The neural basis of vestibular evoked myogenic potentials (VEMPs). The cVEMP is a specific indicator of saccularfunction.

Ian S Curthoys^1*Laura Fröhlich²Leigh McGarvie¹Leonardo Manzari³Shinichi Iwasaki⁴Julia Dlugaiczyk^5,6Gabor Michael Halmagyi⁷

• ¹The University of Sydney, Sydney, Australia
• ²Universitatsklinikum Bonn, Bonn, Germany
• ³MSA ENT Academy Center, Cassino, Italy
• ⁴Nagoya City University Gratuate School of Medicine, Nagoya, Japan
• ⁵University Hospital Zürich, Zurich, Switzerland
• ⁶Universitat Zurich, Zürich, Switzerland
• ⁷Royal Prince Alfred Hospital, Camperdown, Australia

VEMPs are widely used clinical vestibular tests and their interpretation is derived from the original proposal by Colebatch that the cVEMP is due to saccular activation by air conducted sound (ACS). This was based on previous extensive evidence that sounds selectively activate saccular afferent neurons and not semicircular canal neurons at clinical testing conditions. We revisit that earlier data and the results since. Despite that evidence, a recent partial review of the neural basis of cVEMPs has raised the possibility that canal afferents may be activated at clinical test frequencies and intensities and contribute to VEMPs, which would weaken their clinical specificity. This possibility is largely based on evidence that ACS clicks activate canal afferents in the rat – but not in the guinea pig. We show that result from the rat study is due to the very high sound pressure levels used - intensities which were far higher than those in the guinea pig study. When ACS stimuli of comparable intensity are used for both species at comparable clinically effective frequencies and intensities (~110dB SPL), otolithic neurons are activated in both species but canal activation by ACS clicks is negligible (and so most probably in humans also). Furthermore, the evidence from lesion and electrical stimulation studies and human clinical data complements the neural evidence and confirms the cVEMP to ACS is a specific test of saccular function. Audibility curves show rats have very high thresholds for frequencies most commonly used for human VEMP testing (500-1000Hz). That result questions the applicability of results of sound evoked vestibular responses from rat for understanding human vestibular function. Conclusion. There is a range of stimulus intensities and frequencies appropriate for measuring VEMPs. Using stimulus values within this range, neural evidence from rats and guinea pigs shows minimal contribution of canal afferents, meaning that VEMPs are specific tests of otolith function and that any contribution from semicircular canals to human cVEMPs (tones at 500Hz, clicks at 100 dB nHL) is negligible. Using stimulus values outside that range will stimulate semicircular canal afferents as well as otolith afferents, thus compromising the otolithic specificity of VEMPs.