Modulation of high-frequency vestibuloocular reflex during visual tracking in humans

1. Humans may visually track a moving object either when they are stationary or in motion. To investigate visual-vestibular interaction during both conditions, we compared horizontal smooth pursuit (SP) and active combined eye-head tracking (CEHT) of a larger moving sinusoidally at 0.4 Hz in four normal subjects while the subjects were either stationary or vibrated in yaw at 2.8 Hz. We also measured the visually enhanced vestibuloocular reflex (VVOR) during vibration in yaw at 2.8 Hz over a peak head velocity range of 5-40°/s. 2. We found that the gain of the VVOR at 2.8 Hz increased in all four subjects as peak head velocity increased (P < 0.001), with minimal phase changes, such that mean retinal image slip was held below 5°/s. However, no corresponding modulation in vestibuloocular reflex gain occurred with increasing peak head velocity during a control condition when subjects were rotated in darkness. 3. During both horizontal SP and CEHT, tracking gains were similar, and the mean slip speed of the target's image on the retina was held below 5.5°/s whether subjects were stationary or being vibrated at 2.8 Hz. During both horizontal SP and CEHT of target motion at 0.4 Hz, while subjects were vibrated in yaw, VVOR gain for the 2.8-Hz head rotations was similar to or higher than that achieved during fixation of a stationary target. This is in contrast to the decrease of VVOR gain that is reported while stationary subjects perform CEHT. 4. In a control experiment in which subjects carried out vertical SP and CEHT while they were vibrated in yaw at 2.8 Hz, we found that three of four subjects showed an increase in horizontal VVOR gain at 2.8 Hz compared with that achieved during fixation of a stationary target; such an increased horizontal gain would not be required to reduce retinal image slip in the vertical plane. 5. On the basis of these findings, we draw the following conclusions. 1) During sinusoidal oscillations at 2.8 Hz, the gain of the VVOR is adjusted in accordance with peak head velocity in order to hold retinal slip of the image of the visual target below ~5°/s. 2) During visual tracking of a moving target while the subject is in motion, there are two potential sources of retinal image slip: imperfect visual tracking and an inadequate VVOR. When tracking deteriorates, it becomes necessary to increase the gain of the VVOR to levels that prevent additional retinal image slip, so that vision is not compromised. 3) The increase of horizontal VVOR gain that occurs during both horizontal and vertical visual tracking while subjects are in motion may not be wholly due to retinal slip per se, but may also involve a nonvisual mechanism that effectively constrains retinal image slip to levels that permit clear vision.