Validity of a wrist-worn consumer-grade wearable for estimating energy expenditure, sedentary behaviour, and physical activity in manual wheelchair users with spinal cord injury.

Abstract

To evaluate the validity of a consumer-grade wearable for estimating energy expenditure, sedentary behaviour, and physical activity in manual wheelchair users with spinal cord injury (SCI). Fifteen manual wheelchair users with SCI (C5-L1, four female) completed activities of daily living and wheelchair propulsion (2-8 km·h-1). Wrist-worn accelerometry data were collected using consumer-grade (z-Track) and research-grade (ActiGraph GT9X) devices. Energy expenditure was measured via indirect calorimetry. Linear regression was used to evaluate the prediction of criterion metabolic equivalent of task (MET) by each accelerometer's vector magnitude (VM). Area under the receiver operating characteristic curve (ROC-AUC) evaluated the accuracy of VM for discriminating between physical activity intensities and for identifying accelerometer cut-points. Standardised β-coefficients for the association between z-Track and ActiGraph VM for criterion MET were 0.791 (p < 0.001) and 0.774 (p < 0.001), respectively. The z-Track had excellent accuracy for classifying time in sedentary behaviour (ROC-AUC = 0.95) and moderate-to-vigorous physical activity (ROC-AUC = 0.93); similar values to the ActiGraph (ROC-AUC = 0.96 and 0.88, respectively). Cut-points for the z-Track were ≤37 g·min-1 for sedentary behaviour and ≥222 g·min-1 for moderate-to-vigorous physical activity. This study supports the validity of a consumer-grade wearable to measure sedentary time and physical activity in manual wheelchair users with SCI.