Characterizing within-subject variability in quantified measures of balance control: A cohort study

To longitudinally assess individuals using quantified measures, we must characterize within-subject variability (WSV) of the measures.

What is the natural within-subject variability (WSV) that can be expected in postural control over 3+ days?

Thirteen individuals without orthopedic or neurologic impairment (mean(SD) = 55 (9) years; 76 (18) kg; 11 females/2 males) were recruited from a community workplace and consented to participate. Participants stood quietly with eyes closed (QEC) on a force platform (5 x 1 min x 6 days) in two stances: comfortable and narrow. We recorded center of pressure (COP) and calculated COP-based balance parameters. To analyze variance components, we applied a linear mixed model for repeated measures, calculating within-subject standard deviation (SDws) from the pooled variance not attributable to between-subject variability. To estimate WSV, we scaled SDws by a confidence interval (CI) factor (e.g. WSV at the 95%CI = WSV₉₅ = SDws * 1.96) and report WSV₉₅ for a range of conditions previously reported in the literature and the following measures previously found sensitive to or predictive of health: (primary) WSV₉₅ of root-mean square amplitude of medial-lateral COP during QEC (RMSml); (secondary) WSV₉₅ of COP ellipse area (COPa); (secondary) WSV₉₅ of mean medial-lateral COP velocity (COPvml) during QEC.

WSV₉₅ was estimated at RMSml = 0.8 mm, COPa = 99mm², and COPvml = 1.1 mm/s among healthy, middle-aged participants standing comfortably for one recommended data duration (4 × 30 s trials). A look up table provides values for alternate protocols that have been suggested in the literature and might prove relevant for clinical translation.

This work advances longitudinal assessment of individuals using quantified measures of postural control. Results enable practitioners/researchers to assess an individual's progress, maintenance, or decline relative to WSV at a defined CI level.