Source hierarchy
For any medical or regulatory claim we look for sources in this order and cite the highest-ranked one we can find:
- FDA drug labels, approval databases (Drugs@FDA), safety communications and other official regulatory material.
- ClinicalTrials.gov registrations and posted results.
- PubMed-indexed systematic reviews and meta-analyses.
- Randomised controlled trials in humans.
- Other peer-reviewed human studies (observational, open-label, case series, case reports).
- Preclinical animal or cell studies.
- Reputable expert analysis and commentary.
Competitor blogs, vendor pages, forums and social media are used only to understand what readers are asking. They are never cited as evidence for a medical claim.
Assigning an evidence level
A grade applies to a specific claim for a specific population and outcome. We ask, in order: Is there regulatory approval for this indication? Are there adequately powered randomised trials, and do they agree? Is the human evidence controlled or uncontrolled? Is there any human evidence at all? Do credible studies conflict? Does the claim exceed what the data show? The answers map to the six levels shown on this page. When a peptide has different grades for different uses, the profile shows the grade for its most-discussed use and the research-area pages show area-specific grades.
Regulatory status is separate
We record U.S. regulatory status from FDA sources and display it beside, never instead of, the evidence grade. Six statuses are used: FDA-approved for this use; FDA-approved for a different use; investigational; not FDA-approved; withdrawn or discontinued; and status requires verification. The last one is deliberate: if we cannot confirm a status from primary regulatory material at review time, we say so rather than guess. Regulatory positions change; the “updated” date on each page tells you when we last checked.
Citation standards
- Every citation links to the real source (a DOI, PubMed record, ClinicalTrials.gov entry, FDA document or publisher page).
- Each reference lists study type; where relevant we also record population, sample size, primary result, key limitations and year.
- Inline citations are numbered in order of first use and link to the reference list, which links back to the text.
- We do not cite abstracts we could not read, press releases as evidence of efficacy, or preprints without labelling them.
- An automated citation audit checks that every reference URL resolves; failures are fixed or the claim is removed.
Language rules
We lead with the reader’s question and answer it early. We explain mechanisms without implying that a mechanism proves an outcome. We separate demonstrated benefits, possible benefits, theoretical mechanisms, anecdotal claims and unsupported marketing. We avoid hype words (“miracle”, “cure”, “game-changing”, “melts fat”, “reverses aging”, “guaranteed”). We do not provide dosing, protocols or sourcing guidance.
Updates and freshness
Every page shows a published date and an updated date. Regulatory pages and profiles for peptides with active trials are re-checked at least quarterly; substantive changes are noted in the article’s correction history and on the corrections page.
AI assistance
We use AI tools to draft, structure and check content. Every AI-assisted draft is created as an unpublished draft that cannot appear on the public site until a human editor has verified each claim and citation against primary sources and explicitly approved it. See the editorial policy.