The canonical rubric behind every evidence-tier badge across PeptideDecoded. Every deep-dive, comparison, hub card, and MedicalWebPage JSON-LD field is graded against these four criteria — read this page to evaluate our tier assignments (or the tier assignments other sites skip). Author credentials, contact channels, and a general overview live at /about.
Three tiers — Strong, Tier-2, Anecdotal — rendered as high, med, low on the deep-dive pages and the MedicalWebPage.evidenceTier JSON-LD field. Each tier maps to a concrete worked example you can click into.
highMultiple positive human RCTs at scale, an FDA-approved indication (or equivalent regulator approval), and results reproduced across independent research groups.
medLimited human RCTs (often one or two small-to-moderate trials), strong mechanistic or animal data, or reproducible across delivery routes but with only one positive RCT (the topical case for GHK-Cu is the canonical example).
lowNo published human RCTs for the marketed indication. Animal, mechanistic, case-report, or community-only data — credible enough to investigate, not enough to claim efficacy.
Every peptide in the catalog is scored against the same four criteria. Assigning a tier badge is the last step after all four are evaluated — no single criterion on its own decides the tier.
How many randomized controlled trials in humans exist for the marketed indication, and how many of those are positive? Strong requires ≥ 3 independent positive RCTs; Anecdotal means zero peer-reviewed human RCTs for the indication, regardless of how strong the mechanism looks in vitro.
Has the evidence been through a peer-reviewed journal indexed on PubMed? Preprints on medRxiv or bioRxiv are explicitly capped at Tier-2, even when the n is large, because they have not yet been independently verified.
Studies below n = 30 per arm are considered preliminary even when positive. A single positive pilot trial at n = 12 cannot lift a peptide from Anecdotal — it caps it at Tier-2 unless combined with a power-adequate replication or strong meta-analytic support.
Is the result reproduced across multiple independent research groups, or does it rely on a single lab? BPC-157 is the canonical counter-example — its entire positive evidence base comes from the Sikiric lab in Zagreb, which lowers its reproducibility score and contributes to the Anecdotal tier assignment.
How the four criteria resolve into the assigned tier for each catalog peptide. Numbers match the comparison reference; check both for the same row to confirm the assignment.
| Peptide | Tier | RCTs | Peer-review | Sample size | Reproducibility | One-line justification |
|---|---|---|---|---|---|---|
| Semaglutide (GLP-1) | High | ≥ 10 RCTs (STEP, SUSTAIN, SELECT) | PubMed, NEJM, Lancet | n > 1,000 per arm | Multiple independent groups · FDA-approved | Reproduced efficacy and cardiovascular outcomes across at least 10 large RCTs. |
| GHK-Cu (topical) | Med | 1–2 RCTs (Lorentic 2012, cosmetic aging endpoints) | PubMed, peer-reviewed | n = 67–120 | Reproduced across cosmetic-research groups · topical only | Single RCT with positive cosmetic endpoints, mechanistic support, no injectable RCT. |
| GHK-Cu (injectable) | Low | 0 RCTs | — | — | Case reports only | No published human RCTs for injectable use; cosmetic claims extrapolated from topical data. |
| BPC-157 | Low | 0 RCTs | Preclinical only | Animal n = variable | Single-lab dominant (Sikiric) | Animal evidence dominated by a single research group; no human RCT exists. |
| TB-500 (Thymosin β-4) | Low | 0 RCTs for tendon repair | Preclinical + FDA orphan-drug designation | Animal n = variable | Goldstein lab dominant; multiple independent confirmations in animal models | Animal wound-healing and angiogenesis data; no human efficacy RCT; FDA orphan-drug for corneal repair. |
| Ipamorelin | Low | 0 RCTs for musculoskeletal / anti-aging indications | Preclinical + small PK studies | n < 30 for indication RCTs | Single sponsor-driven PK trials | Selective GH secretagogue mechanism well-characterized; no efficacy RCTs for any marketed indication. |
| Sermorelin | Low | Diagnostic-only RCTs (Reutens 1996, Walker 1994) | PubMed, peer-reviewed | n < 30 per arm | Diagnostic endpoints only | Approved for diagnostic GH-stimulation testing; no efficacy RCT for adult anti-aging — withdrawn 2008. |
A tier assignment is only as useful as the evidence trail behind it. We search for the primary record, verify what the study actually tested, and render the source so a reader can inspect it without guessing which claim it supports.
