Article Text

Download PDFPDF

Ocular vestibular evoked myogenic potentials in superior canal dehiscence
  1. S M Rosengren1,
  2. S T Aw2,
  3. G M Halmagyi2,
  4. N P McAngus Todd3,
  5. J G Colebatch1
  1. 1
    UNSW Clinical School and Prince of Wales Medical Research Institute, Randwick, Sydney, NSW, Australia
  2. 2
    Neurology Department, Royal Prince Alfred Hospital, Camperdown, NSW, Australia
  3. 3
    Faculty of Life Science, University of Manchester, Manchester, UK
  1. S M Rosengren, Institute of Neurological Sciences, Prince of Wales Hospital, Randwick, NSW 2031, Australia; s.rosengren{at}


Objective: Patients with superior canal dehiscence (SCD) have large sound-evoked vestibular reflexes with pathologically low threshold. We wished to determine whether a recently discovered measure of the vestibulo-ocular reflex—the ocular vestibular evoked myogenic potential (OVEMP)—produced similar high-amplitude, low-threshold responses in SCD, and could differentiate patients with SCD from normal control patients.

Methods: Nine patients with CT-confirmed SCD and 10 normal controls were stimulated with 500 Hz, 2 ms tone bursts and 0.1 ms clicks at intensities up to 142 dB peak SPL. Conventional VEMPs were recorded from the ipsilateral sternocleidomastoid muscle to determine threshold, and OVEMPs were recorded from electrode pairs placed superior and inferior to the eyes. Three-dimensional eye movements were measured with scleral dual-search coils.

Results: In patients with SCD, OVEMP amplitudes were significantly larger than normal (p<0.001) and thresholds were pathologically low. The n10 OVEMP in the contralateral inferior electrode became particularly large with increasing stimulus intensity (up to 25 μV) and with up-gaze (up to 40 μV). Sound-evoked (slow-phase) eye movements were present in all patients with SCD (vertical: upward; torsional: upper pole away from the affected side; and horizontal: towards or away from the affected side), but began only as the OVEMP response became maximal, which is consistent with the surface potentials being produced by activation of the extraocular muscles that generated the eye movements.

Conclusions: OVEMP amplitude and threshold (particularly the contralateral inferior n10 response) differentiated patients with SCD from normal controls. Our findings suggest that both the OVEMPs and induced eye movements in SCD are a result of intense saccular activation in addition to superior canal stimulation.

Statistics from

Superior canal dehiscence (SCD) is a recently described condition in which a dehiscence (defect) in the bone overlying the superior semicircular canal leads to vestibular hypersensitivity to sound and pressure.1 2 Patients with SCD have pressure- and sound-induced vestibular symptoms (Tullio phenomenon), and a characteristic feature of SCD is large sound-evoked vestibular evoked myogenic potentials (VEMPs) with low threshold.3 4

A new test of vestibular function—the ocular vestibular evoked myogenic potential (OVEMP)—has also recently been described, in which sound-evoked, vestibular-dependent myogenic responses can be recorded from near the eyes, similar to the VEMP in the neck.5 In this test, distinct responses, believed to originate from extraocular muscle EMG, can be recorded from sites superior and inferior to each eye in normal subjects following unilateral, short-duration acoustic stimulation. As air-conducted (AC) sound is known to selectively activate the saccule,6 7 AC-evoked VEMPs and OVEMPs are therefore both arguably measures of saccular reflex function. Although VEMPs test the integrity of the sacculo-collic pathway, OVEMPs test the sacculo-ocular pathway and add to the available range of vestibular tests.5

In the current paper, we provide evidence of reduced thresholds and enhanced OVEMP amplitudes in patients with SCD, the latter separating patients from normal controls more clearly than VEMP amplitudes. We also determined that, in patients with SCD, OVEMPs do not reflect changes in the corneo-retinal dipole. They represent changes in extraocular muscle activation that are required to produce the eye movements. Our findings support a combined saccular and superior canal contribution to short-latency vestibular-ocular reflexes in SCD.


