Details

Purpose: To present complemental analytical approaches for accelerometer data to estimate patterns and compositions of 24-hour movement behaviors and their association with cardiometabolic health.

Rationale: Wearable accelerometers provide possibilities to quantify hour-by-hour physical activity, sedentary time and sleep over 24-hours across a number of days. This has opened new avenues to examine in more detail how various activity patterns vary within or between days as well as how allocation of movement behaviors during the 24-hour day vary and change over time. All movement behaviors are interrelated and this can be accounted in the Compositional data analysis when examining variation and changes in 24-h movement behaviors. Additionally, data-driven approaches can handle multidimensional and correlated data, which makes them appropriate candidates for understanding the complex interrelationship among movement behaviors. These new analytical approaches provide possibilities to increase our understanding of the interplay between movement behaviors and their pattern with cardiometabolic health.

Objectives:

1) To introduce and compare data-driven and compositional data analysis to examine 24-hour accelerometer data.

2) To present and synthesize how 24-hour movement behaviors and their patterns associate with cardiometabolic health.

3) To highlight the potential of new analytical approaches for novel data discoveries.

4) To discuss the future possibilities of utilizing repeated 24-hour accelerometer data for studying associations with cardiometabolic and other chronic diseases.

Summary: First, the chair (prof. Sari Stenholm) provides an overview about movement behaviors and cardiometabolic health. Dr. Kristin Suorsa starts by introducing compositional data analysis based on repeated accelerometer data. She will then share findings about changes in 24-h movement behaviors and obesity indicators in relation to the transition from work to retirement. Dr. Vahid Farrahi continues by describing data-driven approaches. He will describe person-centered approaches for profiling 24-hour movement behaviors to identify novel profiles (e.g. active couch potatoes) and how they may explain differences in cardiometabolic health. Dr. Tuija Leskinen will present another data-driven approach, group-based trajectory analysis, to identify profiles of daily physical activity patterns (e.g. weekend warriors) and assess their associations with cardiometabolic biomarkers. Finally, the discussant associate prof. Annemarie Koster will summarize the presentations and highlight directions for future studies.

Format: Prof. Stenholm provides the introduction (5min) followed by three individual presentations by Drs. Suorsa, Farrahi and Leskinen (15min/presentation). Associate prof. Koster will conclude with summary (5min) and facilitate discussion among  presenters and audience (20min).

Interaction: The live audience will be prompted to ask questions during the discussion.

Speaker