Modeling Accelerometer-Based Physical Activity Using a Measurement Error Approach

• Activity Number: 161
• Type: Topic Contributed
• Date/Time: Monday, August 4, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Korean International Statistical Society
• Author(s): Jungwha (Julia) Lee*+ and Jing Song and Dorothy Dunlop
• Companies: Northwestern University and Northwestern University and Northwestern University
• Keywords: accelerometer ; physical activity ; measurement error

Abstract

Physical activity (PA) is a modifiable lifestyle factor for many chronic diseases with established health benefits. PA outcomes using accelerometers are measured and assessed in many studies, but there are limited statistical methods analyzing accelerometry data. We describe a measurement error modeling approach to estimate the distribution of habitual physical activity and the sources of variation in accelerometer-based physical activity data from a sample of adults with or at risk of knee osteoarthritis. We model both the intra- and inter-individual variability in measured physical activity. Our model allows us to account for and adjust for measurement errors, biases, and other sources of intra-individual variations.