The study was conducted in three parts. First, standard VEMPs were recorded using 2 ms tone bursts to determine VEMP threshold. In the second part, OVEMPs were recorded using two types of sound: 2 ms tone bursts delivered at fixed intensity levels relative to each subject’s individual VEMP threshold, and 0.1 ms high-intensity clicks. In the third part, binocular 3D eye movements were measured using magnetic search coils in response to similar 0.1 ms clicks. The sound stimulus intensities used were high and, although none of our subjects had ill effects, similar studies should always use the lowest effective intensities and follow relevant national guidelines.


Nine patients with superior canal dehiscence participated (3 females, 6 males; mean age 51 years, range 40–63 years). All patients had dehiscence of the bone overlying the superior semicircular canal on high-resolution CT imaging reformatted off-line into the superior canal plane.8 9 In 1 patient, the dehiscence was bilateral, and in all others it was unilateral (5 right-sided, 3 left-sided), giving 10 affected ears. VEMP and OVEMP data from 10 normal controls, reported previously in detail, were used for comparison (7 females, 3 males; aged 24–42 years).5 The controls were selected to have low, but normal, VEMP thresholds. All participants gave informed consent and the study was approved by the local ethics committees.

Vestibular evoked myogenic potentials

Subjects lay supine on a chair reclined to approximately 30 degrees above the horizontal and lifted their heads to activate the sternocleidomastoid (SCM) muscles. VEMP recordings were made in response to 500 Hz, 2 ms tone bursts delivered via calibrated headphones (TDH 49; Telephonics, Farmingdale, USA). VEMP threshold was measured by systematically reducing the intensity in 3 dB steps, and was defined as the lowest stimulus intensity that produced a reliable response on at least two of three trials. Recording and reference electrodes were placed on the middle of the SCM belly and medial clavicle, respectively. Stimuli were generated by customised software (1401plus; Cambridge Electronic Design, Cambridge, UK). The maximum tone burst amplitude was 10 V peak to peak (pp) (equivalent to 142 dB SPL peak). A total of 256 stimuli were delivered at a rate of 5 Hz. Amplification and averaging were performed using previously reported techniques.10 Peak-to-peak amplitudes were divided by the mean rectified EMG over the pre-stimulus interval (“corrected value”).

Ocular vestibular evoked myogenic potentials

OVEMPs were recorded in all patients and normal controls in response to the 500 Hz, 2 ms AC tone bursts described above. The subjects reclined as above and directed their gaze straight ahead to a target situated ∼2 m away. Stimuli were delivered at a rate of 6 Hz for 200–500 trials. Stimulus intensity was adjusted to each subject’s VEMP threshold. All subjects were stimulated at 3 dB below and 18 dB above VEMP threshold, and the patients with SCD were additionally stimulated with higher intensities in steps of 6 dB. The maximum intensity depended on VEMP threshold, meaning that different numbers of affected ears were stimulated at the highest intensity levels (at 18 dB, 24 dB and 36 dB above threshold, n = 10 (all subjects); at 42 dB n = 5 and at 48 dB, n = 2). OVEMPs were recorded in a subset of 8 patients with SCD in response to 0.1 ms clicks at a fixed high intensity of 142 dB peak SPL to compare OVEMP responses with eye movements measured separately using the same stimulus.

Surface potentials were recorded using pairs of 9 mm Ag/AgCl electrodes. For the tone burst stimuli, the superior active (inverting) electrode was placed on the orbital margin between the eye and eyebrow and referred to an electrode directly above the eyebrow. The inferior active electrode was placed on the orbital margin below the eye and referred to an electrode approximately 15 mm below it on the cheek. The electrodes thus formed a vertical line down each side of the face, with the earth placed on the sternum. For the 0.1 ms click stimuli, all electrodes described above were “active” and were referred to an electrode placed on the clavicle. Using this referential montage, the OVEMP bipolar (differential) montage and a traditional EOG montage were constructed offline by subtracting pairs of electrodes. In three patients, an additional recording was made during up-gaze using 0.1 ms clicks. The potentials were sampled at 10 kHz for 60 ms, from 10 ms before to 50 ms following stimulus onset. Amplification and averaging techniques were reported previously.5 Negative potentials at the active electrodes were displayed as upward deflections.

