Explained
What does the research actually show?
Built from index records dated 2026-10-13 Compiled by an automated pipeline, checked by The Peptide Observer editorial desk How this site works
Nothing in the records this site holds settles the question of safety either way, and for most entries here the records are not studies in people at all.
This is the honest answer to the most common question asked about peptides, and it is an answer about records rather than about compounds.
What we actually know
Each record this site indexes is sorted into one of four kinds by the subject headings the literature index assigned to it. Counted across every entry this site describes:
| Kind of study | Records | Share of all records |
|---|---|---|
| In a dish, outside a living body | 0 | none |
| In rats or mice | 154 | 48 per cent |
| A trial in people | 0 | none |
| Kind of study not stated | 167 | 52 per cent |
Not one of the 27 entries described here has a single record that the literature index marks as a trial in people.
A record counted as a human trial is counted from its own publication type in the literature index. Counting a trial is not the same as reading one, and no study behind these counts has been read in full.
What is not known
- Whether a compound causes harm. The adverse-effect row is empty for every entry described here, and an empty row records that nothing was taken from a source, never that nothing happens.
- What amount anything was reported at. No amount was taken from any source, so there is no figure on this site to read as guidance, and none is intended as any.
- How long any of this stays in a body. A reported half-life is absent for most entries, because no pharmacokinetics source was retrieved.
- Whether the records kept here point the same way as the far larger number that were left out.
- Counted exactly: the adverse-effect row is empty for 27 of the 27 entries described here, the reported-amount row is empty for 27 of them, and no half-life was retrieved for 27 of them.
How to read this
- Read the counts rather than any single record. One rodent study is a reason to look further, not a result.
- Where a count is zero, the page prints the zero. A missing bar would be easy to mistake for a small one.
Read next
Where this comes from
Study kinds are derived from the subject headings and publication types the literature index assigned to each record. Each entry's record list carries the identifier for every record counted.