New Direction: Mediation Analysis in Perinatal Epidemiological Studies

Abstract

In traditional epidemiology, a multivariable regression model is often applied to evaluate the relationship between an exposure and an outcome while adjusting for confounders and taking into account of interactions. In 1992, Robins and Greenland introduced the mediator counterfactual framework concept, which has helped researchers disentangle the complex causal pathway between exposure and outcome. In the past one decade, mediation analysis has been used in various epidemiological studies, including reproductive and perinatal research. This analysis method has been incorporated into various statistical software packages, including the CAUSALMED procedure in SAS, among various others. Mediation analysis is not only used to explain causal relationships between exposure, mediator, interaction and outcome but also provide solid evidence for decision makers to implement policy.