Does thenumber of stepsneededfor UCM gait analysisdiffers between healthy and stroke?

The basis for the uncontrolled manifold (UCM) approach is the variability among repetitions of a motor task. Thus, reliable results might be influenced by the number of trials. This study estimated the number of steps needed for UCM analysis of stroke gait and if it is the same for healthy subjects. Twenty-five volunteers participated, sixteen in the stroke group (age 59.0 ± 7.5 years, ten hemiparesis at right), and nine in the healthy group (age 59.2 ± 4.9 years). We applied the UCM analysis over each lower limb's single support phase (SSP). The center of mass in the sagittal plane was the task variable, and the ankle, knee and hip joint angles, the elemental variables. The results obtained with 40 steps were used as a reference and compared with those obtained separately from 10, 20, and 30 steps. The mean values of the curves along the SSP were compared between the sets of steps. Further, for each volunteer, we calculated the Pearson correlation between the 40 steps curve and those obtained with other numbers of steps. Our results indicate that (1) the number of steps necessary to perform UCM analysis of stroke gait is larger than those necessary in healthy condition, (2) the synergy index is less sensitive to the number of steps than the UCM components (V_UCM and V_ORT), and (3) the analysis of the UCM over time requires a more significant number of steps than the mean values.