THE SUPPLEMENT EVIDENCE REGISTER

Every claim graded against the studies behind it.

Grades are per claim, never per compound — creatine for muscle is not creatine for cognition. Mouse results never dress up as human results.

press enter to search the register · 20 compounds · 48 claims · 40 studies
HOW GRADING WORKS — TWO AXES, NEVER COLLAPSED
A
Multiple human RCTs, consistent direction, independently replicated, hard endpoints.
B
At least one well-conducted human RCT, or consistent human observational data with a plausible mechanism.
C
Human data only, but small, short, mixed, or limited to surrogate biomarkers.
D
Animal or in-vitro only. No human evidence for this claim.
E
Tested in humans and found not to work, or contradicted by the better studies.
The grade says how much we know. Effect size — large / moderate / small / negligible — is shown separately: a compound can be grade A with an effect too small to matter. That combination is one of the most useful things this register can tell you.