Training Blindfolded in a Virtual Environment with a Moving Platform: No Within-Session Gains but Robust Long-Term Gains in Task-Specific Postural Stability

Balance training in a virtual environment (VE) simulating travel on a bumpy road, induced large reductions in body sway within the training session and further reductions between-sessions [1]. These gains were well-retained over a 12-week interval. Here, to test how much this learning depended on visual input, we tested seven young adults who practiced maintaining balance in the same platform movements, but blindfolded. We hypothesized that the balance skill acquisition time-course would be invariant to eliminating visual input (sensory input reweighting). Center of Pressure displacements (CoP-d) were analyzed for each task iteration during the training session to assess “online” learning, and at 24 hours, 4 and 12-weeks post-training, to assess “offline” gains and retention. The results were compared to the published results obtained from participants training eyes-open, in the same VE conditions. Unlike the eye-open condition, CoP-d remained unchanged throughout the training session - no “online” stability gains. However, initial blindfolded performance equaled the lowest task-related CoP-d attained, at the end of the session, when training eyes-open. Nevertheless, blindfolded practice resulted in “offline” reductions in CoP-d by 4 and 12-weeks post-training. The results suggest that blindfolded balance performance may rely on pre-established balance skills that are independent of visual input and require little, within-session, adaptation. The initial, large stability gains reported in eyes-open training may reflect minimizing of redundant anticipatory postural adjustments to specific VE scenes. However, blindfolded training on a moving platform can lead to long-lasting balance gains, via a slow post-training, “offline”, process.