What's in a risk factor? "He who strikes the ball"

Abstract

A major aim of epidemiology is to explore the prognosis and aetiology of diseases, in order to improve treatment and prevention. To avoid misleading interpretations of observed associations between a modifier and the occurrence of a disease, the all-purpose idiom "risk factor" should be replaced by 3 locutions with narrower meanings: a risk marker is a modifier which is associated (correlated) with disease prevalence (case-control studies) or incidence (cohort studies); a risk marker truly becomes a risk factor if its experimental correction (intervention studies) does improve the disease incidence or prognosis; a risk factor is promoted to the rank of cause if it is proved to be necessary (sine qua non) for the occurrence of the disease. A more rigorous vocabulary avoids premature claims for unsubstantiated treatments or preventions, and helps defining sound priorities in aetiological research.