Influence of upper limb activity on the step count and accuracy of sleep time of a wristband-type physical activity tracker

Abstract

Background A wristband-type consumer physical activity tracker (PAT) is commonly used in rehabilitation to assess an individual's physical activity. However, under the free-living setting, the wristband-type PAT tends to overestimate step counts when compared with the researchstandard criterion. Also, daily rhythm characteristics, such as sleep time, are difficult to monitor accurately based solely on self-reporting. Purpose To identify the conditions measured as step counts by a wristband-type consumer PAT when using the upper limbs in daily living, and the measurement accuracy of the sleeping time estimated from the wristband-type PAT. Methods Forty participants (20 females, mean age 32.65 ± 9.52 years) were enrolled in two experiments in this study. In Experiment 1, we measured the influence of upper limbs activity (movement speed and distance) on step counts of wristband-type and waist holder-type PAT in two upper limb tasks. In Experiment 2, we verified the measurement accuracy of two sleep times by wristband-type PAT using a self-reported survey for 3 days. Results The results of Experiment 1 revealed that the step counts using wristband-type PAT were influenced by upper limbs activity depending on movement distance (F (1, 19) = 31.705, p < 0.001) but not speed (F (1, 19) = 2.669, p < 0.117). Whereas, there was no relationship between step counts and upper limb activity in waist holder-type PAT. The results of Experiment 2 showed that the sleep times of wristband-type and self-report had a strong correlation (coefficient value = 0.93, p < 0.001). Conclusions This PAT is useful for capturing changes in the amount of physical activity and the daily rhythm within the individual. It can be expected to be used for rehabilitation support centered on upper limb activity and daily rhythm.