Eye movements

Eye movements were recorded in 8 patients with SCD (9 ears), and data from 7 patients was previously reported by Aw et al.9 Scleral dual-search coils (Skalar, Delft, The Netherlands) were used to measure three-dimensional eye movements in a two-axis rotator.11 Each patient lay supine and was secured to the rotator with belts, and the patient’s head was held with an individually moulded thermoplastic full-face mask (Polyflex II Roylan; Smith and Nephew, Warwick, UK ) bolted to the head holder. Pre-calibrated dual-search coils were placed onto both of the patient’s eyes after application of topical anaesthesia (Alcaine 0.5%; Alcon, Sydney, Australia ). During the recording, the patient was in darkness and viewed a 2 mm fixation LED located 600 mm away. Thirty 0.1 ms clicks at 145 dB peak SPL were presented at a rate of 5 Hz using headphones (TDH 49). Binocular eye positions were recorded in three axes—horizontal (z), vertical (y) and torsional (x). Eye positions in space-fixed coordinates expressed as rotation vectors were computed in 3D from the dual-search coil signals and converted to Euler’s angles. The trials were averaged and the epochs reduced to 10 ms before and 50 ms after the stimulus. Angular positions and amplitude of response were expressed in millidegrees (mdeg). Additional details of the recording technique have been reported.9

Data analysis

Data from all patients and controls were re-coded in terms of responses ipsilateral and contralateral to the side of the dehiscence, allowing the results from all patients to be averaged together, regardless of the side of the dehiscence. The normal control data from both ears were averaged in the same manner. Similarly, the direction of eye movement in the horizontal and torsional planes was reported in terms of movement towards or away from the affected/stimulated side. Positive potentials corresponded to horizontal movement away from the side of stimulation, upward vertical movement and torsional rotation of the upper pole away from the side of stimulation. Latencies and amplitudes of all responses were measured at the first response peak after removal of any amplifier DC offset. For the OVEMPs, DC offset was calculated using the entire pre-stimulus interval, and for eye movements, it was calculated over the period 0–5 ms after stimulus onset (before any response), as there was slow drift in several patients. Values reported in the text are mean ±SD and in the graphs mean ±SEM. The SCD and control groups were compared using between-subjects t-tests (two-tailed) and ANOVAs, and within-group analyses were made with repeated measures tests. Where post-hoc t-tests were performed, the Bonferroni procedure was used to correct for multiple comparisons. Linear regression was used to assess OVEMP response “gain” (defined as the slope of the relationship between amplitude and stimulus intensity). All statistical procedures were performed using SPSS statistical software version 14.


Vestibular evoked myogenic potentials

All 9 patients with SCD had large VEMP responses with low threshold following stimulation of the affected ear/s with 2 ms tone bursts. The mean threshold in the patients with SCD was 101.2±5.5 dB peak SPL and in the normal controls (chosen for their low normal thresholds) was 116.7±4.7 dB peak SPL. The difference in threshold of ∼15 dB peak SPL between the groups was highly significant (p<0.001). In the subjects stimulated at the maximum intensity of 142 dB peak SPL (SCD, n = 9; controls, n = 11), there was no significant difference in mean VEMP amplitude (SCD, 2.9 ±0.5; controls, 3.0±0.8 corrected peak-to-peak value; p = 0.840; fig 1A).

Figure 1 (A) Vestibular evoked myogenic potential (VEMP) amplitudes from the ipsilateral sternocleidomastoid muscle in patients with superior canal dehiscence (SCD) and normal controls at 142 dB peak SPL. Values are corrected peak-to-peak amplitudes, which correct for differences in background muscle contraction. The VEMP amplitudes of both groups overlapped. (B) Ocular vestibular evoked myogenic potential (OVEMP) contralateral inferior n10 amplitudes in patients with SCD and normal controls at two intensity levels (139–142 and 133–136 dB peak SPL). There is little overlap in amplitude between the groups, suggesting that an n10 amplitude of more than 5 μV may be diagnostic for SCD (assuming similar stimulation and recording conditions to the present study are used).

