Latent Class and Genetic Stochastic Process Models: Implications for Analyses of Longitudinal Data on Aging, Health, and Longevity

• Activity Number: 648
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Konstantin Arbeev* and Liubov Arbeeva and Igor Akushevich and Alexander Kulminski and Svetlana Ukraintseva and Anatoliy Yashin
• Companies: Duke University and Duke University and Duke University and Duke University and Duke University and Duke University
• Keywords: stochastic process model ; mortality ; health ; aging ; longitudinal data

Abstract

We present two modifications of the stochastic process model of aging (SPM): the latent class SPM (LCSPM) and the genetic SPM (GenSPM). The LCSPM allows applications to populations consisting of latent subpopulations with distinct patterns of longitudinal trajectories of biomarkers that can also have different effects on the time-to-event outcome in each subpopulation. The GenSPM aims at applications analyzing genetic effects on the longitudinal trajectories and time-to-event outcomes taking into account observed characteristics affecting the probability of the presence of an allele/genotype in the genome of an individual. This case assumes that genetic information is available for a sub-sample of participants of the longitudinal study or for the entire sample. The GenSPM allows joint analyses of information from genotyped and non-genotyped subsamples which results in an increase in the power compared to analyses of the genotyped subsample alone. We present simulation studies and discuss practical applications of these approaches.