UC Irvine

Vision contributes to the control of walking, balance regulation, orientation, and anticipatory control, which are important for spatial navigation. How the amount of visual information available, as distinct from its accuracy, shapes balance control during walking remains poorly understood. This study used virtual reality to determine how graded visual occlusion, imposed alongside visual and physical gait speed perturbations, affects gait and balance control during treadmill walking. Eight healthy adults (Age 20.6 ± 2.8 years) completed nine 5-minute walking trials on a split-belt treadmill in a custom virtual environment using a 3×3 design consisting of three walking conditions crossed with three levels of visual occlusion. During baseline walking, participants walked at 1.0 m/s with the optical flow matched to the treadmill speed. During optical flow trials, participants continued to walk at 1.0 m/s while we sinusoidally varied the relative translation speed of the visual environment between 0.5 to 2.0 times the treadmill speed. Finally, during speed modulation trials, belt speed varied sinusoidally between 0.7 m/s and 1.4 m/s while optical flow remained matched to treadmill speed.Gait performance was quantified using normalized mean step length and step width, their associated variabilities, margin of stability (MoS), and foot placement control. Mean step lengths and step width changed little across the nine trials while the step width variability increased with occlusion in every trial type by 6.0% during baseline walking, 9.1% during optical flow modulation, and 16.4% during speed modulation from the lowest to highest level of visual occlusion. Mean MoS during speed modulation trials was lower than during baseline walking and optical flow modulation. MoS variability was also substantially higher during speed modulation, exceeding that observed in the other two trial types by approximately 23.4% to 27.3%. Foot placement predictability increased across trials during baseline walking and speed modulation but remained relatively consistent during optical flow modulation trials. Each manipulation revealed different responses to increasing visual occlusion during regular and speed-perturbed walking. Greater visual occlusion was consistently associated with increased step width variability and stronger coupling between center of mass state and subsequent foot placement, whereas mean spatiotemporal measures remained relatively stable across conditions. These findings support graded visual manipulation in virtual reality as a method for assessing balance control and developing balance training protocols.