A Bayesian finite mixture of bivariate regression model for causal mediation analyses.

Abstract

Building on the work of Schwartz et al, Joint Bayesian analysis of birthweight and censored gestational age using finite mixture models in Statistics in Medicine, we propose a Bayesian finite mixture of bivariate regression model for causal mediation analyses. Using an identifiability condition within each component of the mixture, we express the natural direct and indirect effects of the exposure on the outcome as functions of the component-specific regression coefficients. On the basis of simulated data, we examine the behavior of the model for estimating these effects in situations where the associations between exposure, mediator, and outcome are confounded or not. Additionally, we demonstrate that this mixture model can be used to account for heterogeneity arising through unmeasured binary or categorical mediator-outcome confounders. Considering gestational age as a potential mediator, we then illustrate our mediation mixture model to estimate the natural direct and indirect effects of exposure to inhaled corticosteroids during pregnancy on birthweight using a cohort of asthmatic women from the province of Quebec (Canada).