Ocular vestibular evoked myogenic potentials

2 ms tone burst stimuli

In all patients with SCD (9 patients, 10 ears), OVEMP responses were present following stimulation of the affected ear/s, and consisted of the p8, n10 and n13 responses previously described by Todd et al.5 for normal subjects (fig 2A). When the SCD and control groups were compared at a fixed high intensity of 139–142 dB peak SPL (ie, at or close to maximal output of our system), the patients with SCD had much larger OVEMPs (p<0.001; table 1). Of the five OVEMP potentials, the contralateral inferior n10 was largest; however, this was significant for the patients with SCD (p = 0.001 to 0.02) but not the controls. There was a particularly marked difference between the amplitude of the n10 potential in the SCD and control groups at maximal intensity (11.1±6.8 μV vs 1.9±1.4 μV; p = 0.004, equal variances not assumed). There was some overlap in the n10 amplitude ranges of the two groups at this intensity (SCD patients, 3.3–25.4 μV; controls, 0.4–4.9 μV); however, 7 of 10 SCD cases, but no controls, had response amplitudes well over 5 μV (fig 1B). The findings were similar using a lower intensity of 133–136 dB peak SPL (SCD patients, 2.1–24.7 μV; controls, 0.5–4.8 μV; fig 1B). In contrast, when the SCD and control groups were matched in terms of intensity level (ie, both stimulated at 18 dB above threshold), the OVEMP amplitudes were not significantly different between the two groups (p = 0.207; fig 2A, table 1). There were no significant differences in latency between the SCD and normal responses at either intensity (p = 0.113, 114).

Figure 2 (A) Mean ocular vestibular evoked myogenic potentials (OVEMPs) recorded from 9 patients with superior canal dehiscence (SCD) (10 ears) and 10 normal controls in response to stimulation with 500 Hz, 2 ms tone bursts. All figures depicting group mean OVEMPs are displayed as if the viewer is looking at the patient’s face and the patient is stimulated in the right ear. Electrode positions were superior or inferior to the eyes (sup, inf) and ipsilateral or contralateral to the side of stimulation/dehiscence (Ipsi, Contra). The normal controls (black) were stimulated at 18 dB above VEMP threshold, which corresponded to within 5 dB of maximal stimulation (max.). The SCD patients were also stimulated at 18 dB above threshold (light grey), as well as within 3 dB of the maximal intensity level (dark grey; corresponding to 30–48 dB above their individual VEMP thresholds). Negative potentials are shown as upward deflections. The SCD responses were similar to normal responses at matched intensity levels with respect to VEMP threshold, but were much larger with maximal stimulation. (B) OVEMPs recorded from a single patient (right SCD) in response to stimulation with 2 ms tone bursts at fixed levels relative to his VEMP threshold (ie, at 3 dB below threshold, and in steps of 6 dB above threshold from 18–42 dB above threshold). The OVEMPs became very large with increasing intensity, particularly the n10 in the contralateral inferior electrode, which in this patient reached 25.4 μV. Note: parts A and B have different gains.
Table 1 Amplitudes and latencies of OVEMPs recorded in 9 patients with SCD and 10 normal controls in response to stimulation with 500 Hz 2 ms tone bursts

In the patients with SCD, the OVEMPs became very large with increasing stimulus intensity, up to 25.4 μV in one patient in the contralateral inferior electrode at 42 dB above VEMP threshold (fig 2B). The increase in OVEMP size with increasing intensity was linear for each potential (mean r2  =  0.84–0.96). However, there was a significant difference in “gain” (or slope) between the responses (fig 3; p<0.001). The contralateral inferior n10 tended to have a greater gain than the other responses, although this was significant compared with the contralateral superior p8 only (2.2 vs 0.4 μV/dB, respectively; p = 0.001). OVEMP thresholds were similar to VEMP thresholds, as OVEMPs were always absent at 3 dB below VEMP threshold (fig 4).

