Skip to content
Peptideworth
Menu

Evidence review

How to Tell a Peptide Claim Is Outrunning Its Evidence

Twelve tells, each with a real case from the registry or the literature. Overstated peptide claims tend to fail on one of them, and usually the same one.

By Grant Delaney, Research Editor

The shape of the problem

Almost nobody lies about peptides outright. The claims that mislead are usually assembled from true statements about a slightly different thing — a different species, a different molecule, a different route, a different indication, or a different endpoint.

What follows is a list of the specific substitutions, each with a case you can check.

Tell 1: the species swap

The claim is about people. The evidence is about rodents.

A 2026 review of injectable peptide therapy for orthopaedic and sports medicine physicians found that thymosin beta-4 and its derivative TB-500 promoted angiogenesis and tissue repair in preclinical models, and in the same breath that human orthopaedic data are lacking1.

The preclinical half travels; the second half rarely does. When a claim's citations are all animal work, the honest sentence is "this has been shown in animals", and that sentence sells much less.

Tell 2: the molecule swap

TB-500 is N-acetylated LKKTETQ — the seven-residue segment at positions 17 to 23 of thymosin beta-4. Thymosin beta-4 is a 43-residue protein. A 2026 review lists them as separate entries2.

Decades of research on the protein is not decades of research on the fragment. Our TB-500 page keeps the two apart line by line, because a page that lets them blur is useless.

Tell 3: the route swap

The published phase 3 evidence for thymosin beta-4 is an ophthalmic solution instilled in an eye, in 18 patients with neurotrophic keratopathy, with the primary healing comparison reported at p = 0.06563.

A result from an eye drop is not evidence for a subcutaneous injection into a hamstring. Route is part of the claim.

Tell 4: the indication swap

A ClinicalTrials.gov search for ipamorelin returned 3 studies in August 2026. One completed phase 2 trial studied post-operative ileus in 117 participants4. The other studied recovery of gastrointestinal function after bowel resection in 3205.

Neither studied growth hormone secretion or body composition. A compound can have a real trial history that has nothing to do with what it is discussed for.

Tell 5: the tier swap

A marker moved, and the claim is about an outcome. Inflammation markers, IGF-1, body composition on a scan — all real measurements, none of them an event.

Converting a marker into an outcome takes a trial like SELECT: 17,604 patients, a mean 39.8 months of follow-up, and a composite of cardiovascular death, non-fatal myocardial infarction and non-fatal stroke — 6.5% versus 8.0%, hazard ratio 0.80, 95% CI 0.72 to 0.906. That is what the upgrade costs. The full argument is in surrogate endpoints.

Tell 6: the endpoint swap

The number in the headline is not the trial's primary endpoint.

The retatrutide phase 2's declared primary was percent change in body weight at week 24, where the 12 mg arm showed −17.5% against −1.6%. The widely quoted −24.2% is the 48-week figure, a secondary endpoint7.

Both numbers are real and the trial reported both. Quoting the secondary as though it were the finding is the move worth catching — see what a phase 2 result means.

Tell 7: the phase halo

"Phase 3" sounds like the end of the argument. The thymosin beta-4 phase 3 above enrolled 18 people3. STEP 1, also a phase 3, enrolled 1,961.

Phase is a stage, not a size and not a quality grade. What size actually buys is set out in why n=30 is not n=3,000.

Tell 8: the plan quoted as a result

A registered trial is a filing, not a finding.

The BPC 157 hamstring trial, NCT07437547, was first posted in February 2026 with an estimated primary completion of February 2027 and an estimated study completion of February 20288. No result from it can exist in August 2026. Any claim resting on it is resting on a protocol — the reading method is in how to read a record.

Tell 9: the silence

Completed trials that never reported are not neutral. ARISE-3, a phase 3 of a thymosin beta-4 ophthalmic solution with 700 participants, completed its primary endpoint in November 2020 and had no results posted on ClinicalTrials.gov as of August 20269.

Whether that silence means anything is unknowable in any single case. Whether it is common is measurable, and it is: see registered but never reported.

Tell 10: the unbounded universal

"No studies show harm." "Nothing has ever been reported." "Clinically proven."

These cannot be sourced, because no search establishes a universal negative. The repairable version names the search, the count and the date — "a ClinicalTrials.gov search for X returned N studies in August 2026, and none lists Y".

If a claim cannot be rewritten in that form, it is not a claim about evidence.

Tell 11: the identity gap

A protocol is being discussed for a substance nobody has characterised.

An analytical study of products sold online as TB500 and TB1000 reported that their content is not systematically consistent with the products' own former descriptions10. And FDA's published notes on peptides nominated for compounding repeatedly cite possible immunogenicity risk from aggregation and peptide-related impurities, with added complexity in characterising the active ingredient11.

When the material is unestablished, everything downstream of it is provisional — which is why what a certificate of analysis proves is a narrower list than people expect.

Tell 12: the disclaimer that contradicts the pitch

A product described as being for research use, discussed alongside an injection schedule for a person.

