Supplement Research & Evidence Reviews

How FactoWiki evaluates evidence

Sources prioritised

• PubMed — indexed human studies, located and cited by record.
• NIH, including the Office of Dietary Supplements, and the National Eye Institute and other institutes where relevant.
• NCCIH — the National Center for Complementary and Integrative Health.
• FDA — safety communications, monographs and regulatory decisions.
• Systematic reviews, including Cochrane reviews.
• Meta-analyses of randomised trials.
• Randomised controlled trials, especially large, independent and placebo-controlled ones.

Hierarchy of evidence

Human evidence > animal evidence > mechanistic evidence. Human trials measuring real outcomes (symptoms, blood markers, clinical events) carry the most weight. Animal studies can suggest what to test but do not show benefit in people. Laboratory and mechanistic studies explain how something might work, never whether it does. Within human evidence, large randomised placebo-controlled trials and well-conducted systematic reviews outrank small, open-label or observational studies.

Ingredient research ≠ finished-product proof

A trial tests a specific ingredient, form, dose and population. A commercial supplement usually combines several ingredients, often at undisclosed or different amounts, and is rarely tested itself. So each review keeps three questions separate: is there human evidence for the ingredient; at what dose; and has any finished product been tested. We do not describe products as “clinically proven” on the strength of ingredient research.

How disagreements are handled

Where studies or reviews disagree, we describe the disagreement and why it may exist — trial size, funding, extract type, population — rather than choosing the most positive result. Each page reports population, dose, duration, result, limitations and safety.

More on our process: Research Team · Editorial policy · How we review