Modeling Physical Activity Distributions for Populations

• Activity Number: 640
• Type: Topic Contributed
• Date/Time: Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Sarah Nusser*+ and David Osthus and Bryan Stanfill and Wayne A. Fuller and Alicia L. Carriquiry
• Companies: Iowa State University and Iowa State University and Iowa State University and Iowa State University and Iowa State University
• Keywords: physical activity ; survey ; measurement error

Abstract

There is considerable concern about whether US adults engage in sufficient levels of physical activity. One way to evaluate activity patterns in populations is to estimate the distribution of activity for individuals. We have developed methods for estimating distributions of usual daily energy expenditure as expressed as metabolic equivalents (METs) expended over time, which adjusts for the effect that body size has on an individual's energy expenditure. Using self-reported activity data translated to MET-hours and data from an objective physical monitor, we develop a model for the distribution of individual mean MET-hours in the population. We also explore models for METs-hours from bouts of moderate or vigorous activity per week lasting at least 10 minutes, which can be more directly compared to US activity guidelines. We illustrate these methods using data are from the Iowa Physical Activity Measurement Survey.