부산대학교 도서관

Postural adjustments are essential for balance control and to reduce risk of falling. One emerging method to train reactive postural control consists in exposing individuals to safe and controlled destabilizing perturbations that intend to simulate changing conditions that can lead to falls. Studies using virtual reality suggest that visual perturbations engage mechanisms of motor adaptation, increase electrocortical activity and modulate balance performance. What is not yet clear is the impact of trunk and limb muscles activation on the postural adjustments responsible to maintain balance control. This paper aims to map the response of trunk and limb muscles to visual perturbations, and compare them to those of physical perturbations. Additionally, our study includes vertical perturbations (i.e. balance disturbances in the vertical plane) known to be a major cause of falling. Therefore, this paper also compares muscles responses to both horizontal and vertical perturbations. Fourteen healthy participants (ten males; age: 27±4; BMI: 23.8±2.6 kg/m 2 ) stood on a moveable platform within a virtual reality system projecting visual scenes over a 360° dome-shaped screen such that the participant appeared to be standing in the middle of a room. Concomitantly, the electrical activity of tibialis anterior, gastrocnemius, rectus femoris, hamstring, rectus abdominis, paraspinal, external oblique and deltoid muscles was captured. Amid a larger protocol, this paper reports on randomly presented 1) visual perturbations; i.e. the virtual room moves during 0.35 seconds a distance corresponding to 14 cm in four directions (forward - FP, backward - BP, upward - UP, downward - DP), each repeated three times; and 2) physical perturbations (12cm displacement in one second) for the four directions and two sensory conditions: static camera (SC; virtual room remains static) and dynamic camera (DC; corresponding transitions in the visual scenery). We calculated three muscle activation parameters: onset latency, duration of activation, and magnitude. Separate 2-factor repeated-measures ANOVA were applied for each outcome measure across factors of perturbation direction (FP, BP, UP and DP) and condition (VIS, SC, DC). Forward visual perturbations led to longer onset latencies when compared to upward and downward visual perturbations (e.g. in the gastrocnemius: respectively, 443±56.6 ms vs. 326±39.6 ms and 334±51.1 ms, P