Estimating Correlation Among Multiple Longitudinal Outcomes of Disease Progression in Patients Who Developed Primary Open-Angle Glaucoma

• Activity Number: 231
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Feng Gao*+ and J. Philip Miller and Julia Beiser and Ling Chen and Mae Gordon
• Companies: Washington University School of Medicine in St. Louis and Washington University School of Medicine in St. Louis and Washington University School of Medicine in St. Louis and Washington University in St. Louis and Washington University School of Medicine in St. Louis
• Keywords: multivariate longitudinal data ; joint modeling ; latent class growth model ; linear mixed effect model

Abstract

Primary open-angle glaucoma (POAG) is one of the leading causes of blindness and characterized by a gradual loss of vision over time. The course of disease progression in POAG is often described using longitudinal studies, but most longitudinal models focus on characterizing change over time in a single outcome variable. However, the availability of multivariate longitudinal data provides a unique opportunity to understand the interrelationship among multiple outcomes and to gain insightful information on monitoring of disease progression. In this paper, using longitudinal data from 279 participants who developed POAG in the Ocular Hypertension Treatment Study (OHTS), we assessed the joint evolution among functional indices (mean deviation and pattern standard deviation), structural measurement (horizontal cup-disc ratio), and visual acuity. Using recent emerging techniques for multivariate longitudinal data, we assessed not only how the evolution of one response is correlated to the evolution of another response ("association of evolution"), but also how the correlation between outcomes changes over time ("evolution of association").