Validation of the Apple Watch for Estimating Moderate-to-Vigorous Physical Activity and Activity Energy Expenditure in School-Aged Children.

Sunku Kwon,Youngwon Kim,Wonwoo Byun,Yang Bai,Ryan D Burns,Timothy A Brusseau

doi:10.3390/s21196413

Abstract

The Apple Watch is one of the most popular wearable devices designed to monitor physical activity (PA). However, it is currently unknown whether the Apple Watch accurately estimates children’s free-living PA. Therefore, this study assessed the concurrent validity of the Apple Watch 3 in estimating moderate-to-vigorous physical activity (MVPA) time and active energy expenditure (AEE) for school-aged children under a simulated and a free-living condition. Twenty elementary school students (Girls: 45%, age: 9.7 ± 2.0 years) wore an Apple Watch 3 device on their wrist and performed prescribed free-living activities in a lab setting. A subgroup of participants (N = 5) wore the Apple Watch for seven consecutive days in order to assess the validity in free-living condition. The K5 indirect calorimetry (K5) and GT3X+ were used as the criterion measure under simulated free-living and free-living conditions, respectively. Mean absolute percent errors (MAPE) and Bland-Altman (BA) plots were conducted to assess the validity of the Apple Watch 3 compared to those from the criterion measures. Equivalence testing determined the statistical equivalence between the Apple Watch and K5 for MVPA time and AEE. The Apple Watch provided comparable estimates for MVPA time (mean bias: 0.3 min, p = 0.91, MAPE: 1%) and for AEE (mean bias: 3.8 kcal min, p = 0.75, MAPE: 4%) during the simulated free-living condition. The BA plots indicated no systematic bias for the agreement in MVPA and AEE estimates between the K5 and Apple Watch 3. However, the Apple Watch had a relatively large variability in estimating AEE in children. The Apple Watch was statistically equivalent to the K5 within ±17.7% and ±20.8% for MVPA time and AEE estimates, respectively. Our findings suggest that the Apple Watch 3 has the potential to be used as a PA assessment tool to estimate MVPA in school-aged children.

Highlights

There were no significant differences in age, height, weight, Body mass index (BMI) percentile, and waist circumference between boys and girls
The Mean absolute percent errors (MAPE) in Apple Watch 3 was 1% for Moderate-to-vigorous physical activity (MVPA) estimate compared to the K5 measures
Compared to the measures from indirect calorimetry, the Apple Watch 3 comparably estimated time the spent in MVPA and activity energy expenditure (AEE) under the simulated free-living condition in elementary school students

Summary

Introduction

Moderate-to-vigorous physical activity (MVPA) in children offers numerous health benefits, including the prevention of childhood obesity [1], improved bone strength [2], and improved cardio-respiratory endurance [3,4]. Given that physical activity (PA) monitoring plays an important role in promoting children’s MVPA [5], PA assessment, which focuses on estimating MVPA time and activity energy expenditure (AEE), can be useful for PA monitoring. Such quantification of activity intensities, time, and AEE in real-time has been made possible through the advancement of micro-technology and the deployment of various consumer-based wearable devices, such as activity trackers and smartwatches.

Objectives

Methods

Results

Discussion

Conclusion