The Whole Day Matters: Moving Beyond Analyzing Physical Activity, Sedentary Behavior and Sleep in Isolation
Wednesday, June 10, 12-1 pm ET.
Traditional approaches to analyzing physical activity, sedentary behavior, and sleep data have commonly examined these behaviors in isolation, despite the fundamental constraint that all time-use behaviors engaged in over the course of a day are codependent and sum to 24 hours. This webinar will introduce attendees to compositional data analysis (CoDA), an emerging analytic approach in the field of time-use epidemiology that accounts for the mutually exclusive and exhaustive nature of 24-hour movement behavior data, whereby time spent in one behavior reduces time available for others. Although CoDA is particularly well-suited for continuous, device-based time-use data from research-grade accelerometers and consumer wearables, this webinar will also discuss its application to self-reported measures, with emphasis on the careful selection of instruments that best align with compositional modeling assumptions. The session will begin with an accessible overview of CoDA, including core concepts and adequate reporting, and explain why broader adoption of this framework is needed to address limitations of traditional statistical models, which often overlook the perfect collinearity inherent in 24-hour time-use data and may yield biased estimates or misleading interpretations. A multidisciplinary panel of researchers representing physical activity, sedentary behavior, and sleep science will then discuss what motivated their adoption of CoDA, perceived barriers to wider uptake in the field, and how they are currently applying CoDA in observational and intervention research. The webinar will conclude by highlighting advanced and emerging applications of CoDA, including time-use reallocation..... To read the full webinar description please check our website at https://www.sbm.org/training/webinars