Measurement of sedentary time and physical activity in rheumatoid arthritis: an ActiGraph and activPAL™ validation study

Accurate measurement of sedentary time and physical activity (PA) is essential to establish their relationships with rheumatoid arthritis (RA) outcomes. Study objectives were to: (1) validate the GT3X+ and activPAL3(μ)™, and develop RA-specific accelerometer (count-based) cut-points for measuring sedentary time, light-intensity PA and moderate-intensity PA (laboratory-validation); (2) determine the accuracy of the RA-specific (vs. non-RA) cut-points, for estimating free-living sedentary time in RA (field-validation). Laboratory-validation: RA patients (n = 22) were fitted with a GT3X+, activPAL3(μ)™ and indirect calorimeter. Whilst being video-recorded, participants undertook 11 activities, comprising sedentary, light-intensity and moderate-intensity behaviours. Criterion standards for devices were indirect calorimetry (GT3X+) and direct observation (activPAL3(μ)™). Field-validation: RA patients (n = 100) wore a GT3X+ and activPAL3(μ)™ for 7 days. The criterion standard for sedentary time cut-points (RA-specific vs. non-RA) was the activPAL3(μ)™. Results of the laboratory-validation: GT3X—receiver operating characteristic curves generated RA-specific cut-points (counts/min) for: sedentary time = ≤ 244; light-intensity PA = 245–2501; moderate-intensity PA ≥ 2502 (all sensitivity ≥ 0.87 and 1-specificity ≤ 0.11). ActivPAL3(μ)™—Bland–Altman 95% limits of agreement (lower–upper [min]) were: sedentary = (− 0.1 to 0.2); standing = (− 0.7 to 1.1); stepping = (− 1.2 to 0.6). Results of the field-validation: compared to the activPAL3(μ)™, Bland–Altman 95% limits of agreement (lower–upper) for sedentary time (min/day) estimated by the RA-specific cut-point = (− 42.6 to 318.0) vs. the non-RA cut-point = (− 19.6 to 432.0). In conclusion, the activPAL3(μ)™ accurately quantifies sedentary, standing and stepping time in RA. The RA-specific cut-points offer a validated measure of sedentary time, light-intensity PA and moderate-intensity PA in these patients, and demonstrated superior accuracy for estimating free-living sedentary time, compared to non-RA cut-points.