Responses of primary vestibular neurons to galvanic vestibular stimulation (GVS) in the anaesthetised guinea pig

Previous studies in humans and animals which have shown that DC galvanic vestibular stimulation (GVS) induces horizontal and torsional eye movements have been interpreted as being due to a preferential activation of primary vestibular afferents innervating the horizontal semicircular canals and otoliths by GVS. The present study sought to determine in guinea pigs whether GVS does indeed selectively activate primary horizontal canal and otolith afferents. Constant-current GVS was passed between electrodes implanted in the tensor-tympani muscle of each middle ear or between electrodes on the skin over the mastoid. During this stimulation, responses from single primary vestibular neurons were recorded extracellularly by glass microelectrodes in Scarpa's ganglion. Afferents from all vestibular sensory regions were activated by both surface and tensor-tympani galvanic stimulation. Tensor tympani GVS was approximately 10 times more effective than surface GVS. At larger current intensities irregularly discharging afferents showed an asymmetrical response: cathodal stimulation resulted in a larger change in firing (increase) than anodal stimulation (decrease), whereas regularly discharging afferents responded symmetrically to the two polarities of GVS. Across all afferents tuned for different types of natural vestibular stimulation, neuronal sensitivity for GVS was found to increase with discharge variability (as indexed by CV*). Anterior canal afferents showed a slightly higher sensitivity than afferents from other vestibular sensory regions. Hence, the present study concluded that GVS activates primary vestibular afferents innervating all sensory regions in a uniform fashion. Therefore, the specific pattern of GVS-induced eye movements reported in previous studies are not due to differential sensitivity between different vestibular sensory regions, but are likely to reflect an involvement of central processing.