Effects of Stepping Training With Reduced Visual Cues in a Virtual Environment in Chronic Stroke Patients: A Randomized Controlled Trial

To take a step and appropriately place the foot in the environment is an ability that may be impaired after stroke. This study aimed to determine if a virtual reality (VR) stepping training with gradual removal of on-line visual cues about the feet or target would be superior to the same training with more visual cues for persons with chronic stroke.

Participants were randomly assigned to either a control group (CON, n=19) or an experimental group (EXP, n=19) to receive stepping training within a fully immersive VR environment. During practice, images of the feet and the target were presented to CON, while these images were gradually removed for EXP. The outcome measures included walking speed and gait symmetry ratio, error distance and variation during targeted stepping, and toe clearance and landing distance during obstacle crossing stepping.

For walking, there were no significant between- or within-group differences. For targeted stepping, EXP significantly reduced the unaffected limb error distance (p=0.013), but not CON (p=0.238). For obstacle crossing stepping, when the unaffected limb led, EXP significantly reduced the landing distance in both the affected and unaffected limb (p=0.002), but not CON (p=0.595).

Gradual removal of visual cues during stepping training in a VR environment resulted in better accuracy in the unaffected limb targeted stepping and in both limbs during the unaffected limb leading obstacle crossing. This type of training may be considered for inclusion in stroke rehabilitation, with careful consideration of the amount of visual cues provided.