Every page follows the same four steps.
1. Research
We start with primary and authoritative sources: government health agencies, clinical practice guidelines from major medical organizations, systematic reviews and meta-analyses, and individual randomized trials where they matter. We read the source, not a summary of it, and we record it in a central source registry so that the same study is cited the same way everywhere.
2. Review
An editor reads the finished draft against its sources. Anything that cannot be traced to a source is removed. Anything that overstates the evidence is rewritten. The editor also checks that the article answers the question a reader actually came with, that technical terms are explained, and that the reading level is appropriate.
3. Verify
For supplement products, the Supplement Research Editor checks the ingredient list, doses, price, guarantee and manufacturer claims against the product label and the seller's own website. If something cannot be confirmed, the page says "not independently verified". A product page is not published until the formula, manufacturer and guarantee have been confirmed.
4. Update
Every article shows its last review date. Pages are re-checked on a rolling schedule, sooner when guidelines or product formulas change. Readers can report an error through the contact page; see our corrections policy.
How we rate ingredient evidence
An ingredient does not have one evidence score. It has one for each use it is studied for, because the evidence for berberine and blood glucose is not the same as the evidence for berberine and weight. Each rating reflects human research only:
| Rating | Meaning |
|---|---|
| Strong | Consistent benefit across multiple well-designed randomized trials or a robust meta-analysis |
| Moderate | Benefit in several trials or a meta-analysis, with some limitations in size, duration or quality |
| Limited | A small number of trials, small samples, or short durations; results may not be reliable |
| Mixed | Trials disagree, or benefit appears in some outcomes but not others |
| Insufficient | Little or no human research; evidence is mainly laboratory or animal studies |
How we look at finished products
A product is not its ingredients. Even when individual ingredients have good evidence, the finished formula may use different doses, different forms, or a proprietary blend that hides the amounts. On every product page we state which of these applies:
- Clinical evidence exists for the finished product — rare, and we link the study.
- Evidence exists for individual ingredients only — the common case. We say which ingredients, at what doses, and whether the product's doses match.
- No relevant evidence found — we say so.
We separate manufacturer claims from established medical evidence, and we do not present one as the other.