Visualising accelerometer-based 24/7 human movement behaviour data: an umbrella review and framework development from the LABDA project.

Abstract

Background: Regular Physical Activity (PA) is important for disease prevention and health promotion. PA has been assessed through surveys, questionnaires, and devices such as accelerometers. Alongside PA, Sedentary Behaviour (SB) and sleep are the main components of 24/7 movement behaviours, and their adequate measurement is important for assessing health outcomes. Many different metrics to summarise 24/7 movement behaviours are used; however, little attention has been paid to visualising these metrics. Data visualisation is likely to impact the way results are communicated and understood by different audiences. This study systematically reviews 24/7 movement behaviour metrics, presents an overview of their visualisations, and develops a framework to guide context-specific visualisation choices.

Methods: An umbrella review was conducted in February 2025 in Scopus and Web of Science. Included papers were reviews of any type, with any human population and study design, having at least one of the three 24/7 movement behaviours as exposure or outcome measured through accelerometers, and clearly reporting the outcome metrics. Data extraction and an adapted thematic data analysis were performed in April 2025. The overview of the visualisations used for the metrics identified in the review and thematic analysis was created through non-systematic web searches and use of Microsoft Copilot. Finally, a framework was created based on the sender-receiver model for effective communication.

Results: In total, 93 reviews were included, with a total of 5667 articles reporting on 134 unique output metrics based on accelerometer data. The most common metrics were step counts and time spent in Moderate-to-Vigorous Physical Activity (MVPA). The non-systematic web searches showed that most researchers use bar charts, line graphs, or pie graphs to visualise 24/7 movement behaviour data, while Copilot input provided more options of visualisations. The resulting framework was the product of an iterative process aggregating the previous results, providing clear guidance for organising metrics and their corresponding visualisations.

Conclusions: This study structures and summarises types of visualisations of accelerometer-derived metrics to describe 24/7 human movement behaviour data. Future research is needed to apply the framework in practical contexts and investigate how the visualisations are perceived by different audiences.