How we grade evidence.
Every benefit on this site carries one of five grades, from Moderate to None. The grade is our reading of published human trials on the ingredient. It says how much proof exists, not how well something will work for you.
What each
grade means.
| Moderate | Several controlled human trials point the same way. Not proven beyond doubt, but more than a hint. |
|---|---|
| Some | A few controlled trials, usually small, with broadly consistent results. |
| Early | One or two small or preliminary human studies. Interesting, not reliable. |
| Traditional | Long use in Ayurveda or other traditions, with little or no controlled human research. |
| None | A popular claim for which we found no human evidence worth quoting. |
We do not use a "Strong" grade. For the ingredients we cover, we have not found evidence that deserves it.
The rules
we follow.
Human trials first. A benefit is graded on controlled studies in people. Test-tube and animal results do not earn a grade.
Size and funding are stated. Where a trial was small, short or funded by an ingredient maker, we say so.
Retractions count. If a study is retracted, we stop relying on it and say that it was withdrawn.
Sources are linked. Each guide lists the studies and public health pages it draws on, so you can read them yourself.
No disease claims. We describe what was studied. We do not say any ingredient treats, cures or prevents a disease.
About
the guides.
Who writes Skeptics guides?
The Skeptics editorial team writes them from published human research and public health sources. We are not doctors and the guides are not medical advice.
What counts as evidence?
Controlled trials in people, and reviews that pool them. Laboratory and animal studies are mentioned only to explain a caution, never to support a benefit.
Do the grades apply to Skeptics products?
No. A grade describes research on the ingredient in general. It is not a clinical test of any particular product, ours included.
How often are guides reviewed?
Each guide carries the date it was last reviewed. We update a guide when a significant new trial is published, or when a study we relied on is corrected or retracted.
What if you get something wrong?
Tell us and we will check it. If we were wrong we will correct the page and change the review date.