Limited transfer and retention of locomotor adaptations from virtual reality obstacle avoidance to the physical world

Abstract

AbstractLocomotor training based in virtual reality (VR) is promising for motor skill learning, with transfer of VR skills in turn required to benefit daily life locomotion. This study aimed to assess whether VR-adapted obstacle avoidance can be transferred to a physical obstacle and whether such transfer is retained after 1 week. Thirty-two young adults were randomly divided between two groups. A control group (CG) merely walked on a treadmill and an intervention group (IG) trained crossing 50 suddenly-appearing virtual obstacles. Both groups crossed three physical obstacles (transfer task) immediately after training (T1) and 1 week later (T2, transfer retention). Repeated practice in VR led to a decrease in toe clearance along with greater ankle plantarflexion and knee extension. IG participants crossed physical obstacles with a lower toe clearance compared to CG but revealed significantly higher values compared to the VR condition. VR adaptation was fully retained over 1 week. For physical obstacle avoidance there were differences between toe clearance of the third obstacle at T1 and the first obstacle at T2, indicating only partial transfer retention. We suggest that perception–action coupling, and thus sensorimotor coordination, may differ between VR and the physical world, potentially limiting retained transfer between conditions.