Evaluation of lower limb proprioception is usually done under static conditions. Sensory inputs are partially gated during movements, however. This gating might explain the lack of correlation between static proprioceptive testing and motor performance in clinical populations. The goal of the present study was therefore to determine if proprioception could reliably be estimated during actual walking.

Twenty healthy subjects walked on a treadmill while wearing a robotized ankle-foot orthosis (see Noel et al., 2008; 2 visits within 2–7days). Short-duration force perturbations (150ms; n=96) of graded amplitude (1–7Nm; producing ankle deviations of 1–12°) were applied during the swing phase of gait, once every 5–10strides. Participants pressed a hand-held button each time they felt a perturbation. To determine proprioceptive threshold, % detection was then plotted as a function of ankle deviations. A sigmoidal curve was fitted to individual subject data. Proprioceptive threshold was defined as the ankle deviation at which subjects detected 50% of the perturbations. Intraclass correlation coefficient (ICC) were then calculated to determine the test-retest reliability.

Median proprioceptive threshold was 5.25° (range 1.04°–9.14°; n=20). The ICC was 0.792, indicating a good reliability.

This pilot study suggests that a threshold for movement error detection can be reliably obtained during walking in healthy subjects. This proprioceptive threshold is higher than what is reported during static tasks, likely reflecting the sensory gating occurring during movement, and emphasizing the need to perform this assessment during walking. Further studies are now underway to determine if this method can be used as a tool for proprioception assessment in clinical populations.