studied a compounded preparation rather than the approved medicine? pass = a compounded product was the test article; partial = compounded products are involved but their identity rests on self-report or is mixed with other classes; stop = no compounded product was examined.", "humanAdministration": "Did this study give the product to a person and observe what happened? pass = administered and observed; partial = human exposure is reported but the study did not administer; stop = no administration.", "comparator": "Was there a control group or an originator comparator? pass = present; partial = a reference standard or a non-compounded reporting group, not a control arm; stop = none.", "equivalenceTested": "Does the study test whether a compounded product performs the same as the approved one? pass = equivalence or bioequivalence tested; partial = composition or signal compared without a clinical equivalence claim; stop = not tested.", "independentOfStake": "Is the work free of a party with a commercial interest in the answer? pass = no such party; partial = disclosed ties that do not place a product on either side; stop = authored, funded or staffed by a party selling one side of the comparison." }, "verification": "Each record was read from the indexed PubMed record or the publisher's own article page, including conflict-of-interest and funding statements, rather than from a search snippet.", "notReproducedHere": "The 106-record PubMed screen that establishes how few studies administered a compounded product is stated in the editorial evidence map but not captured as a record list, so this dataset cannot demonstrate it." }, "summary": { "sourcesAudited": 6, "administeredToHumans": 2, "testedEquivalence": 0, "freeOfStake": 1, "totalParticipantsAdministered": 95 }, "gateDistribution": { "compoundedArticle": { "pass": 4, "partial": 1, "stop": 1 }, "humanAdministration": { "pass": 2, "partial": 1, "stop": 3 }, "comparator": { "pass": 1, "partial": 2, "stop": 3 }, "equivalenceTested": { "pass": 0, "partial": 2, "stop": 4 }, "independentOfStake": { "pass": 1, "partial": 2, "stop": 3 } }, "sources": [ { "id": "chun-2025", "label": "Chun 2025 real-world cohort", "recordType": "retrospective cohort", "journal": "Diabetes, Obesity & Metabolism", "published": "2025-03", "pmid": "39776038", "doi": "10.1111/dom.16162", "testArticle": "Compounded semaglutide with cyanocobalamin, once weekly, subcutaneous", "participants": 94, "duration": "3 months", "reportedOutcome": "4.11 kg (SD 2.77), or 4.57 % of body weight (SD 2.96)", "statedLimit": "No control group, single commercial site, underlying data available on request only. The paper's title omits the cyanocobalamin its abstract reports.", "interestDeclaration": "All authors have received salary support from Restore Hyperwellness, the wellness chain whose customers were studied.", "gates": { "compoundedArticle": "pass", "humanAdministration": "pass", "comparator": "stop", "equivalenceTested": "stop", "independentOfStake": "stop" }, "sourceUrl": "https://doi.org/10.1111/dom.16162" }, { "id": "heor-self-experiment-2025", "label": "HEOR analysis with one participant", "recordType": "self-experiment", "journal": "International Journal of Pharmaceutical Compounding", "published": "2025-01", "pmid": "39921911", "doi": null, "testArticle": "Compounded tirzepatide, weekly", "participants": 1, "duration": "approximately 4 weeks", "reportedOutcome": "Not extracted here; the record is reported by its own author, who was the sole participant.", "statedLimit": "One participant, no control, no blinding. Sole affiliation is a one-person consultancy; the journal is published for the compounding industry. The paper states it is the first article focused on compounded tirzepatide.", "interestDeclaration": "Author and participant are the same person; published in the compounding industry's own journal.", "gateNotes": { "independentOfStake": "Coded partial rather than stop, and the difference matters. The sole affiliation is a one-person consultancy and the venue is the compounding industry's own journal, so a tie to one side of the comparison is disclosed and real. What is not established is that the author, a funder or an employer sells a compounded GLP-1 product — which is what stop asserts, and which the other three stop codings evidence by name through employment, funding or authorship. Reading a sales interest into a named individual from the publication venue alone would apply a standard not used on the rest of this set. The methodological problems are severe but belong elsewhere: one participant, no control and no blinding are carried by the comparator and equivalenceTested gates and by statedLimit." }, "gates": { "compoundedArticle": "pass", "humanAdministration": "pass", "comparator": "stop", "equivalenceTested": "stop", "independentOfStake": "partial" }, "sourceUrl": "https://pubmed.ncbi.nlm.nih.gov/39921911/" }, { "id": "lilly-b12-impurity-2026", "label": "Novel tirzepatide/B12 impurity", "recordType": "laboratory analysis", "journal": "Expert Opinion on Drug Safety", "published": "2026-05", "pmid": "42010938", "doi": "10.1080/14740338.2026.2663185", "testArticle": "Ten samples of compounded tirzepatide with B12 from compounding pharmacies, medspas and telehealth providers across seven US states", "participants": 0, "duration": null, "reportedOutcome": "Potency down to 43 % of the labelled amount; a previously unknown tirzepatide-B12 adduct at up to 10 % of total polypeptide content, with NMR showing altered tirzepatide structure.", "statedLimit": "Ten samples. Establishes that the adduct occurs and is widespread among products tested; does not measure how often, and tested no people. Clinical significance unknown, and the authors say so.", "interestDeclaration": "This manuscript was funded