"Research chemical" is not a regulatory category. FDA's compounding framework evaluates substances as nominations to the 503A and 503B bulks lists, and publishes the safety concerns it identifies for each — a process with no tier called "research use"11. The phrase is doing something else, and what it is doing is worth knowing.

Six questions that settle most of it

What species was it in?

What exact molecule, by what route, in whom?

Is the quoted number the declared primary endpoint, at the declared timepoint?

How many people, and is that enough to see what is being claimed?

Was it blinded, and does that matter for this endpoint? Sometimes it does not — see open-label vs blinded.

Who paid, and did the results get posted? Both are one click away, and both are informative.

What a strong claim looks like

For contrast, here is one that survives every tell above. Semaglutide, in humans, by the studied route, in a named population, at the declared primary endpoint, blinded, in 17,604 people, sponsored by a company that posted and published the result, measuring events rather than markers6.

That is the standard. Very little of what is sold as a peptide is supported by anything of that shape, and saying so plainly is not scepticism — it is the accurate description. Our injury recovery and weight loss pages apply the same tests compound by compound, and the rest of this cluster lives in research.

Frequently asked questions

What is the most common way a peptide claim overstates its evidence?

Substitution. The claim is about humans and the evidence is about rodents; or about an injection and the evidence is an eye drop; or about one molecule and the evidence is about a longer protein containing it. Each substitution is individually small and each is checkable against the cited source.

How should an absence of evidence be stated?

With a named search, a count and a date. 'A ClinicalTrials.gov search for ipamorelin returned 3 studies in August 2026, and neither of the two completed phase 2 trials studied growth hormone secretion' is checkable. 'No studies show harm' is not, because no search can establish a universal negative.

Does a registered trial mean a compound has evidence behind it?

No. NCT07437547, a phase 2 trial of BPC 157 for hamstring strain, was first posted in February 2026 with an estimated primary completion of February 2027 and study completion of February 2028. A registration is a plan, and no result from that trial can exist yet.

What does a claim that survives all these tests look like?

SELECT is the example: semaglutide in 17,604 humans with pre-existing cardiovascular disease, blinded, at the declared primary endpoint, measuring events rather than markers — 6.5% versus 8.0% for a composite of cardiovascular death, non-fatal myocardial infarction or non-fatal stroke, hazard ratio 0.80 (95% CI 0.72 to 0.90). Very few peptide claims are built on anything of that shape.

References

  1. Mayfield CK et al (2026). Injectable Peptide Therapy: A Primer for Orthopaedic and Sports Medicine Physicians. Am J Sports Med. https://pubmed.ncbi.nlm.nih.gov/41476424/
  2. Mendias CL; Awan TM (2026). Safety and Efficacy of Approved and Unapproved Peptide Therapies for Musculoskeletal Injuries and Athletic Performance. Sports Med. https://pubmed.ncbi.nlm.nih.gov/41966639/
  3. Sosne G et al (2022). 0.1% RGN-259 (Thymosin ß4) Ophthalmic Solution Promotes Healing and Improves Comfort in Neurotrophic Keratopathy Patients in a Randomized, Placebo-Controlled, Double-Masked Phase III Clinical Trial. Int J Mol Sci. https://pubmed.ncbi.nlm.nih.gov/36613994/
  4. U.S. National Library of Medicine (2017). NCT00672074 — Safety and Efficacy of Ipamorelin for Management of Post-Operative Ileus. ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT00672074
  5. U.S. National Library of Medicine (2017). NCT01280344 — Safety and Efficacy of Ipamorelin Compared to Placebo for the Recovery of Gastrointestinal Function. ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT01280344
  6. Lincoff AM et al (2023). Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med. https://pubmed.ncbi.nlm.nih.gov/37952131/
  7. Jastreboff AM et al (2023). Triple-Hormone-Receptor Agonist Retatrutide for Obesity - A Phase 2 Trial. N Engl J Med. https://pubmed.ncbi.nlm.nih.gov/37366315/
  8. U.S. National Library of Medicine (2026). NCT07437547 — BPC 157 for Acute Hamstring Muscle Strain Repair. ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT07437547
  9. U.S. National Library of Medicine (2022). NCT03937882 — Assessment of the Safety and Efficacy of RGN-259 Ophthalmic Solutions for Dry Eye Syndrome: ARISE-3. ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT03937882
  10. Delcourt V et al (2023). TB500/TB1000 and SGF1000: A scientific approach for a better understanding of misbranded and adulterated drugs. Drug Test Anal. https://pubmed.ncbi.nlm.nih.gov/36482504/
  11. U.S. Food and Drug Administration (2026). Category 2 of the Bulk Substances Nominated Under Sections 503A or 503B of the Federal Food, Drug, and Cosmetic Act. FDA.gov. https://www.fda.gov/drugs/human-drug-compounding/safety-risks-associated-certain-bulk-drug-substances-nominated-use-compounding

Medical disclaimer: This content is for general educational purposes only and is not medical advice, diagnosis, or treatment. Always consult a licensed healthcare professional before starting, stopping, or changing any treatment.

Continue reading