Evaluation of validity of commercial devices for monitoring physical activity in physically active adult persons

Više senzorni uređaji koji prate tjelesnu aktivnost i pri tome koriste akcelerometar u suradnji s ostalim fiziološkim podacima su dizajnirani da omoguće jednostavno praćenje tjelesne aktivnosti. Jedan od takvih uređaja trenutno na tržištu je popularna fitnes narukvica Fitbit Charge 2. Cilj ovog diplomskog rada bio je utvrditi valjanost mjernog uređaja Fitbit Charge 2 za praćenje intenziteta tjelesne aktivnosti uspoređujući ga s uređajem znanstvene razine Polar H7. Popratno je s kriterijskim uređajem uspoređen i uređaj istog proizvođača starije generacije, Polar OH1. Šesnaest mladih odraslih, tjelesno aktivnih sudionika, prosječne dobi 25 godina, mjereni su s tri uređaja za vrijeme trčanja proizvoljnim intenzitetom približno dva kilometra. Fitbit Charge 2 je u odnosu na Polar H7 podcijenio frekvenciju srca za 9,8% (Fitbit: 135,6±12,7 o/min, Polar H7: 150,3± 9,5 o/min, p=0,001), a opću kalorijsku potrošnju za 7,5% (Fitbit: 142,8±16,6 kcal, Polar H7: 154,3±18,7 kcal, p=0,053). Između originalno mjerenih frekvencije srca i kalorijske potrošnje između dva uređaja nije utvrđena značajnija pozitivna, niti negativna povezanost. Za usporedbu, između dva Polar uređaja nisu utvrđene statistički značajne razlike, a utvrđena je pozitivna povezanost između njihovih originalno mjerenih pokazatelja. Kod naknadno izvedenog metaboličkog ekvivalenta, koji ukazuju na relativni intenzitet i energetsku potrošnju, nije uočena značajna razlika između Fitbit Charge 2 uređaja i Polar H7 uređaja (Fitbit: 9,4±2 MET, Polar H7: 10,3±2,5 MET, p=0,043). Iz ovog istraživanja proizlazi da Fitbit Charge 2 nije adekvatno valjan za neposredno mjerenje frekvencije srca i kalorijske potrošnje. No, postoje naznake da bi se podaci prikupljeni Fitbit Charge 2 uređajem u istraživačkoj praksi mogli koristiti za praćenje intenziteta aktivnosti mladih, tjelesno aktivnih odraslih ljudi ukoliko se originalni podaci transformiraju u relativne jedinice poput metaboličkog ekvivalenta. Ipak, obzirom na određena ograničenja trenutnog istraživanja, buduće studije bi trebale potvrditi navedenu pretpostavku. Multi-sensor devices that monitor physical activity and use an accelerometer in conjunction with other physiological data are designed to allow easy monitoring of physical activity. One such device currently on the market is the popular fitness bracelet Fitbit Charge 2. The aim of this study was to determine the validity of the measuring device Fitbit Charge 2 for monitoring the intensity of physical activity by comparing it with a scientific level device Polar H7. The device of the same manufacturer of the older generation, Polar OH1, was compared with the criterion device. Sixteen participants, young adults, physically active, average age 25 years, were measured with three devices while running at an arbitrary intensity of approximately two kilometers. Fitbit Charge 2 underestimated the heart rate compared to Polar H7 by 9.8% (Fitbit: 135.6 ± 12.7 bpm, Polar H7: 150.3 ± 9.5 bpm, p = 0.001), and total caloric consumption by 7.5% (Fitbit: 142.8 ± 16.6 cal, Polar H7: 154.3 ± 18.7 cal, p = 0.053). No significant positive or negative association was found between the originally measured heart rate and caloric expenditure between the two devices. For comparison, no statistically significant differences were found between the two Polar devices, and a positive correlation was found between their originally measured indicators. In the subsequently derived metabolic equivalent, which indicates relative intensity and energy consumption, no significant difference was observed between the Fitbit Charge 2 device and the Polar H7 device (Fitbit: 9.4 ± 2 MET, Polar H7: 10.3 ± 2.5 MET, p = 0.043). From this research, it follows that Fitbit Charge 2 is not adequately valid for direct measurement of heart rate and caloric expenditure. However, there are indications that the data collected by the Fitbit Charge 2 device in research practice could be used to monitor the activity intensity of young, physically active adults if the original data are transformed into relative units such as metabolic equivalent. However, given certain limitations of current research, future studies should confirm this assumption.