Vitamin D and heart health are linked in a large body of research. The first question is whether a paper describes an association, a laboratory mechanism or the results of giving people a supplement. Those forms of evidence can help each other, but they do not carry the same message. A lower blood level in people with heart disease does not, by itself, show that raising it prevents a heart attack.

Follow the evidence to actual events

VITAL did not find a significant reduction in its main cardiovascular outcome with vitamin D compared with placebo. The trial counted events rather than assuming that a change in a blood marker would translate into a clinical benefit. Results from a generally healthy prevention population should be described that way. They do not automatically settle treatment of a deficiency or every question about established cardiovascular disease. Equally, those distinctions should not be used to dismiss a substantial null result.

Be precise about smaller findings

Blood pressure, cholesterol and measures of vessel function can be useful research outcomes. If one changes, the article should say so without substituting the phrase ‘prevents heart disease’. That larger claim requires evidence at the level of disease outcomes. The biological and observational interest has not translated into a general cardiovascular-prevention benefit in VITAL. Heart-health coverage should explain the difference between a plausible mechanism and a demonstrated reduction in events. Readers should not have to discover that distinction in the small print.