We start with targeted PubMed searches combining the peptide, indication, and study design terms such as randomized, controlled, clinical trial, or systematic review. For each potentially relevant record, we verify the title, authors, journal, publication year, and PMID against the PubMed entry. We then check the study design, population, sample size, comparator, endpoint, and reported direction of effect — reviewing the full text when available rather than relying on an abstract alone.
Mechanistic, animal, case-report, and preprint records remain useful context, but they are labeled as such and cannot be presented as human efficacy evidence. A citation is not counted toward the RCT criterion simply because it contains the word “trial” or appears in a review's reference list.
We check ClinicalTrials.gov when a registered study could change a page's evidence picture. The review includes the registration record, recruiting or completed status, enrollment, intervention, primary outcome measures, posted results, and any linked publications. A registered, ongoing, terminated, or completed study without posted results is not treated as proof of efficacy; when results are available, we distinguish the registry record from the peer-reviewed publication and assess both.
| Source | What we verify | How it appears on a page |
|---|---|---|
| PubMed record | Title, authors, journal, year, PMID, design, population, endpoint, and full-text findings. | Direct source link or linked PMID beside the claim or evidence-table row. |
| ClinicalTrials.gov record | Registration, status, enrollment, outcome measures, posted results, and linked publications. | Registry link labeled as registration or results context, never as a substitute for published efficacy. |
Deep-dive evidence tables use a short, italic citation treatment so the source and its limitation stay adjacent to the evidence label. Where a stable source record exists, we link directly to it; otherwise the PMID or trial identifier is shown for verification.
Published study: BPC-157 preclinical evidence — source record: PMID 32628526; the adjacent label identifies the evidence as animal data, not human efficacy.
Registered study: Clinical trial registration and status — ClinicalTrials.gov record search; registration context is kept distinct from a peer-reviewed outcome report.
We look for conflicts before assigning weight to a study or claim, then disclose known conflicts instead of silently treating them as irrelevant.
During review, we record author affiliations, funder affiliations, sponsor-funded studies, and whether the sponsor designed the study, controlled the data, or supported publication. Sponsor funding does not automatically invalidate a result, but it is relevant context alongside study design, prespecified endpoints, attrition, and independent replication.
PeptideDecoded's paid-guide model funds the site through guide purchases. We do not sell peptides, recommend vendors or clinics, accept vendor sponsorship, or use affiliate links to direct readers to a supplier. The editorial operation, paid guides, vendors, clinics, and any affiliate relationships are kept separate so a commercial relationship cannot determine a tier assignment.
When a known conflict, sponsor role, or relevant limitation could change how a reader interprets the evidence, we surface it in the article, citation note, or evidence table. The full sourcing, disclosure, and corrections policy is available on Editorial Standards.
Every page PeptideDecoded publishes that carries an evidence-tier badge can be traced back to this rubric. Use this list to jump into the worked-example pages and compare the tier assignment against the underlying evidence.
Peptide deep-dives
Peptide comparisons
Long-form guides, regulatory log, and dashboard
The rubric is applied by the editorial team during the evidence review documented on the Editorial Standards page. Tier assignments are revisited when a new RCT is published, when an existing peer-reviewed finding is replicated or challenged, when FDA actions reclassify a peptide's regulatory status, or as part of the 6-month content-review cycle. Tier badges on deep-dive pages and MedicalWebPage JSON-LD fields are updated in the same edit.
If a peptide's tier changes — for example, if the first positive BPC-157 human RCT is ever published — the deep-dive, the comparison pages, the relevant hub cards, and the MedicalWebPage.evidenceTier JSON-LD field will all be re-evaluated against this rubric and updated atomically. The "as of" date in the JSON-LD dateModified field is the authoritative record of when the last tier reassessment happened.
For questions about a specific tier assignment or to flag evidence we may have missed, see the contact channels on the Editorial Standards page.
This is a methodology document, not medical advice. The rubric described here grades the strength of scientific evidence for different peptide indications — it does not recommend any peptide for any individual, condition, or use case.
PeptideDecoded does not sell peptides, recommend vendors, provide dosing protocols, or make claims about safety or efficacy for any specific individual. Work with a qualified physician who can evaluate your full medical history, current medications, and individual risk factors before making any decision about peptides.