Otolith and semicircular canal contributions to the human binocular response to roll oscillation

Three normal human subjects were oscillated about their naso-occipital axis in a supine position at 0.4 Hz and 0.1 Hz, both in darkness and in the light with a structured fixation target. The same subjects were oscillated in roll about an upright position, at the same frequencies, in darkness; and also about axes directed 20 degrees and 40 degrees to the left and to the right of the midsagittal plane, at 0.4 Hz, in darkness. Three-dimensional binocular eye movements were recorded using video-oculography. All stimuli induced a predominantly torsional nystagmus with small disconjugate head-vertical (skew) and conjugate head-horizontal components. For roll oscillation, the torsional slow phase velocity gain was higher in the light and generally increased with the stimulation frequency. In darkness, only one subject had significantly higher torsional gains in the upright compared to the supine position (12% difference), suggesting that the otolith contribution to the roll response is minimal at the frequencies tested. The slow phase velocity gain of the skew increased with fixation in two subjects. but decreased in the third subject; these changes were related to changes in eye vergence. In the upright position, with oblique axes of rotation, the head-vertical eye movements were asymmetrical, with the outermost eye executing a larger amplitude movement. The disconjugate head-vertical eye movements observed can be explained by the pattern of vertical semicircular canal stimulation and their connections to the extraocular muscles. In humans, skewing of the eyes may compensate for the eccentricity of the foveae which lie in the temporal retina and undergo vertical translations in opposite directions when the eyes tort.