One supplement, more than one grade
An EBL grade does not rate a supplement. It rates the strength of the human evidence that a supplement helps with one specific thing. That is why the same supplement can carry more than one grade on its page.
Most supplements are used for more than one purpose, and the evidence behind those purposes is rarely equal. A supplement can have strong clinical trial evidence for one use and weak, or even failed, evidence for another. If we collapsed all of that into a single letter, we would have to do one of two things. Round up, which oversells the weak uses. Or round down, which hides the strong ones. Both are marketing. Neither is honest.
So we grade each use on its own evidence, using the same scale every time. The scale never moves. What moves is the question being asked.
How good and how replicated The same finding reproduced by a separate research group. A result seen once can be chance or bias; a replicated one is harder to explain away. the human trials are. A is multiple replicated RCTs Randomized controlled trial: people are assigned to the treatment or a comparison by chance, so the groups differ only in what they received. It is the design that can show cause and effect. with a consistent signal. B is a strong but more limited RCT base. C is small, mixed, or failed-to-replicate data. D is mechanism or animal only.
An A is reserved for outcomes that are real and hard to game. Softer measures are capped lower no matter how clean the trials are.
Hard clinical means mortality, heart attacks, fractures, disease. Objective functional means a measured outcome that is the goal itself, like measured strength or VO2max The most oxygen the body can use during all-out exercise, a standard measure of aerobic fitness.. Blinded clinician-rated means a validated scale scored by a clinician who does not know which treatment the patient received, like the Hamilton depression scale. Surrogate biomarker is an objective stand-in like LDL Low-density lipoprotein cholesterol, the cholesterol fraction most closely tied to plaque in the arteries. or HbA1c A blood test that reflects average blood sugar over the previous two to three months.. Self-report is subjective, like a questionnaire. Only the first three can earn an A.
A worked example: vitamin D
Vitamin D is the clearest case. It works when you are deficient and adds little when you are not. Same molecule, two very different answers, because they are two different questions resting on two different bodies of evidence.
Hip-fracture and fall reduction in deficient or older populations. Graded B rather than A because the benefit does not extend to already-replete adults, so the signal is not consistent across populations.
The large VITAL trial was null for invasive cancer and cardiovascular events, and its bone arm was null for fractures in replete adults. Real endpoints The outcome a trial measures to decide whether the treatment worked, such as heart attacks, blood pressure or a symptom score., replication failed.
On each supplement page you will see a separate grade for each use we have reviewed, and beside every grade the actual human trials it stands on. A high grade for one use is not a green light for all of them. Read the grade that matches what you are actually trying to do.
This is the part most supplement information gets wrong. A product is sold as one item, so it gets discussed as if it carries one level of proof. Evidence does not work that way. Grading each use on its own is more work and it is less tidy, but it is the only way the grade can mean what it says.
A grade reflects the evidence for one specific use. The same supplement can score high for a use with strong trials and low for a use where the trials are weak or negative. The grade that counts is the one that matches the goal.