Forecasting Adult Weight Status Trends in Canada with Longitudinal Population-based Data

Abstract

The prevalence of obesity in Canada and other Western countries has increased dramatically over the last 30 years. This project modelled future prevalence of weight status trends under a Markov framework by estimating transition probabilities between categorical weight states. A microsimulation was developed to track individual weight status histories through discrete-time updates. Model and simulation parameters were derived from a nationally representative longitudinal survey which followed 17,000 participants over a 20 year period. A series of weight prevention and intervention strategies were simulated to determine their potential impact on weight status trends and population health utility.