Photo: Unknown photographer / Wikimedia Commons, Public domain

In vitro, animal or human? How to read peptide research claims

Updated 14 Oct 20262 min read

In short

Identify the study type first. Cell and animal studies show what a molecule can do in a model; only controlled human studies test effects in people, and then only for the product and population studied. Check also that the research is about the same molecule that is being sold.

Supplier pages and social media often summarise peptide research in a single confident sentence. The studies behind those sentences vary enormously. Some are experiments on cells in a dish, some on mice, and a few on people, from small observations to controlled trials. Knowing which kind of study you are looking at is the single most useful skill for reading peptide research. This guide explains the levels and how this atlas labels them.

The levels of evidence

Study type	What it can show	What it cannot show
In vitro (cells, tissues, enzymes)	That a molecule can affect a biological process under controlled conditions	Whether that happens in a whole organism, at realistic concentrations
Animal studies	Effects in a living system, including distribution, metabolism and toxicity signals	Whether results transfer to humans: species differ, and models simplify disease
Human observational	Associations, for example between natural levels of a molecule and a condition	Cause and effect, or the effect of giving the molecule
Human controlled trials	Effects of a defined intervention against a comparison group	Effects outside the population, product and conditions studied
Reviews	A synthesis of existing studies	Anything the underlying studies did not show; quality depends on what was included

Three common errors

Mechanism treated as outcome

Gene-expression and cell-signalling studies show that a molecule can influence a pathway. They do not show that it produces a measurable result in a person. The GHK-Cu literature, for example, includes extensive gene-expression analysis. It is valuable research, but it is mechanistic.

The wrong molecule

Research cited for one product sometimes concerns a related but different molecule. Our library flags two cases. Much TB-500 marketing cites research on full-length thymosin beta-4. And the human study usually cited for “CJC-1295” used the long-acting DAC form, not the “no DAC” peptide many vendors sell.

One laboratory’s results treated as consensus

When most studies on a compound come from one group, independent replication has not yet happened. That is a feature of the field’s maturity, not proof of a problem. It is still a reason for caution.

How this atlas labels evidence

Every reference on this site carries a type label: cell / in vitro, animal study, human study, review, official / regulatory, chemical database, analytical method, supplier document or general reference. Library entries also list the types of evidence that exist for each compound, and a “known limitations” section. A guide’s “evidence mix” shows how its references break down by type.

What this means for research-product listings

A research-use-only product is sold for laboratory work. Evidence from animal or cell studies is the normal starting point for further laboratory research. It is not a basis for any use in people, and this atlas never presents it as one.

5 sources