Source Standards
This page sets out what we count as a source, how we rank sources against each other, and what we will not cite as evidence.
The hierarchy
Higher tiers outrank lower ones where they disagree.
Tier 1 — preferred
- Cochrane systematic reviews
- Systematic reviews and meta-analyses in indexed journals
- Indian Council of Medical Research (ICMR) publications
- AIIMS and National Institute of Nutrition (NIN) publications
- World Health Organization guidance
- Ministry of AYUSH official research publications
- FSSAI regulations and guidance
Tier 2
- Randomised controlled trials in indexed journals
- Government health portals — National Health Portal, Ministry of Health and Family Welfare
- Peer-reviewed observational studies
Tier 3 — context only, never clinical evidence
- Classical texts: Charaka Samhita, Sushruta Samhita, Ashtanga Hridaya
- Practitioner interviews, attributed and credentialed
Why Tier 3 is separated so firmly
Classical texts tell us what a tradition holds and how a practice has been used for centuries. That is real knowledge and we cite it as such.
It is not clinical evidence of efficacy. Conflating the two — citing a classical text as though it demonstrated that something works — is the central epistemic failure of this entire content category, and it is how a great deal of misleading health content gets an air of authority.
So in every sources block, classical citations carry a distinct visual marker from research citations. You can always tell at a glance which kind of claim you are being shown.
What we do not cite as evidence
- Other content-mill health websites, including similarly named ones
- Supplement and nutraceutical brand marketing
- Social media claims and influencer content
- Anecdote and testimonial
- Wikipedia — useful as a gateway to primary sources, not a source itself
- AI-generated summaries
How we cite
Every article carries at least three sources. Each citation shows a type indicator — systematic review, RCT, observational, government, regulation, classical text — so you can weigh it without clicking through.
Where a source sits behind a paywall we say so. Where we are relying on an abstract rather than a full text, we say that too.
When sources conflict
We say they conflict. A body of evidence that points in two directions is a real and common finding, and flattening it into a confident answer would be the more dishonest choice. In those cases the evidence label will usually be “some evidence, limited studies” and the article will explain the disagreement.
Broken links
Links rot. Annual re-review includes repairing broken sources. If you find a dead or wrong citation, tell us at [email protected].