Psychometric Properties of Free-Living Step-Based Metrics (Daily Steps and Peak Cadence) in Multiple Sclerosis.

Abstract

We examined the reliability, precision, and clinically detectable change of step-based metrics (daily steps, peak 30-min cadence [Peak-30_CAD], and peak 1-min cadence [Peak-1_CAD]) over 6 months in the absence of intervention and evaluated the construct validity through correlations with laboratory-assessed walking and gait performance among persons with multiple sclerosis (MS). Cross-sectional study. University-based laboratory. Seventy-eight ambulatory adults (18-64y) with MS. Not applicable. Free-living physical activity (via ActiGraph GT3X accelerometer), the Timed 25-Foot Walk, 6-minute walk, gait assessment (gait velocity, step length, and time), and disability status (the 12-item MS Walking Scale, Patient-Determined Disease Steps, and Self-Report Expanded Disability Status Scale) were measured before and after 6 months without any intervention. Step-based metrics were stable with no significant changes across time (P>.05) and demonstrated good test-retest reliability (intraclass correlation coefficients: 0.80-0.85) and acceptable precision (SEM%s:14.4%∼24.3%). The minimal detectable changes at 95% CIs (MDC₉₅) values for Peak-30_CAD, Peak-1_CAD, and daily steps were 25.6 steps/min, 31.0 steps/min, and 2909.2 steps/d, respectively. There were consistent, strong associations between peak cadence with walking tests, gait parameters, and disability status at both time points (|r_s|=.52-.79), even after controlling for daily steps (|pr_s|=.25-.58; P<.05). Walking represents an important clinical endpoint in people with MS, yet it is often measured in controlled settings using performance-based tests that might not reflect real-world status. Our findings support step-based metrics via accelerometry as reliable and valid measures of free-living ambulatory performance and may inform the inclusion of these metrics in clinical trials among people with MS.