Editorial Policy

Sourcing standards

• Prefer primary sources (PMID/DOI, trial registries, regulatory documents).
• Preprints are allowed but must be clearly labeled as not peer reviewed.
• Every substantive claim should be backed by at least one citation.

Evidence labeling

We label content by study type and (when possible) a simple confidence tier.

• Study type: in vitro, animal, observational, RCT, systematic review, meta-analysis.
• Confidence tier: low, moderate, high (based on design, consistency, and limitations).

Claims and language

• We avoid overstating animal or in vitro findings as human outcomes.
• We separate study results from interpretation and call out limitations.
• We prioritize effect sizes and endpoints over hype.

Corrections

• Substantive edits should be recorded in a changelog.
• If you find an error, submit a fix via the repo workflow or the contact page.