A Multistate Time-Dependent Model of Health Transitions During Aging

• Activity Number: 475
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Arnold Mitnitski*+ and Xiaowei Song and Kenneth Rockwood
• Companies: Dalhousie University and Dalhousie University and Dalhousie University
• Keywords: multistate model ; stochastic process ; Poisson distribution ; health status ; mortality ; aging

Abstract

We present a multi-state stochastic model to describe longitudinal changes in health status in people from late middle age. Health status in individuals is represented by the number of adverse health related characteristics that they have accumulated. The ratio of this number to the total number of deficits, known as a frailty index accounts for the heterogeneity of people at the same age. Using data from the Canadian National Population Health Survey (age 55+, n=4330, every 2 years, 7 cycles) we calculated a frailty index based on 33 health deficits. We demonstrate that transitions in the number of deficits during any time period can be approximated using a time-dependent Poisson distribution with the mean decelerating over time, according to the square-root-of-time kinetics characteristic for stochastic processes (e.g. diffusion, Brownian motion) whereas the probability of death accelerates. The model has a high degree of precision, 'explaining' over 98% of the observed variance. Note that a particularly useful aspect of this approach is that it permits simultaneous estimation of the probabilities of both death and transitions in different directions and to many degrees.