by Eli Lilly and Company; all five authors are Lilly employees. Lilly manufactures the approved tirzepatide products.", "gates": { "compoundedArticle": "pass", "humanAdministration": "stop", "comparator": "partial", "equivalenceTested": "partial", "independentOfStake": "stop" }, "sourceUrl": "https://doi.org/10.1080/14740338.2026.2663185" }, { "id": "novo-immunogenicity-2026", "label": "Impurities and potential immunogenicity", "recordType": "laboratory analysis", "journal": "Pharmaceutical Research", "published": "2026-07", "pmid": "42533250", "doi": "10.1007/s11095-026-04146-9", "testArticle": "Products from 36 commercial suppliers across four continents, including 13 injectable compounded semaglutide products, alongside follow-on and oral products", "participants": 0, "duration": null, "reportedOutcome": "Every compounded and follow-on sample showed an impurity profile distinct from the originator, including amino-acid deletions and additions and unidentified impurities. Under light exposure, compounded semaglutide diverged significantly in strength, total impurities and high-molecular-weight protein.", "statedLimit": "The impact of these impurity profiles on safety and efficacy has not been tested clinically; increased immunogenicity is potential, not demonstrated. Follow-on products are a separate class from US compounded products and their findings must not be merged.", "interestDeclaration": "Every author is a Novo Nordisk employee and shareholder, or was at the time of the work. Novo Nordisk supplied the originator comparators and manufactures the approved semaglutide and liraglutide products.", "gates": { "compoundedArticle": "pass", "humanAdministration": "stop", "comparator": "pass", "equivalenceTested": "partial", "independentOfStake": "stop" }, "sourceUrl": "https://doi.org/10.1007/s11095-026-04146-9" }, { "id": "faers-2026", "label": "FAERS disproportionality analysis", "recordType": "pharmacovigilance", "journal": "Expert Opinion on Drug Safety", "published": "2026-03", "pmid": "40285721", "doi": "10.1080/14740338.2025.2499670", "testArticle": "81,078 GLP-1 receptor agonist reports 2018-2024, of which 707 involved compounded products", "participants": null, "duration": "2018-2024", "reportedOutcome": "Elevated reporting odds ratios for preparation errors (48.92; 12.63-189.6), contamination (19.00; 4.24-85.03), compounding and manufacturing issues (8.51; 5.17-14.0) and hospitalisation (2.35; 1.94-2.83); lower for administration errors (0.29) and dosing errors (0.24).", "statedLimit": "A reporting odds ratio is a disproportionality measure, not a risk estimate, and does not establish cause. The database is voluntary and unverified, the denominator is unknown, and the window coincides with sustained press attention. The authors state the study was not designed to show cause and effect.", "interestDeclaration": "Geisinger and Binghamton University, with no product on either side. The lead author sits on the board of the Partnership for Safe Medicines, which contributed $850 to the work; a co-author was on a Pfizer- and Lilly-supported research team years earlier. Disclosures, not disqualifications.", "gates": { "compoundedArticle": "partial", "humanAdministration": "partial", "comparator": "partial", "equivalenceTested": "stop", "independentOfStake": "partial" }, "sourceUrl": "https://doi.org/10.1080/14740338.2025.2499670" }, { "id": "ada-statement-2025", "label": "American Diabetes Association statement", "recordType": "position statement", "journal": "Diabetes Care", "published": "2025-02", "pmid": "39620926", "doi": "10.2337/dci24-0091", "testArticle": "None. A professional-society position, not a study.", "participants": 0, "duration": null, "reportedOutcome": "No US Pharmacopeia monograph exists for semaglutide or tirzepatide, so no external standard defines a correct compounded preparation and no required test verifies one. Some compounders add vitamin B12 or B6, and some use salt forms in place of the base used in approved products.", "statedLimit": "A position statement is not a measurement. It establishes the absent standard, not an outcome in any patient.", "interestDeclaration": "Funded from ADA general revenue; the statement declares that no other entity, including industry, provided support. The only industry-free source in this set.", "gates": { "compoundedArticle": "stop", "humanAdministration": "stop", "comparator": "stop", "equivalenceTested": "stop", "independentOfStake": "pass" }, "sourceUrl": "https://doi.org/10.2337/dci24-0091" } ], "screenedNotRead": [ { "label": "Compounded semaglutide and tirzepatide products use unique formulations but efficacy and safety largely unknown", "journal": "Annals of Pharmacotherapy", "published": "2026-09", "pmid": "41689811", "recordType": "editorial", "reason": "Surfaced by the screen; not read in full, so no gate is coded." }, { "label": "Compounded GLP-1 drugs remain prevalent after drug shortage", "journal": "JAMA", "published": "2026-08", "pmid": "42566208", "recordType": "research letter", "reason": "No abstract indexed. Full text required before any citation." }, { "label": "Postshortage compounded GLP-1 RA market in two states", "journal": "JAMA Health Forum", "published": "2026-07", "pmid": "42467450", "recordType": "cross-sectional secret-shopper study", "reason": "Characterises the market after the shortages ended rather than the evidence for equivalence; surfaced on the freshness check and not read in full." } ], "excluded": [ { "label": "Rubio 2025 retrospective telehealth cohort", "publisher": "Athenaeum Scientific Publishers", "reason": "Checked 18 August 2026: zero records in Europe PMC, zero journals in DOAJ under that publisher, and absent from the 106-record PubMed result set. Surfaced by the consumer SERP, not usable as evidence." } ] }