Abstract
Humans use multisensory information to maintain balance. Especially, vision plays an important role. A tilted visual frame can induce postural sway toward the frame tilt (Isableu et al., 1997, Exp. Brain Res), suggesting the influence of visually estimated direction of gravity on postural control. In this study, we systematically changed the frame tilt and measured the postural adjustment and subjective visual vertical. Healthy young adults participated in the postural measurement and the rod adjustment task. In the postural experiment, standing participants observed an 8 m × 8 m × 20 m virtual room for 10 s through a head mounted display (HMD). The room tilted between -42° and 42° around the sagittal axis. Participants’ head movement was collected by the HMD’s tracking system. Head displacements were averaged for the mediolateral head positions relative to the initial position for each trial. In the rod adjustment task, the same visual stimuli were presented as the postural experiment, while seated participants were asked to manually adjust a rod at the room center to be gravitationally vertical. Subjective visual verticals were defined as the angular errors of the adjusted rod from the gravitational vertical. Consistent with the previous reports, we found biased head displacement toward the room tilt. The amount of the head displacement changed non-linearly as a function of the room tilt. The head displacement was the highest around the 20° room tilt, returning to the baseline as the room tilt further increased. The subjective visual vertical in the rod task showed consistent changes with the head displacement, with the maximum bias at around 20° room tilt. The weighted vector sum model (Mittelstaedt, 1983), can account for both the head displacement and subjective visual vertical, suggesting shared representation of frame of reference between perception and motor control as to the gravity information.