Editorial Policy
Last updated: 2026-03-04
Evidence Grading System
Every ingredient-condition claim receives a letter grade (A through F) based on the totality of available peer-reviewed evidence.
A
Strong Evidence — Multiple high-quality RCTs with consistent positive results.
B
Moderate Evidence — At least one well-designed RCT with positive results.
C
Limited Evidence — Preliminary positive findings from small studies.
D
Weak Evidence — Only observational or preclinical data.
F
No/Negative Evidence — No credible evidence or potential harm.
Data Sources
- PubMed / MEDLINE (National Library of Medicine)
- PubMed Central — Full-text articles under CC-BY license
- Semantic Scholar (Allen Institute for AI)
- ClinicalTrials.gov
Review Process
- Rechercher — Systematic PubMed search for ingredient + condition
- Screen — Relevance filtering by study type and population
- Extract — PICO data extraction from each included study
- Grade — Algorithmic grading based on study quality, consistency, and sample size
- Review — Human review of algorithmic output before publication