Figure 3 The effect of increasing stimulus intensity on response amplitude for all five ocular vestibular evoked myogenic potential (OVEMP) responses. The mean amplitudes from all available subjects in response to 500 Hz, 2 ms tone burst stimulation are shown, and error bars are standard error of the mean (SEM). At intensities +18 to +36 dB, n = 10, whereas at +42 dB, n = 5. The slope (ie, gain) of the contralateral inferior n10 was higher than the slope of the other potentials (this remained the case when only the 5 subjects tested at each intensity level were considered). Contra, contralateral; inf, inferior; ipsi, ipsilateral to the stimulated side; sup, superior electrode position.
Figure 4 Vestibular evoked myogenic potentials (VEMPs) and ocular vestibular evoked myogenic potentials (OVEMPs) recorded from a single patient (right superior canal dehiscence (SCD), same as in fig 3) in response to stimulation with 500 Hz, 2 ms tone bursts at fixed levels relative to his VEMP threshold (3 dB below threshold (–3) to 42 dB above threshold (+42)). The VEMPs were recorded in the sternocleidomastoid muscle ipsilateral to the side of stimulation/dehiscence, whereas the OVEMPs were recorded beneath the contralateral eye (contralateral inferior n10 response). Note that the calibration bars and time axes are different for the two responses as conventional recording conditions were used. Negative potentials are shown as upward deflections. The responses had similar thresholds and increased linearly with increasing stimulus intensity.

0.1 ms click stimuli

OVEMPs were also present in the 8 patients with SCD (9 ears) stimulated with clicks. Using the referential montage, initial negative potentials were recorded in all electrodes, apart from the contralateral superior electrode, in which a small positivity was present prior to the negativity (fig 5A; see also Todd et al.5). The bipolar OVEMP montage constructed offline produced OVEMPs similar to those reported above (fig 5B). A traditional EOG montage was also constructed by subtracting the electrodes directly above and below each eye, and demonstrated a marked asymmetry in the size of the response between the eyes, with the contralateral response being significantly larger (fig 5C; 23.5±12.8 μV vs. 2.7±1.2 μV; p = 0.005).

Figure 5 Mean extraocular and eye movement responses recorded from 8 patients with superior canal dehiscence (SCD) (9 ears) in response to stimulation with 0.1 ms clicks. (A) Extraocular responses were recorded using a referential electrode montage, with active electrodes at two positions above and below each eye (ie, superior and inferior to each eye (sup, inf) with upper and lower electrode positions (U, L)), and with a reference electrode on the sternum. Stimulus intensity was 142 dB peak SPL. (B) A bipolar ocular vestibular evoked myogenic potential (OVEMP) montage was created offline from the referential traces by subtracting the upper from the lower traces superior to the eye and the lower from the upper traces inferior to the eye. Typical OVEMP responses can be seen. (C) An EOG montage was created by subtracting the inferior upper referential responses from the superior lower responses. Note that negative potentials are depicted as upward deflections, whereas the conventional EOG montage is negative down. The contralateral response was much larger than the ipsilateral response. (D) Mean eye movements were recorded in the same patients on a separate occasion following stimulation with 0.1 ms clicks at 145 dB peak SPL. The movements were conjugate for all three directions. The vertical (dark grey) and torsional (light grey) movements were the same in each patient: the eyes moved up and the upper pole of the eye rotated away from the side of the dehiscence. However, in 6 patients, the horizontal component of the movement was away from the side of the dehiscence and, in 2 patients (3 ears), it was towards the side of the dehiscence (shown as two separate traces).

OVEMPs recorded in three patients with SCD in response to up-gaze with clicks were larger than with neutral gaze, but the enhancement occurred for the contralateral inferior n10 only (fig 6). The n10 response increased in size by 60–276%, and in one patient reached 40.7 μV (61.7 μV with the referential electrode montage).

Figure 6 Ocular vestibular evoked myogenic potentials (OVEMPs) from a single subject with right superior canal dehiscence (SCD) (same as in fig 3 and 4) during neutral- (grey) and up-gaze (black). The OVEMPs were elicited by 0.1 ms clicks at 142 dB peak SPL. The contralateral inferior n10 response became very large with up-gaze, whereas the other responses remained similar. Contra, contralateral; inf, inferior; ipsi, ipsilateral to the stimulated side; sup, superior electrode position.

Eye movements

The 0.1 ms clicks produced transient eye movements in all 8 patients, analogous to the slow-phase component of nystagmus (9 ears; fig 5D). In the vertical and torsional directions, the movements were conjugate and the same in each patient. The eyes moved upwards and the upper pole rotated away from the affected/stimulated side. In contrast, whereas the eye movements in the horizontal plane were always conjugate, they were not uniform in direction across patients. In 6 patients, both eyes moved away from the affected side, whereas both eyes moved towards the affected side in 2 patients (3 ears). This difference was not related to the laterality of the dehiscence. The peak amplitudes and latencies and onset latencies of the eye movements are shown in table 2. Horizontal eye movements were significantly smaller than vertical and torsional movements (p<0.001); however, there were no differences between ipsi- and contralateral eye movements or in response latency.

