How to read a study
Every peptide page on Learn tags each finding as Cell, Animal or Human. Here's why that label matters more than the finding itself.
The evidence ladder
Cells in a dish. Fast and cheap. Shows whether a molecule can do something to cells under controlled conditions. It can't show what happens in a whole body, where the molecule has to survive digestion or the bloodstream, reach the right tissue, and not cause problems elsewhere.
Mice and rats, usually. A whole living body, so more meaningful. But rodents aren't small people: their metabolism, doses and lifespans are very different, and many results don't carry over. Most research peptides have only reached this step.
People. Ranges from a handful of case reports to huge trials. Within human research, quality varies enormously (see below).
Not all human studies are equal
| Case report or chart review | Looks back at what happened to a few patients. No comparison group, so it can't separate the substance from coincidence, other treatment or natural recovery. |
| Small pilot study | Tests a few people, mainly for safety. Too small to say much about whether something works. |
| Randomized controlled trial | People are randomly split into the substance and a placebo, often without knowing which they got. This is the strongest kind of single study. |
| Large phase 3 trials | Several big controlled trials. This is the level of evidence that leads to FDA approval. |
Five questions to ask about any claim
- Cells, animals or people? If it doesn't say, assume cells or animals.
- How many? 8 rats and 8,000 people are not the same evidence.
- Was there a comparison group? Without one, you can't know what caused the result.
- Who ran it? Results repeated by independent groups count for more than one lab's.
- Is there a source? A real claim links to a study you can look up on PubMed. "Studies show" with no link is marketing.
Next
Next: where peptides stand with the FDA and in sport, in plain English: Where peptides stand legally.