New metrics for analysis and presentation of device-based indices of physical activity : Exploration of a novel way of reducing accelerometer data into analytical and translational indices of physical activity

Abstract

Background: Device-based measurements of human behavior are becoming the standard tool to collect precise, valid, and free-living indices of habitual physical activity. However, differences between brands and lack of consensus regarding data reduction strategies, hampers between and within study comparisons. New metrics such as the analytical (average acceleration and intensity gradient) and translational metrics (MX) proposed by Rowlands et al. (2019a) might overcome some of these limitations. Currently the new metrics have not been applied to large representative data sets and comparisons of measurement properties between traditional and new accelerometer metrics are scarce. Purpose: Provide normative new metrics for population estimates of PA and compare measurement properties of new and traditional count-based measures of physical activity. […] Conclusion: We display normative values of new metrics for device-based indices of physical activity in a representative sample of Norwegian adults and older, showing the feasibility and usefulness of this approach. Furthermore, the new approach provides a way of presenting data in a way that is interpretable to a wider audience. Lastly, the appropriateness of the new metrics is further strengthened by the similar predictive properties compared to the traditional metrics. Given the new approach’s ability to overcome central limitations of device-based measurements of physical activity, this might be a feasible way forward in order to aid comparability between studies of physical activity and health. Keywords: accelerometer-assessed physical activity, analytical metrics, translational metrics, average acceleration, intensity gradient(IG), MX metrics, and counts per minute (CPM).