Background: The aim of this paper is to analyse how epidemiology can contribute and assess interventions to reduce inequalities in health. This goal is best attained through an in-depth understanding of the construction of inequalities along the life course. Furthermore, it is concordant with a conception of causality stating that "the idea of causality has lost any other meaning than the pragmatic designation of the point in the causality chain where an intervention would be the most effective".
Discussion: In a cross-sectional approach, these chains of causality may be seen as Chinese boxes; in a longitudinal approach, they refer to life course epidemiology. Based on these models, a classification of interventions might be proposed: the approach by individual risk factors refers to the interventions toward vulnerable populations; the Chinese box approach leads to intervention directed toward non proximal factors, i.e., the environment, social factors and where the individual lives; the life course approach leads to intervention on fundamental causes. Understanding these mechanisms would thus lead to a better use and coordination of interventions to reduce health inequalities.