What changed?

What WPF corrected or newly verified, what registries posted and what regulators decided, newest first. Each WPF change has its own citable page with the earlier and updated records. Three kinds are kept apart: a trial result or a regulator’s decision is a fact about the outside world; a WPF correction is a change to what WPF published, and the earlier record stays available.

22 of 22 changes.

  1. · WPF editorial correction

    Claim check corrected (Peptide Claims & Evidence 1.5.0 to 1.6.0): trial registration versus posted results

    Version 1.5.0 of the check said that in WPF's sample every completed or terminated trial that was opened had posted results. The sample includes the four registered BPC-157 trials, and one of them, the completed BPC-157 supplement study NCT07752381, had no results posted on ClinicalTrials.gov, so that sentence was wrong for the sample it described. Version 1.6.0 states the finding for that study, names the four BPC-157 records, adds the registry record to the evidence, and makes no statement about any other trial. The verdict stays Overstated. The registry gives no reason for the absence of results and does not mark the study as an FDA-regulated drug or device trial, so WPF draws no inference about why. Version 1.5.0 stays available unchanged at its own address.

    Source: WPF public correction record

    Affected: Claim check: trial registration versus posted results (/research/check/registered-trial-posted-results). Versioned records: 1.5.0 and 1.6.0.

  2. · WPF editorial correction · Statement about WPF's own sample corrected · verdict unchanged

    Trial registration and posted results: a completed BPC-157 supplement study in the sample had no results posted

    The verdict stayed Overstated. Version 1.5.0 said every completed or terminated trial in the sample that was opened had posted results. NCT07752381 is one of the four BPC-157 registrations in that sample, it was already recorded as completed in the BPC-157 tendon check, and it had no results posted, so that sentence was wrong for the sample it described on the original check date. The registry shows no reason for the absence and does not mark the study as an FDA-regulated drug or device trial, so WPF draws no inference about why no results are posted.

    Source: ClinicalTrials.gov NCT07752381 (checked 2026-11-22)

    Affected: Claim check: trial registration versus posted results

    Before and after, exact records and source locators · Correction Log entry

  3. · WPF editorial correction

    Claim checks corrected (Peptide Claims & Evidence 1.3.0 to 1.4.0): research use only, and BPC-157 safety

    Research use only: the check now states the claim as "A product labelled research use only is not a drug" and ties its Contradicted verdict to FDA's warning letter to Xcel Research LLC of 10 December 2024, the seven products it names (RETA retatrutide, CagriLean, cagrilintide, mazdutide, SEMA semaglutide, survodutide and sermorelin) and the website claims FDA cited as evidence of intended use. The earlier second clause, "and it is fine to use on yourself", is removed because the letter does not address it. BPC-157 safety: human dose figures were removed, the knee-pain series is stated as 16 patients followed (12 treated with BPC-157 alone) and as a source that reported pain outcomes, not safety data, and the registrations are named. Verdicts are unchanged (Contradicted by evidence; Insufficient evidence). Version 1.3.0 of all twelve checks stays available unchanged at its versioned address. AI-assisted source verification, checked 2026-11-15; human review none recorded.

    Source: WPF public correction record

    Affected: Claim checks: a research use only label (/research/check/research-use-only-label-product) and BPC-157 and safety in humans (/research/check/bpc-157-human-safety)

  4. · WPF editorial correction

    Claim checks corrected (Peptide Claims & Evidence 1.4.0 to 1.5.0): Humanin mouse lifespan, and SURMOUNT-5 confidence intervals

    Humanin: the check said no mouse lifespan result was found, because only abstracts had been read. The full text of the 2020 Aging paper (PMID 32575074) reports a mouse lifespan test of a humanin analogue started at 18 months of age in female mice, with no significant difference in lifespan between treated and control groups (Figure 3F, where the two survival curves overlap), alongside lower IGF-I. The check now records that null result and the authors' own caveat that it may reflect a sub-optimal dose or a late start; its Overstated verdict is unchanged. SURMOUNT-5: the check attributed the 95% confidence intervals (-21.4 to -19.1 for tirzepatide and -14.9 to -12.6 for semaglutide) to the registry record. They are from the NEJM abstract and each belongs to one arm's own mean. The registry posts the two arm means with standard errors and a between-group difference of -6.5 (95% CI -8.1 to -4.9), which the abstract does not report. The check now says so. The figures and verdicts are unchanged. Version 1.4.0 remains available unchanged. This work is AI-assisted source verification; documented human review: none recorded.

