Reliability, validity and responsiveness of physical activity monitors in patients with inflammatory myopathy

Abstract

Idiopathic inflammatory myopathies (IIMs) cause proximal muscle weakness, which affects the ability to carry out the activities of daily living. Wearable physical activity monitors (PAMs) objectively assess continuous activity and potentially have clinical usefulness in the assessment of IIMs. We examined the psychometric characteristics for PAM outcomes in IIMs. Adult IIM patients were prospectively evaluated (at baseline, 3 months and 6 months) in an observational study. A waist-worn PAM (ActiGraph GT3X-BT) assessed average step counts/minute, peak 1-minute cadence, and vector magnitude/minute. Validated myositis core set measures (CSMs) including manual muscle testing (MMT), physician global disease activity (MD global), patient global disease activity (Pt global), extramuscular disease activity (Ex-muscular global), HAQ-DI (HAQ disability index), muscle enzymes, and patient-reported physical function were evaluated. Test-retest reliability, construct validity, and responsiveness were determined for PAM measures and CSMs, using Pearson correlations and other appropriate analyses. A total of 50 adult IIM patients enrolled [mean (s.d.) age, 53.6 (14.6); 60% female, 94% Caucasian]. PAM measures showed strong test-retest reliability, moderate-to-strong correlations at baseline with MD global (r = -0.37 to -0.48), Pt global (r=-0.43 to -0.61), HAQ-DI (r = -0.47 to -0.59) and MMT (r = 0.37-0.52), and strong discriminant validity for categorical MMT and HAQ-DI. Longitudinal associations with MD global (r=-0.38 to -0.44), MMT (r = 0.50-0.57), HAQ-DI (r = -0.45 to -0.55) and functional tests (r = 0.30-0.65) were moderate to strong. PAM measures were responsive to MMT improvement ≥10% and moderate-to-major improvement on ACR/EULAR myositis response criteria. Peak 1-minute cadence had the largest effect size and standardized response means. PAM measures showed promising construct validity, reliability, and longitudinal responsiveness; especially peak 1-minute cadence. PAMs are able to provide valid outcome measures for future use in IIM clinical trials.