Reporting adherence, validity and physical activity measures of wearable activity trackers in medical research: A...

Wearable activity trackers are gaining traction in medical research, providing both real-time and remote monitoring of physical fitness. Activity trackers offer an excellent source of personalized physical activity data from patients, as well as healthy individuals, that would provide insights into healthcare analytics and user-feedback on health status. In addition, these activity trackers would also allow researchers to monitor symptom severity and assist clinicians in providing their patients a more holistic care. Despite the promise of wearable device technology, there is still a lack of standardization in the medical literature regarding the analysis and reporting of adherence, validity and physical activity data generated by these activity trackers.

We performed a systematic review to identify the activity tracker-derived measures and evaluate the relations of reported adherence, validity, and physical activity types across currently available literature.

The searches were performed using Pubmed and Embase databases. Studies enrolling at least 1,000 human subjects regardless of health or disease status, using activity trackers of any brand used to track step count, distance, heart rate, energy expenditure or activity intensity, were included. Studies have been published between 2009 to March 2021, with editorials, systematic reviews, meta-analysis, grey literature, validation studies, study protocols and studies using smartphone trackers being excluded. This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

A total of 27 studies met the eligibility criteria and were included in the review, with a total of 514,418 and 1,186,530 subjects recruited in observational and interventional studies, respectively. Apart from ActiGraph (n = 11, 41%), Fitbit (n = 4, 15%) and Axivity (n = 3, 11%) were found to be the most commonly used activity trackers in both types of studies. The wear duration of activity trackers ranged from 1 day to 59 months, with 1 week being the most common length (n = 16, 59%). The most frequently collected physical activity measure was activity intensity (n = 21, 78%), followed by step count (n = 9, 33%) and energy expenditure (n = 2, 7%). Most studies defined a valid day as wear-time of at least 10 h within 1 day (n = 10, 37%), and a valid interval as a week with at least 3 valid days (n = 8, 30%).

This systematic review reveals the diverse analysis and reporting of activity tracker data in the medical literature. Future studies will need to evaluate the feasibility on adopting minimum reporting thresholds of data generated by wearable activity trackers.