Table 3.

Objective 1: ROC curve-generated RA-specific triaxial (VM) accelerometer count-based cut-points

Epoch (1-min)	RA-specific count-based cut-points (VM activity counts/min)	Sensitivity	1-Specificity	AUC
Sedentary time	≤ 244	0.99	0.03	1.00
LPA	> 244–< 2502	–	–	–
MPA	≥ 2502	0.87	0.11	0.94

RA-specific count-based cut-points were developed for sedentary time, LPA and MPA, based on average GT3X+ activity counts (VM) and METs during steady-state VO₂ (± 0.5 ml/min/kg [min 4–6 of each activity]). LPA count-based cut-points were defined using the upper cut-point threshold of sedentary time and the lower cut-point threshold of MPA. AUC demonstrated accuracy of the RA-specific count-based cut-points (0.90–1.00 = excellent; 0.80–0.89 = good; 0.70–0.79 = fair; 0.60–0.69 = poor; < 0.60 = failure)

RA rheumatoid arthritis, VM vector magnitude, AUC area under the curve, LPA light-intensity physical activity, MPA moderate-intensity physical activity, – does not apply