Analysis of Semi-Continuous Longitudinal Physical Activity Data

• Activity Number: 321
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Peter John De Chavez*+ and Lei Liu and Bonnie Spring and Juned Siddique
• Companies: Northwestern University and Northwestern University and Northwestern University Feinberg School of Medicine and Northwestern University
• Keywords: semi-continuous ; two-part model ; longitudinal ; physical activity ; accelerometer ; activity bout

Abstract

We present the results of a lifestyle intervention trial, the Make Better Choices (MBC) study, designed to test which combination of diet and activity change goals yields the greatest overall healthy lifestyle change. MBC was conducted as a randomized controlled trial where each participant was asked to wear an accelerometer to objectively measure physical activity during three study periods: 14-day baseline, 7-day treatment 1 (50% goal), and 14-day treatment 2/3 (100% goal). In order to assess the effect of the intervention on a participant's ability to meet current exercise guidelines, the outcome, referred to as an "activity bout", was defined as a moderate-vigorous physical activity (MVPA) that elapsed for 10 or more minutes, with allowance for interruptions of 1-2 minutes. Note that the values of this outcome variable are a combination of high frequency of zero values and skewed positive continuous values; hence the data are "semi-continuous". We thus describe the use of a two-part model for semi-continuous longitudinal data to analyze the effect of the MBC interventions on MVPA.