A joint modelling approach for longitudinal studies

• 成果类型: Article
• 署名作者: Zhang, Weiping; Leng, Chenlei; Tang, Cheng Yong
• 署名单位: Chinese Academy of Sciences; University of Science & Technology of China, CAS; University of Warwick; National University of Singapore; University of Colorado System; University of Colorado Denver; Children's Hospital Colorado; University of Colorado Anschutz Medical Campus
• 刊物名称: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
• ISSN/ISSBN: 1369-7412
• DOI: 10.1111/rssb.12065
• 发表日期: 2015
• 页码: 219-238
• 关键词: GENERALIZED ESTIMATING EQUATIONS; maximum-likelihood-estimation; semiparametric estimation; COVARIANCE-STRUCTURES; linear-models; selection; matrix

摘要

In longitudinal studies, it is of fundamental importance to understand the dynamics in the mean function, variance function and correlations of the repeated or clustered measurements. For modelling the covariance structure, Cholesky-type decomposition-based approaches have been demonstrated to be effective. However, parsimonious approaches for directly revealing the correlation structure between longitudinal measurements remain less well explored, and existing joint modelling approaches may encounter difficulty in interpreting the covariation structure. We propose a novel joint mean-variance correlation modelling approach for longitudinal studies. By applying hyperspherical co-ordinates, we obtain an unconstrained parameterization for the correlation matrix that automatically guarantees its positive definiteness, and we develop a regression approach to model the correlation matrix of the longitudinal measurements by exploiting the parameterization. The modelling framework proposed is parsimonious, interpretable and flexible for analysing longitudinal data. Extensive data examples and simulations support the effectiveness of the approach proposed.