Continuous physical activity recording - Consumer-based activity trackers in epidemiological studies

Abstract

Physical activity is an important modifiable lifestyle factor that can improve general health and reduce the risk of disease. Currently, collecting data on physical activity in epidemiological studies are generally limited to long-term but self-reported and inaccurate physical activity questionnaires and/or using short-term but objective and more accurate accelerometers. Consumer-based activity trackers are designed for long-term objective data collection and can therefore potentially be used to close this gap. The objective of this dissertation was therefore to explore and develop new methods for collecting data on physical activity in epidemiological studies using consumer-based activity trackers. The four included papers apply different methods to explore the objective from multiple angles. Results includes an overview of how activity tracker sensor support has changed over time, recommendations when choosing an activity tracker model for future physical activity research, recommendations for increasing activity tracker wear time among participants in clinical studies, as well as knowledge about activity tracker validity and physical activity trends during the Norwegian COVID-19 lockdown in 2020. Finally, the dissertation describes a system for automatic and continuous data collection using consumer-based activity trackers from multiple providers. We show the usability of this system by accessing and analysing historic activity tracker data from participants who wore a tracker before-, during-, and after the COVID-19 lockdown period. The proposed system can be a valuable addition to existing methods for physical activity assessment by contributing to closing the above-mentioned method gap.