Our method

How we rate the evidence

Every grade on Vitaproof answers one question: how sure can we be that this supplement does this for these people? We use the four certainty levels of GRADE, the rating system used by many health agencies and guideline panels, including the vaccine advisory committee of the US CDC.

Two separate things: the evidence and the product

An evidence grade belongs to a claim: one supplement, one outcome, one group of people. A product fact belongs to a pack on a shelf: its dose, its price per day, where we found it. A cheap, well-made product cannot raise a supplement's grade, and no brand can pay to change either. Sponsored placements, when they come, will be labelled and will never change a grade, a description or a score.

The four grades

A

High certainty

In GRADE terms: “We are very confident that the true effect lies close to that of the estimate of the effect.” Usually several good randomised trials agree.

6 claims in our database
B

Moderate certainty

The true effect is probably close to what studies found, but it could be substantially different. Often good trials with some inconsistency, or a smaller body of evidence.

9 claims in our database
C

Low certainty

The true effect may be substantially different from what studies found. Typically small, short or mixed studies. Promising is not the same as proven, and we say so.

13 claims in our database
✕

Not supported

Either good studies found no meaningful benefit (for example vitamin D and cancer in the VITAL trial), or the evidence is so weak that the true effect is likely very different. The claim summary always says which.

10 claims in our database

How a grade is set

  1. One grade per claim. Vitamin D can be A for bone health in people with low levels and ✕ for preventing cancer. We never give a supplement one overall grade.
  2. We start from expert evidence summaries, such as the NIH Office of Dietary Supplements fact sheets, the NIH's National Center for Complementary and Integrative Health, Cochrane reviews and large trials. Every claim links to its source.
  3. Study design sets the starting point. In GRADE, evidence from randomised trials starts as high certainty and evidence from observational studies starts as low.
  4. Five things lower certainty: risk of bias, inconsistency between studies, indirectness (for example studies in a different group of people), imprecision and publication bias.
  5. Three things can raise it for non-randomised studies: a large effect, a dose-response relationship, and confounding that would have made the effect look smaller.

Honest about our limits: we have not run our own formal GRADE reviews. We apply the GRADE certainty levels to the conclusions of the sources we link. If a source calls the evidence weak, mixed or insufficient, the grade is C or lower. If we cannot verify a claim in a public source, it does not go into the database. We would rather leave a field empty than guess.

Safe amounts: two limits, shown side by side

  • EFSA tolerable upper intake level (UL): the European Food Safety Authority's maximum daily intake from all sources (food, fortified foods and supplements) that is unlikely to cause harm over a lifetime.
  • BfR recommended maximum for supplements: the German Federal Institute for Risk Assessment's suggested maximum per day in a food supplement, which takes into account what people already get from food. It is often much lower than the UL.

When a product's daily dose is above the BfR maximum, we note it next to the product.

Products: what we show today

  • Price per daily dose, calculated only when the price, the pack size and the daily dose are all stated on the source page. If one is missing or the dose is a range, we leave it empty.
  • The source page and the date we checked it, the PZN (German pharmacy product number), and notes such as “EPA/DHA content not stated on source”.

Product score (coming soon)

We are preparing a 0–100 product score. This is the weighting we plan to use; we will publish the full rules before the first score goes live.

  • Dose matches what the evidence used30
  • Form of the ingredient20
  • Price per daily dose20
  • Independent lab testing15
  • Clean formulation (no unnecessary extras or megadoses)15

Found a mistake?

Tell us at admin.vitaproof@gmail.com with a link to the source. We check it and correct the database.

Sources for this method

Database as of 2026-11-11: 21 supplements, 38 graded claims, 17 products. Try the supplement check.