Reading the research
Most bold claims about peptides rest on cell or animal studies. Here is how to tell what kind of evidence sits behind a claim, and how we label it.
The four evidence levels we use
| Label | What it means | What you can conclude |
|---|---|---|
| FDA-approved | The compound passed large, controlled human trials and the FDA approved it for at least one use. | Benefits and common side effects are well measured for the approved use and dose. Other uses may still be unproven. |
| Late-stage trials | Large human trials (phase 2b or 3) have reported results, but the US has not approved it. | Effects in people are fairly well described. Rare risks and long-term safety are still being learned. |
| Early human data | Small human studies exist: pilot studies, phase 1 or 2, case series, or old trials that never went further. | There are hints of an effect in people, but results can easily be due to chance, bias or small numbers. |
| Preclinical only | Studied only in cells or animals. | Nobody knows whether it works, at what dose, or how safe it is in people. |
Why animal results often fail in people
Most drugs that look promising in animals never make it to approval. Reviews of drug development put the overall success rate from first human testing to approval at roughly one in ten, and lower in some fields. Rodents clear drugs faster, heal faster and are often given doses far larger than a person would receive. Many studies use small groups of genetically identical animals and are run by a single lab.
So when a page says a peptide "healed tendons in rats", read it as "worth testing in people", not "works in people".
What the trial phases mean
- Phase 1: a small group, often healthy volunteers. The question is safety and how the body handles the drug.
- Phase 2: a few dozen to a few hundred patients. Early signs of benefit and the best dose range.
- Phase 3: hundreds to thousands of patients, usually compared against placebo or a standard treatment. This is the evidence regulators rely on.
- Phase 4: monitoring after approval, where rarer side effects often show up.
Questions to ask about any study
- Who was studied? Cells, animals or people? How many?
- Was there a comparison group? Without a placebo or control, improvement may be natural recovery or expectation.
- Was it randomized and blinded? If people knew what they got, results can be biased.
- What was measured? A lab marker moving is not the same as people feeling or living better.
- Who paid for it, and has anyone repeated it? A single lab's results, repeated only by that lab, deserve caution.
- Was it published in a peer-reviewed journal? Conference posters, preprints and company press releases are a starting point, not a conclusion.
How we write each profile
- Animal and cell results are labeled as such, every time.
- Doses appear only if they come from a published human study or an official label. If none exist we say so, rather than repeat numbers circulating online.
- Every factual claim links to a numbered source at the bottom of the page.
- Legal status is checked against current FDA, compounding and WADA lists, with the review date shown.
More in our editorial policy.