A cell finding, an animal experiment and a human trial answer different questions. Eight things to check before deciding what a result means, with examples from this catalogue.
Most claims about research peptides trace back to a study somewhere. The study is usually real. What is often missing is the context: what was tested, in what, by which route, against what, measuring what, and whether anyone repeated it. This guide walks through those checks using compounds from the catalogue, so that a reader can place any result at its correct evidence stage.
The first question is what kind of experiment it was. A cell study puts a compound onto cultured cells and measures what changes; it can show a mechanism is possible, not that it happens in a body. An animal study shows what happens in a living organism, usually a mouse or rat, under conditions the experimenters chose. A human trial shows what happens in people, but only in the people enrolled, by the route used, for the time measured.
The catalogue contains a clear example of the gap. AOD-9604, a fragment of growth hormone, reduced body weight and fat in obese mice (Heffernan et al. 2001). Six randomised human trials followed; the largest, a 24-week oral study in about 500 adults with obesity, found no weight-loss effect beyond diet and exercise, and development stopped in 2007 (Moré and Kenley 2014). Both results are true. The mouse result did not predict the human one.
Ask what was given, how, and to whom. Formulation matters because a pharmaceutical product with known potency and sterility is not the same as research material, and a lyophilised powder is not the same as a branded solution. Route matters because a peptide applied to a wound, inhaled, infused into a vein or injected under the skin reaches different tissues at different concentrations; GHK-Cu’s human evidence is a gel on ulcers, which does not describe any other route. Population matters because a result in patients with a defined disease describes those patients. SS-31’s 2025 US approval is for Barth syndrome, in people above a weight threshold, on a surrogate endpoint, with a confirmatory trial still owed; the phase 3 trial in the broader mitochondrial myopathy population missed its primary endpoints (Karaa et al. 2023). Narrow is the accurate word.
A result means little without knowing what it was compared against. Placebo control and blinding exist because expectation, regression to the mean and the passage of time all produce change on their own. Several Russian clinical studies of Selank and Semax report effects, but their abstracts describe no randomisation, blinding or placebo, and that design, not the language of publication, is what limits them.
Then ask what was measured. A growth hormone pulse in a volunteer study is a hormone measurement, not a body-composition outcome. A rise in plasma BDNF is a blood test, not a cognitive result. Kisspeptin’s volunteer studies measure luteinising hormone, and its IVF study counted mature eggs and pregnancies; neither establishes a registered treatment. The endpoint that was measured is the only thing a study can support.
One study is a starting point. DSIP shows why: two small intravenous trials in chronic insomnia, one of 14 patients reporting improved sleep (Schneider-Helmert 1987) and one of 16 patients whose authors judged the effect weak and unlikely to matter (Bes et al. 1992). Neither has been replicated in decades. Two small trials with opposite conclusions leave a question open, not answered.
Size is the related check. Walum and colleagues showed that the typical intranasal oxytocin study was too small to detect the effects it claimed, leaving the field with an inflated share of false positives (Walum et al. 2016). A three-person pilot, as with Melanotan II, and a case report support different kinds of claim again: the first that something was observed, the second that something should be investigated.
Finally, ask who ran and funded the study. This is not an accusation. It is a reason to look for independent replication. Much of the BPC-157 literature comes from one affiliated group in Zagreb; much of the GHK-Cu review literature comes from the peptide’s discoverer, who holds commercial interests; Epitalon and Pinealon come almost entirely from one Russian programme; several MOTS-c and SLU-PP-332 authors declare interests in companies developing analogues. Sponsor-run trials produced the strongest evidence in the catalogue, for tirzepatide, retatrutide and cagrilintide, and they were large, randomised and published in full. Funding is a flag for scrutiny, not a verdict.
A reader who checks stage, formulation, route, population, comparator, endpoint, replication and source will not need to be told what a result means. They will be able to say it themselves.
Sources checked 2026-09-22