Table 2 Amplitudes and latencies of eye movements recorded in 8 patients with SCD (9 ears) in response to stimulation with 0.1 ms clicks

Eye movements versus EOG responses

Both the morphology and timing of the eye movements were different from the EOG-montage responses (fig 5). The responses produced by the EOG montage began significantly earlier than the eye movements (6.6±0.5 ms vs. 10.4±0.5 ms, respectively; p<0.001), and also peaked significantly earlier (8.7±0.3 ms vs. 20.6±1.1 ms; p<0.001). The eyes were therefore beginning to move only after the EOG (and the OVEMP) responses became maximal.


We have shown that, in patients with SCD, OVEMPs have normal latency and polarity but are characterised by a high amplitude and pathologically low threshold to AC sound, analogous to the conventional VEMP. Patients with SCD are known to have characteristically low VEMP thresholds,4 12 as well as low click-evoked eye movement thresholds.9 The present study has shown that the OVEMP threshold in SCD is similarly reduced, making it an alternate method of diagnosis. However, the amplitude of the OVEMP n10 response may additionally separate SCD from normal responses. Although VEMPs in SCD are typically large, there is significant overlap between SCD and normal amplitudes.12 In contrast, the largest SCD n10 amplitude was approximately 5 times higher than the largest normal n10 amplitude. We suggest that, under similar stimulation and recording conditions to the present study, an n10 amplitude of approximately 3 μV raises the possibility of SCD, and an amplitude of more than 5 μV is virtually diagnostic (ie, neutral gaze; 500 Hz, 2 ms tone bursts; at or above 133 dB peak SPL). A single OVEMP recording could therefore potentially be diagnostic of SCD and may not require a confirmatory threshold estimation.

The difference in amplitude between the VEMP and n10 OVEMP responses may be due to the differing underlying excitability changes. The VEMP is mediated by an inhibitory projection to the ipsilateral SCM.13 In contrast, OVEMPs can be recorded from both eyes, although the contralateral projection appears to be dominant, particularly in SCD patients. Colebatch and Rothwell14 demonstrated that, for the VEMP, an initial surface positivity corresponded to an inhibition of the underlying SCM muscle. As OVEMPs are believed to have a similar origin to VEMPs, it is likely that they may also be interpreted as excitation and inhibition of the extraocular muscles that are close to each electrode site. The contralateral inferior n10 response may therefore better distinguish between SCD and normal controls because it is excitatory and may not saturate as readily as the inhibitory p13 of the VEMP. Indeed, the n10 becomes even larger with up-gaze (up to ∼40 μV), a feature which may be clinically useful to improve OVEMP recordings.

Given the large n10 OVEMP response in the contralateral inferior electrode and the direction of the eye movement in normal subjects, Todd et al.5 proposed that the contralateral inferior oblique muscle had a role in producing the eye movement. The response to up-gaze and the induced extorsion of the contralateral eye in patients with SCD in this study are also consistent with inferior oblique activity. The short-duration AC sounds produced transient eye movements in all patients, analogous to the slow-phase component of nystagmus produced by long-duration acoustic stimulation.15 The eyes moved upwards and the upper pole rotated away from the affected ear, whereas the horizontal component was small and inconsistent between patients. Similar acoustic stimulation in normal subjects produces eye movements with the same direction, but with around 1/30th of the magnitude.9 As AC sound produces both surface potentials and eye movements in SCD, it could be argued that OVEMPs are merely a recording of the corneo-retinal dipole shift produced by the eye movement. Previous EOG type recordings of sound-evoked responses in patients with SCD demonstrated a large short-latency potential, which suggested detection of an eye movement.16 However, whereas the polarity of the EOG response in patients with SCD would suggest an upward eye movement, the surface potentials reach their peak before the eyes begin to move (fig 5; see also Todd et al5). In the patients with SCD, the mean peak vertical movement was ∼50–60 mdeg. