At Natural Compare, we take a rigorous, science-first approach to evaluating natural compounds. Our methodology is designed to provide you with accurate, unbiased information that helps you make informed decisions about your health.
Evidence Hierarchy
We evaluate research based on a hierarchy of evidence, with higher-quality study designs weighted more heavily:
- Meta-Analyses & Systematic Reviews – Combine results from multiple studies to provide the most comprehensive view of the evidence.
- Randomized Controlled Trials (RCTs) – The gold standard for determining cause and effect.
- Cohort & Observational Studies – Follow groups over time to identify associations. Useful but cannot prove causation.
- Animal & In Vitro Studies – Provide preliminary evidence and mechanistic insights. Results may not translate to humans.
- Traditional Use & Case Reports – Historical use and individual cases. Valuable context but lowest level of evidence.
Evidence Rating System
Each compound receives an evidence score from 1-5 based on the quality and consistency of available research:
- 5 – Strong Evidence: Multiple meta-analyses or large RCTs with consistent results
- 4 – Good Evidence: At least one meta-analysis or multiple RCTs showing benefit
- 3 – Moderate Evidence: Some RCTs with mixed or modest results
- 2 – Preliminary Evidence: Limited human studies, mainly animal or in vitro research
- 1 – Insufficient Evidence: Traditional use only or very limited research
Product Evaluation Criteria
When evaluating supplement products, we consider:
- Third-Party Testing: Independent verification of purity, potency, and absence of contaminants
- Bioavailability: Whether the form used is well-absorbed by the body
- Clinical Dosing: Whether the dose matches what was used in clinical studies
- Value: Cost per serving relative to quality
Limitations
- We are educators, not healthcare providers. Our information is for educational purposes only.
- Scientific evidence evolves. We update our content regularly.
- Individual responses vary. What works for one person may not work for another.