This appendix sets out how PBTWeb’s research teams grade the evidence behind a product and which sources they accept. It exists so that when a review says an ingredient is “well supported” or “traditional use only”, you know exactly what that means and can check the reasoning yourself.
I. The Evidence Grades
| Grade | Type of evidence | What it can and cannot tell you |
|---|---|---|
| Grade A | Systematic reviews and meta-analyses of randomised trials | Pooled results from multiple controlled human studies. The strongest support an ingredient can have. |
| Grade B | Individual randomised controlled trials | A single well-run human trial with a placebo group. Strong, but one trial can be wrong; replication matters. |
| Grade C | Observational and cohort studies | Associations in populations. Useful for dietary patterns; cannot prove a supplement caused an effect. |
| Grade D | Small, open-label or pilot human studies | Early human data without a control group. Suggestive at best. |
| Grade E | Animal and laboratory (in vitro) studies | Mechanistic findings in cells or animals. Explain how something might work; do not show that it does in people. |
| Grade F | Traditional use | Documented historical use in a medical tradition. Evidence of safety at traditional doses and of plausibility, not of efficacy. |
A review states the highest grade available for the specific ingredient, form and dose in the product. A grade-A meta-analysis of a 500 mg extract does not support a product containing an undisclosed amount of that extract inside a 150 mg blend.
II. Three Questions Applied to Every Study
- Same ingredient? Species, plant part, extract ratio and standardisation must match the product.
- Same dose? The study dose is compared with the label dose. When the label hides the dose, the comparison is impossible and the review says so.
- Same people? A trial in patients with a diagnosed condition does not describe what happens in healthy adults, and vice versa.
III. Sources We Accept
- PubMed and the peer-reviewed journals it indexes
- NIH Office of Dietary Supplements fact sheets
- National Center for Complementary and Integrative Health
- FDA dietary supplement guidance and warning-letter database
- Cochrane Library systematic reviews
- Professional bodies such as the American Heart Association, American Lung Association and American Academy of Dermatology
IV. Sources We Do Not Accept as Evidence
- Manufacturer websites and sales pages (used only to record claims, never to support them)
- Other review sites and affiliate blogs
- Testimonials, influencer content and before/after photos
- Press releases without an accompanying published study
- Studies funded by the seller unless independently replicated
V. How Grades Appear in Reviews
Each review states the grade for the key ingredient in plain words — for example “grade B: one randomised trial at a higher dose than this label provides” — and links to the source. Where the only support is grade E or F, the review says the product is plausible but unproven, and the verdict is written accordingly.