This page sets out how the articles here are written and the rules I use to decide how strongly a claim can be stated. It exists so you do not have to take my word for any of it — you can check whether I keep to my own rules.
Who writes this
I am Dr Cath. My doctorate is in chemistry. I am not a doctor and not a clinical dietitian, and I do not diagnose anything — I read the literature and explain what follows from it, and what does not. Where a question needs a clinician, I say so plainly instead of offering you a supplement.
Where the sources come from
The basis is literature indexed in PubMed. Every article carries a Sources section with clickable links to PubMed or DOI. If a source cannot be opened and checked, it has no business being here.
Three rules govern which papers make it in:
- A paper has to be about this compound and this population. A study on a different compound, or on animals, is not evidence for a claim about people — not even with a caveat in brackets.
- I do not select sources to fit a conclusion. Where the data are mixed, the article says so instead of pretending the question is settled.
- Two relevant papers beat five where three are from another field.
How I match claim strength to study strength
This is the core of the whole site. The same result can be described honestly or made to sound stronger than it is, and the difference is usually one verb. I hold to a fixed key:
- Meta-analysis of randomised trials → "shown to". The strongest level available to me.
- A single randomised trial → "in one study". One result is not yet settled knowledge.
- Observational study → "is associated with". Never "causes". Correlation is not cause.
- In vitro or animal work → the caveat goes in the same sentence as the finding, not a paragraph later.
- A biochemical mechanism on its own → a hypothesis, not a result. A mechanism says why something might work, not that it does.
The distinction that changes the most: biomarker vs. effect
A supplement can raise the concentration of something in your blood and change nothing you would notice. A rise in a biomarker is evidence that the compound was absorbed — not evidence that it worked. I keep those two apart in every article, because most supplement marketing depends on running them together.
What you will not find here
- Dosing schedules addressed to you. I describe what was dosed in a study and what the result was. "Take X mg for Y weeks" is clinical advice, and I do not give it.
- Promises to treat or prevent disease.
- Body-weight advice, or language that could feed disordered eating.
- Protocols for pregnancy, breastfeeding, children or chronic illness. For those groups I point you to a clinician.
Money — how it works
Some links here are affiliate links. If you buy through one I earn a commission and you pay exactly the same price. That is disclosed at the top of every article that contains one, before you reach the first link.
The rule I hold to: a product appears only when the article's own findings support it. Where a piece ends at "the data do not support this", there is no buy link, because there would be nothing behind it. An article with no affiliate link at all is a normal and acceptable outcome.
What I do when I get it wrong
I correct the article, and I do not pretend it always read that way. If you find an error — a figure that does not match its source, a citation that does not say what I attributed to it, or a claim stated more strongly than the study behind it — tell me. That is the most useful thing anyone can send me.
Disclaimer
Everything on Evidence Dose is educational. It does not replace medical, pharmaceutical or dietary advice, and it is not a basis for diagnosing or treating anything yourself.