Methodology

How we source

HealthspanLedger sources approved doses from FDA labels and study doses from published human trials. When neither exists, we say so. Every figure stays linked to its document and the day we read it. A missing source stops the site from building.

What each source label means

Five plain-language labels, strongest first. You will see them beside the numbers they support.

FDA label

It is on an FDA-approved label. We quote what the label says.

Human trial

It comes from a published trial in people. We say what the trial gave people — not what anyone should take.

limited human data

There is some human data, but not much: a small study, or an approval from outside the US. We name the limit.

indirect evidence

It circulates through clinic protocols or animal studies. We say whose number it is and give it as a range, never a single figure.

no human dose shown to work

No study has established a dose. We say that, instead of printing a number.

The rules we hold ourselves to

Numbers come from the source itself — FDA labels, DailyMed, peer-reviewed journals, ClinicalTrials.gov, USP — never from a summary of one.
How we phrase it matches what we have. A label is quoted, a trial is described, a clinic protocol is attributed to whoever uses it. Nothing here is a recommendation.
Reconstitution arithmetic is just arithmetic, so we state it plainly. The dose you put into it is not, so that always carries its source.
Where no study has established a dose, we say so rather than making one up.
We do not report the result of a regulatory decision that has not been made yet.

Where you can see these rules kept

A dose with no trial behind it — the quantity that circulates, printed as exactly that, with what no study has tested beside it.
A figure quoted from a federal database — FDA side-effect reports, with the day they were read and what a count of reports can and cannot mean.
Who reads this before it ships — and what a review does not cover.