What the grades mean
Every claim on this site carries one. If it doesn't, it's a bug — the build refuses ungraded claims.
AMeta-analysis or large RCT
the strongest evidence there is
BSmall RCT
real trial, limited size or duration — both stated in the text
CObservational
association, not causation
DPreclinical
animal, organoid, or in-vitro — promising is not proven
EPreprint
not yet peer-reviewed
FExpert opinion or anecdote
worth hearing, not evidence