Assessment of the MyWellness Key accelerometer in people with type 2 diabetes

Abstract

Accelerometers are designed to measure physical activity (PA) objectively. The MyWellness Key (MWK) accelerometer has been validated primarily in younger, normal-weight populations. The aims of this study were to examine the accuracy of the MWK against directly measured lab-based exercise and free-living PA in people with type 2 diabetes, many of whom are older and overweight or obese. Thirty-five participants with type 2 diabetes completed the protocol, which included a laboratory-based session and a free-living phase. In the laboratory visit, participants completed a structured treadmill protocol wearing MWKs on each hip (all subjects) and bra cup (women only). The speed where each MWK switched from recording light- to moderate-intensity activity was determined for each MWK worn. In the free-living phase, participants wore the MWK for all waking hours for 2 weeks, and recorded exercise in PA diaries immediately after each exercise session. The mean cut-points between low (“Free”) and moderate (“Play”) intensity for the right and left waist-worn MWKs were 4.1 ± 0.5 km/h and 5.0 ± 0.9 km/h for the bra-mounted MWK; ideal cut-point would be 4.0 km/h. In the free-living phase, the Spearman correlation between PA according to PA diary and the waist-worn MWK was 0.81 (95% confidence interval (CI): 0.76, 0.85; P < 0.001), but only 0.66 (95% CI: 0.53, 0.77; P < 0.001) when on the bra. In conclusion, the waist-worn MWK measured PA volume accurately, and was acceptably accurate at discriminating between low- and moderate-intensity PA in people with type 2 diabetes. The MWK underestimated PA volume and intensity when worn on a bra.