What "evidence-graded"
actually means here.
Every recommendation in the stack check and every guide on this site carries one of three grades. Here is exactly how they're assigned — so you can disagree with us intelligently.
Replicated human evidence
Multiple randomized controlled trials or meta-analyses in humans, with consistent direction of effect. Example: vitamin D for correcting measured deficiency.
Plausible, partially demonstrated
Mechanistic evidence plus limited or mixed human trials. Real effect likely in some contexts, oversold in marketing. Example: zinc lozenges for cold duration.
Mostly marketing
Animal or in-vitro data only, or human trials that failed to replicate. We'll say so plainly. Example: collagen for skin in well-nourished adults.
Sources we grade against
Reference values (RDA, UL) follow the NIH Office of Dietary Supplements and EFSA. Effect claims are graded against peer-reviewed trials and Cochrane reviews, prioritized by study quality — never by what sells. Retail links may earn a commission, but no brand can pay for a grade, a placement, or a recommendation.
Review process
Guides are drafted against this framework, fact-checked against primary sources, and reviewed by a registered dietitian before publishing. Each guide shows its last-updated date; reference values are re-checked when source guidelines change.