Using Smart Bracelets to Assess Heart Rate Among Students During Physical Education Lessons: Feasibility, Reliability, and Validity Study (Preprint)

Abstract

BACKGROUND More and more wrist-worn trackers are examined in health care. However, feasibility, reliability and validity studies for Physical Education (PE) classes remain scarce. OBJECTIVE This article was to examine the reliability and validity of Fizzo Smart Bracelet (FSB)in measuring heart rate (HR) in laboratory and examine the validity and feasibility in PE classes. METHODS In Study 1, 11 healthy subjects (5 males 23.8 ± 1.8 years, 6 females 21.0 ± 1.3 years) completed the test which included running on the treadmill at the speed of 6 km/h and 12 km/h for 12min and 5min respectively for two times. During the test, participants wore two FSBs on their left and right wrists to measure the HR data, resembling a watch. At the same time, the Polar Team2 Pro (Polar) was worn on the chest as the gold standard on their chest. In Study 2, we went to 10 schools and measured 24 students (11 males 14.0 ± 1.4 years, 13 females 13.9 ± 1.1 years) in PE classes. During the PE classes each student worn a Polar on chest and an FSB on the right wrist to measure HR data. At the end of the PE classes, the students and their teachers finished the questionnaire for assessment the feasibility of FSB. To measure reliability, intraclass correlations (ICC), mean difference (MD), standard error of measurement (SEM), mean absolute percentage errors (MAPE) were used. To measure validity, intraclass correlations (ICC), limits of agreement (LOA), mean absolute percentage errors (MAPE) were calculated. Bland-Altman plots were constructed. Percentage value were used to estimate the feasibility of FSB. RESULTS In Study 1 the FSB showed high reliability between the Left Fizzo (LF) and Right Fizzo (RF) (ICC > 0.97, MD < 0.7, SEM < 0.56 and MAPE < 1.45%). The validity of LF and RF were high (ICC > 0.98, MAPE < 1.85%). Bland-Altman plots showed strong correlation. LF (bias: 0.48, LOA: -3.94 ~ 4.89 bpm) and RF (bias: 0.56, LOA: -4.60 ~ 5.72 bpm). Compared to study 1, the validity of FSB was lower in Study 2, but still moderate (ICC > 0.70, MAPE < 9.0%), and the FSB showed broad LOA in the Bland-Altman plots (bias: -2.60, LOA: -38.89 ~ 33.69bpm). FSB slightly underestimate the HR compared with Polar. Most participants considered the FSB is very comfortable and easy to apply and remove. All the teachers thought FSB is helpful. CONCLUSIONS Running on the treadmill in the laboratory, both LF and RF were all accurate. Compare to the laboratory, the validity of FSB was decreased in PE classes but still reached moderate level. FSB is comfortable to wear, easy to apply and remove, it has high application value in PE class setting.