Validity of Wrist-Wearable Activity Devices for Estimating Physical Activity in Adolescents: Comparative Study

Area of Science:

• * Exercise Science
• * Biomedical Engineering
• * Wearable Technology

Background:

• * Wrist-wearable activity devices offer significant potential for clinical applications, particularly in monitoring physical activity (PA) in adolescents.
• * Accurate PA monitoring is essential for assessing current levels and the effectiveness of interventions aimed at increasing PA in adolescents.

Purpose of the Study:

• * To evaluate the validity of 11 commercial wrist-wearable activity devices for measuring total steps and 24-hour total energy expenditure (TEE).
• * To assess device accuracy in healthy adolescents under simulated free-living conditions.

Main Methods:

• * Nineteen adolescents (14-18 years) participated in a 24-hour metabolic chamber study.
• * Participants simultaneously wore 11 different wrist-wearable devices and a research-grade accelerometer.
• * Criterion measures included metabolic chamber energy expenditure (mcTEE) and accelerometer-measured steps (AGsteps).

Main Results:

• * Correlation coefficients for TEE ranged from 0.78 to 0.96, and for steps from 0.20 to 0.57.
• * Mean Absolute Percent Error (MAPE) for TEE ranged from 5.7% to 26.4%, and for steps from 14.2% to 27.6%.
• * Mi Band 2, UP3, Vivofit, and Bong 2s showed equivalent TEE estimates; Bong 2s also showed equivalent step counts.

Conclusions:

• * The Bong 2s demonstrated the highest accuracy for estimating both TEE and total steps in adolescents under simulated free-living conditions.
• * Further research is recommended to validate these devices in real-world settings and diverse physical activities.