Ask three questions of any peptide claim: was it tested in people, was there a randomized control group, and did it measure something you would notice, like pain or strength, rather than a blood level. Most claims made online for research peptides fail the first question.
Higher rungs answer whether something works in people. Lower rungs suggest why it might. Both are useful, as long as nobody quotes a lower rung as if it were a higher one.
A randomized trial is the only design that can separate the drug from everything else that changes over time: healing on its own, the placebo effect, and the tendency of bad weeks to be followed by better ones.
Everything below the dashed line is a reason to run a human study. None of it shows that a peptide works or is safe in a person, and that is where most of the peptides sold online for healing, longevity and cognition sit.
Ineichen et al. 2024, PLOS Biology: 281 therapies from 122 systematic reviews. A 2006 JAMA study found a similar pattern: about a third of highly cited animal studies were later confirmed in human randomized trials.
Most of the attrition is not because the animal studies were faked. It is because rats are not people, the injury models are simplified, and small experiments overstate effects.
The same review found that when an animal result did get tested in people, the human result usually pointed the same direction. The problem is that most never get tested, which is exactly where most research peptides sit today.
6 mg/kg in rats x 6 / 37 = about 0.97 mg/kg
FDA's 2005 guidance converts an animal dose to a human-equivalent dose using body surface area: the rat factor is 6, the adult human factor is 37. So a rat dose is not a per-kilogram human dose; it is roughly a sixth of it. FDA then divides by a default safety factor of 10 to pick a first dose for a safety trial.
That math sets a cautious starting point for testing. It does not tell anyone what amount works, and it assumes the drug is handled the same way in both species. The numbers here are illustrative and do not refer to any peptide.
Peptide X accelerates healing of transected Achilles tendon in rats
Male Wistar rats (n = 8 per group) underwent surgical transection of the Achilles tendon. Animals received peptide X or saline by daily injection for 14 days. Tendons treated with peptide X showed higher load to failure and improved histological scores compared with controls (p < 0.05). These results suggest peptide X may be a potential therapy for tendon injury.
Each claim below is one you will find on vendor sites, forums or social media. Pick the best evidence behind it, then see what the studies actually show.
"Semaglutide lowers the risk of heart attack and stroke."
What is the best evidence behind this claim?
Each morning: the studies that matter, what species they were in, how big they were, and what they measured. Sources linked.