    Source: WPF public correction record

    Affected: Claim checks: Humanin and longevity (/research/check/humanin-extends-lifespan), weight-loss results comparable (/research/check/weight-loss-results-comparable) and tirzepatide and semaglutide in the same trial (/research/check/tirzepatide-semaglutide-same-trial)

  5. · WPF editorial correction

    Clarification of the 1.4.0 to 1.5.0 entry: which SURMOUNT-5 claim check corrected what

    The earlier entry on Claims & Evidence 1.5.0 described the SURMOUNT-5 correction as one change to one check. To be exact, two checks changed for different reasons. The weight-loss comparison claim check (weight-loss-results-comparable) had cited the registry alone for the arm means (-20.2% and -13.7%) and for the P value (P<0.001). The registry posts the arm means, with standard errors, and the publication abstract states them too; the P value comes from the abstract only, because the registry posts none. That check now names each source for what it contains. It never stated the arm confidence intervals. The arm confidence intervals (-21.4 to -19.1 for tirzepatide and -14.9 to -12.6 for semaglutide) had been attributed to the registry in the head-to-head claim check (tirzepatide-semaglutide-same-trial). They come from the publication abstract and each belongs to one arm's own mean; the registry posts the between-group difference of -6.5 (95% CI -8.1 to -4.9), which the abstract does not report. That interval correction belongs to the head-to-head check. No estimate and no verdict changed. The 1.4.0 records of both checks stay available unchanged. AI-assisted source verification; documented human review: none recorded.

    Source: WPF public correction record

    Affected: Claim checks: weight-loss results comparable (/research/check/weight-loss-results-comparable) and tirzepatide and semaglutide in the same trial (/research/check/tirzepatide-semaglutide-same-trial)

  6. · WPF editorial correction · Existing source newly verified · verdict unchanged

    Humanin and mouse lifespan: a mouse survival test in an existing paper was newly verified

    The verdict stayed Overstated. The earlier version said no mouse lifespan result was found in the abstracts read; checking Figure 3F established that the paper did test mouse lifespan and found no significant difference. That is a null result, which fits an overstated claim. WPF newly verified an experiment that already existed in a paper published in 2020; the paper did not change.

    Source: Aging (Albany NY) 2020, PMID 32575074 (PMC7343442) (checked 2026-11-16)

    Affected: Claim check: Humanin and longevity

    Before and after, exact records and source locators · Correction Log entry

  7. · WPF editorial correction · Study classification corrected · verdict unchanged

    Semaglutide SOUL trial: study classification corrected from Phase 1–2 to Phase 3

    The correction changes the study classification of one source record, from Phase 1–2 to Phase 3. It does not change the stored evidence-strength grade (A, 86, provisional), the grade's rationale, any trial result or any claim-check verdict. The grade is an automated baseline that the pipeline which produced it will recompute later for the corrected tier; if that changes the number, it will be recorded as its own change.

    Source: ClinicalTrials.gov NCT03914326 (checked 2026-11-16)

    Affected: Semaglutide profile: source record for PMID 40162642 (SOUL)

    Before and after, exact records and source locators · Correction Log entry

  8. · WPF editorial correction · Source attribution corrected · verdict unchanged

    SURMOUNT-5 confidence intervals: arm intervals come from the abstract, the between-group interval from the registry

    The correction changed source attribution only. Every estimate and every verdict stayed the same. The intervals were already correct; the records now say that the arm intervals come from the publication abstract and the between-group interval from the registry results, so a reader who goes to either source finds the figure where WPF says it is.

