Stochastic Process Model and Its Applications to Analysis of Longitudinal Data

Abstract

Longitudinal studies are widely used in medicine, biology, population health and other areas related to bioinformatics. A broad spectrum of methods for joint analysis of longitudinal and time-to-event (survival) data has been proposed the in last few decades. The Stochastic process model (SPM) represents one possible framework for modelling joint evolution of repeatedly measured variables and time-to-event outcome typically observed in longitudinal studies. SPM is applicable for analyses of longitudinal data in many research areas such as demography and medicine and allows researchers to utilize the full potential of longitudinal data by evaluating dynamic mechanisms of changing physiological variables with time (age), allowing the study of differences, for example, in genotype-specific hazards. SPM allows incorporation of available knowledge about regularities of aging-related changes in the human body for addressing fundamental problems of changes in resilience and physiological norms. It permits evaluating mechanisms that indirectly affect longitudinal trajectories of physiological variables using data on mortality or onset of diseases. In this tutorial we explain the basic concepts of SPM, its current state and possible applications, corresponding software tools and show practical examples of analysis of joint analysis of longitudinal and time-to-event data with this methodology.