ABSTRACT
The objective of this study was to investigate the validity of measured caloric expenditure from a fitness smartwatch designed to measured values in wheelchair users against criterion values from a portable metabolic system. 15 wheelchair users and 15 able-bodied participants completed multiple tasks; wheelchair treadmill routine at 30, 45, and 60 strokes per minute, arm cycle ergometry at 45, 60, and 80 revolutions per minute, and arm cycle ergometry VO2Peak test. There were no interactions for device or task and group (wheelchair users vs. able bodied, p=0.375-0.944) therefore results were pooled across groups for all measures. The smartwatch exhibited poor to moderate caloric expenditure association during wheelchair treadmill routine (ICC<0.39) and arm cycle ergometry (ICC<0.541). Smartwatch underestimated caloric expenditure during the wheelchair treadmill task (Mean differences (Limits of Agreement)) (-2.11 (-8.19-3.96), -3.68 (-12.64-5.28), and -4.51 (-15.05-6.02)) and overestimated during the arm cycle ergometry task (0.89 (-3.10-4.88), 3.40 (-0.31-7.12), and 2.81 (-1.71-7.32)). The smartwatch is currently not well suited to calculate caloric expenditure when performing exercise tasks on a wheelchair treadmill and arm cycle ergometry.
