Estimation of the Heart Rate and Energy Expenditure with a Smart Bracelet during Different Intensities of Exercise: A Reliability and Validity Study (Preprint)

BACKGROUND Wrist-worn wearables have become an increasingly popular way to assess exercise intensity and energy expenditure (EE) in health care and various other settings. However, there is still few research on the validity and reliability of wrist-worn wearables in monitoring exercise parameters at different intensities by people with different levels of physical activity. OBJECTIVE This study assessed the validity and reliability of the Mio Smart Bracelet (Lexin Mio Band 1) in measuring the heart rate (HR) and EE with different physical activity levels during exercise at different intensities. The second objective was to determine if wear position (i.e. dominate or nondominate hand) would cause differences in assessments. METHODS A total of 65 participants (N = 65; 63% female, n = 41; 37% male, n = 24) wore the Lexin smart bracelet (Mio Band 1) and its validity and reliability were assessed against two “gold standard” devices: the Polar H10 heart rate band (for HR) and the gas-analysis system (GAS, Metalyzer 3B) for EE. The participants completed one maximal oxygen uptake (VO2max) test and a running exercise test twice (10-min resting period; 6-min treadmill exercise under three intensities; 10-min recovery period). To measure test-retest reliability, intra-group correlation coefficient (ICC) were determined between the first and second treadmill test. Validity was determined by comparing the Mio with the gold standard (Polar and GAS), using mean differences, mean absolute percentage errors (MAPE), ICC and Bland-Altman analysis. RESULTS For the reliability, the HR consistency between the tests was significant (p < 0.05) in all the states except the Mio in HPAL (P < 0.05), while the EE consistency between the tests, only the right-wrist Mio showed significance in the low-physical-activity level (LPAL) population, and for the moderate-intensity exercise (MIE) and the high-intensity exercise (HIE) in the moderate-physical-activity level (MPAL) population. With regard to the validity, the MAPE between the Polar and the left- and right-wrist Mio bracelets (henceforth “left and right Mios”) ranged from 1% to 7.55%, while the MAPE between the left and right Mios ranged from 10.37% to 32.42%. The HR consistency in the LPAL, MPAL, and high-physical-activity level (HPAL) populations was significant between polar and the left and right Mios at different exercise intensities (P < 0.01), and insignificant only in HPAL group under HIE (P>0.05). As for the EE consistency, it was significant in the left and right Mios under MIE and in the right Mio under HIE in the LPAL group; in the left and right Mios under MIE or HIE, and in the right Mio during recovery in MPAL group; and in the left and right Mios under MIE in the HPAL group (P < 0.05). CONCLUSIONS For HR measurement, the Lexin Mio bracelet showed good reliability and validity among people with different physical activity levels who were exercising under various intensities in a laboratory setting. However, for the estimation of the EE, the Lexin Mio bracelet showed low reliability and validity. CLINICALTRIAL This study was approved by the Experimental Ethics Committee for Sports Science of Beijing Sport University (registration number 2020133H). The participants were informed of the study purpose and procedures and then signed a written informed consent form before enrollment in this study.