    Source: NEJM abstract, PMID 40353578 (checked 2026-11-16)

    Affected: Claim check: tirzepatide and semaglutide in the same trial; Claim check: weight-loss results comparable; Flagship result: SURMOUNT-5, week 72

    Before and after, exact records and source locators · Correction Log entry

  9. · WPF editorial correction

    BPC-157 tendon-healing claim check corrected (Peptide Claims & Evidence 1.2.0 to 1.3.0)

    The Overstated verdict is unchanged. Version 1.3.0 states the 2026 rat study's allocation from its Methods and Figure 1 (eight rats per group; four tendons per group for biomechanical testing and four per group for histological, histochemical and immunohistochemical analysis, so n = 4 per group for each type of analysis), and records the Figure 2 maximum-load values and the total Bonar and Movin scores in Tables I and II with their locations (the full text was read). The knee-pain locator now separates the systematic review's 7 of 12 patients reporting relief for more than six months from the original uncontrolled case series' 11 of 12 reporting improvement, and says that series assessed knee pain, not tendon healing. The four BPC-157 registrations were rechecked and are linked directly; the human-trial conclusion is stated as limited to the sources and registries searched. Animal dose figures were removed from the check. Version 1.2.0 stays available unchanged at its versioned address (checksum prefix c04209c4). AI-assisted source verification, checked 2026-11-15; human review none recorded.

    Source: WPF public correction record

    Affected: Claim check: BPC-157 and tendon healing (/research/check/bpc-157-tendon-healing)

  10. · WPF editorial correction

    SURMOUNT-5 and retatrutide phase 2: how treatment discontinuation was handled (Flagship 1.4.0)

    Release 1.4.0 adds, for SURMOUNT-5 and the retatrutide phase 2 trial, how people who stopped the study drug were handled (from each analysis plan) and the registry's study-completion counts, and states that counts of people who stopped the drug and stayed in the study are not posted. Two re-runnable checks were added and all 22 automated checks were re-run and passed. No displayed value changed. Human review is none recorded.

    Source: Flagship Weight Management 1.4.0 release notes

    Affected: Results surmount-5-weight-72w and retatrutide-phase2-weight-48w

    Cited an older version? Version 1.3.0 record of SURMOUNT-5 (still resolves) still returns what you cited. Updated: Current record, version 1.4.0. Rewritten: discontinuation. Added: participant-flow locator and check, unverified list. Unchanged: resultId, PMID, DOI, registry ID and every displayed number.

  11. · WPF editorial correction

    SURMOUNT-1 estimand disclosed (Flagship Weight Management 1.0.0 to 1.1.0)

    Release 1.0.0 labelled the SURMOUNT-1 result a treatment-regimen estimand and said the public analysis plan had not been reviewed. Release 1.1.0 reviewed it: the plan calls the matching estimand "hybrid", and the registry's posted results are a different, efficacy-style analysis, so they do not check the displayed values. This is disclosed, not resolved. The displayed numbers did not change.

    Source: WPF release history and 1.1.0 changelog

    Affected: Result surmount-1-weight-72w

    Cited an older version? Version 1.0.0 record (still resolves) still returns what you cited. Updated: Current record, version 1.4.0. Rewritten: estimand. Added: analysisPlan, analysisPlanUrl, analysisPlanSections. Also updated: designExplanation, verification, citation, versionedRecordUrl. Unchanged: resultId, PMID, DOI, registry ID and every displayed number.

  12. · WPF editorial correction

    JAMA cohort: supplement and Discussion wording both recorded (Flagship 1.3.0)

    Release 1.3.0 records both the supplement's description of the modified-ITT effects as somewhat attenuated and the Discussion's description of the differences as similar, and adds the full-text Figure 4 reading and the SURMOUNT-1 protocol check. It is the AI-assisted source verification release. Human review is none recorded.

    Source: WPF AI-assisted source verification, 2026-11-12

    Affected: Result jama-intern-med-2024-realworld

  13. · WPF editorial correction

    Retatrutide phase 2: population and comparator fields corrected (Flagship 1.0.0)

    The population and comparator fields of the retatrutide phase 2 result were corrected after automated source validation against the registry and publication. No human adjudication is recorded.

    Source: WPF release history

    Affected: Result retatrutide-phase2-weight-48w

  14. · WPF editorial correction

    Homepage preservation, archive count, and language coverage clarified

    The absolute statement that every observation is preserved was replaced with wording about context, contributor choices, and review and publication. The homepage now reads the public archive count using its existing approved-record exclusions and privacy threshold, with a count timestamp and an unavailable state on query failure. The footer links to Language Integrity and distinguishes the 15 active language choices from the broader 20-language program. No scientific or observation records were changed.

