A Copula-based Fully Bayesian Nonparametric Evaluation of Cardiovascular Risk Markers in the Mexico City Diabetes Study.

Many studies have been carried out to understand and explore cardiovascular risk markers in normoglycemic and diabetic populations. In this study we model the association structure between hyperglycemic markers and cardiovascular risk controlled by triglycerides, body mass index, age, and gender, for the population in The Mexico City Diabetes Study. The existence of an association structure could contribute to the proper assessment of cardiovascular risk markers in this low income urban population with a high prevalence of classic cardiovascular risk biomarkers. The association structure is measured by conditional Kendall's tau, defined through conditional copula functions. The latter are in turn modeled under a fully Bayesian nonparametric approach, which allows the complete shape of copula to vary for different values of the controlled variables.