Evidence tiers

The ladder from cell studies to animal work to human trials — a grading of how directly a piece of research bears on questions about people, and the reason a study count alone says little.

Not all published evidence carries the same weight, and the differences are structural rather than a matter of quality. In-vitro work — cells, tissues, biochemical assays — establishes mechanism under controlled conditions far removed from a whole organism. Animal studies add physiology, at the cost of species differences that translate imperfectly. Human trials range from small safety studies to large randomised controlled trials, and only the upper rungs can support claims about effects in people.

Position on the ladder is a property of a question, not a compound: the same molecule can have hundreds of mechanistic papers and no human data on the outcome anyone actually asks about. That is why a raw study count is a weak signal, and why the compound pages in this site's research library state each compound's evidence maturity and say explicitly when findings are preclinical only — including blends whose components were studied separately but never together.

The tier also bounds what a positive result means. A compound 'shown to' do something in cells has been shown to do it in cells; the phrase carries no implication further up the ladder, and reading it otherwise is the most common way research summaries mislead.

For research purposes only. Not for human consumption, diagnosis, treatment, or prevention of any condition. Nothing on this page is medical advice.