    Source: WPF public correction record

    Affected: Homepage, footer, and Language Integrity page

  15. · WPF editorial correction

    Multilingual corpus accounting superseded

    The August 22, 2026 reconciliation reported a corrected total of 12 corpora / 23,808 units. That figure has since been superseded. Current governed automated-assurance reporting covers 18 Core Global 20 corpora and 36,126 translated units, of which 35,655 pass automated assurance. The August 22 entry remains part of the public record and is not modified. The total returned to 18 later on August 22 because governed corpora were then generated for the six canonical Core Global 20 locales that had never had one — Bengali, Marathi, Nigerian Pidgin, Tamil, Telugu and Urdu — completing the roster, rather than by restoring the six non-roster corpora (Persian, Italian, Dutch, Polish, Ukrainian and Traditional Chinese) that the reconciliation had correctly removed.

    Source: WPF public correction record

    Affected: Public multilingual coverage and assurance accounting

  16. · WPF editorial correction

    Multilingual corpus accounting reconciliation

    Multilingual corpus reporting was reconciled to the canonical Core Global 20 language roster. Six generated corpora associated with a superseded runtime-language roster were removed from Core-20 assurance totals. The reported automated-assurance total changed from 18 corpora / 35,712 units to 12 corpora / 23,808 units at the time of reconciliation. No underlying scientific evidence was deleted; the correction changed language classification and public assurance accounting so Core-20 reporting reflected only corpora belonging to the canonical public language roster at that time.

    Source: WPF public correction record

    Affected: Public multilingual coverage and assurance accounting

  17. · Trial result posted

    NCT05822830: results first posted

    SURMOUNT-5: tirzepatide versus semaglutide in adults with obesity. The registry lists the study as completed with results posted.

    Source: ClinicalTrials.gov NCT05822830 (checked 2026-11-13)

    Affected: Semaglutide research journey

  18. · Regulatory decision

    FDA: Accelerated approval (new molecular entity for FORZINITY (elamipretide) injection

    United States. FORZINITY (elamipretide) injection, NDA 215244 (Stealth BioTherapeutics). FDA states the approval rests on improved knee-extensor muscle strength, a measure reasonably likely to predict benefit, and requires a confirmatory randomized trial. It concerns Barth syndrome only. It says nothing about primary mitochondrial myopathy, the condition tested in MMPOWER-3.

    Source: FDA press announcement, 19 September 2025, "FDA Grants Accelerated Approval to First Treatment for Barth Syndrome"; openFDA Drugs@FDA record NDA215244 (ORIG-1, status AP, 2025-09-19, priority) (checked 2026-11-13)

    Affected: Elamipretide research journey

  19. · Trial result posted

    NCT03098797: results first posted

    Elamipretide in Barth syndrome (TAZPOWER). The registry lists the study as completed with results posted. The U.S. decision below concerns this condition, not primary mitochondrial myopathy.

    Source: ClinicalTrials.gov NCT03098797 (checked 2026-11-13)

    Affected: Elamipretide research journey

  20. · Trial result posted

    NCT03548935: results first posted

    STEP 1: semaglutide 2.4 mg once weekly in adults with overweight or obesity. The registry lists the study as completed with results posted.

    Source: ClinicalTrials.gov NCT03548935 (checked 2026-11-13)

    Affected: Semaglutide research journey

  21. · Regulatory decision

    FDA: New drug application approval (original) for WEGOVY (semaglutide) injection

    United States. WEGOVY (semaglutide) injection, NDA 215256. Other semaglutide products (Ozempic, Rybelsus) are separate applications with their own decisions and are not covered by this record. Approval is regulatory context, not evidence that a trial result applies to a given person.

    Source: openFDA Drugs@FDA record for NDA215256 (submission ORIG-1, status AP, 2021-06-04; latest efficacy supplement 2026-03-19) (checked 2026-11-13)

    Affected: Semaglutide research journey

  22. · Trial result posted

    NCT03323749: results first posted

    MMPOWER-3: elamipretide in primary mitochondrial myopathy, with open-label extension. The registry lists the study as terminated with results posted. Registry reason for stopping: "Part1,double blind portion of the trial did not meet the primary end points".

    Source: ClinicalTrials.gov NCT03323749 (checked 2026-11-13)

    Affected: Elamipretide research journey

What each kind means

Trial result posted
A trial registry posted results. WPF did not change its own record because of this entry.
Regulatory decision
A regulator acted on a product. This is regulatory context, not scientific evidence.
WPF editorial correction
WPF corrected or clarified something it had published. The earlier record stays available.
What changed? — World Peptide Foundation