CMStatistics 2017: Start Registration
View Submission - CMStatistics
B0359
Title: Bivariate copula regression models in cardiovascular disease Authors:  Jenifer Espasandin-Dominguez - University of Santiago de Compostela (Spain)
Pedro Oliveira - EPIUnit-ICBAS-Universidade do Porto (Portugal) [presenting]
Isabel Vila - Centro Hospitalar do Alto Ave (Portugal)
Jorge Cotter - University of Minho (Portugal)
Pedro Cunha - University of Minho (Portugal)
Carmen Cadarso Suarez - Universidad de Santiago de Compostela (Spain)
Abstract: Portuguese population presents a high incidence of stroke that seems to be linked to hypertension, salt consumption, excess weight and other cardiovascular risks. Early vascular aging (EVA) is defined as an accelerated aging process of the blood vessels by which arteries present the characteristics of older chronological ages. It is envisaged that the effects of accumulated cardiovascular risks in vascular aging are reflected in Pulse Wave Velocity and Central Blood Pressure. The data consists of a random sample of dwellers, aged 18-96 years, in Northern Portugal. With the aim of studying the relationship between pulse wave velocity and central blood pressure, depending on gender, age, waist circunference, salt consumption and other clinical covariables, bivariate copula additive models for location, scale and shape for continuous responses will be used. This type of model extends the use of generalized additive models for location, scale and shape to situations in which two responses are modeled simultaneously conditional on some covariables using copulae. The approach permits the two responses and copula parameter to be modelled using additive predictors that allow for several types of covariate effects. The models can be easily used via the SemiParBIVProbit() function in the R package SemiParBIVProbit.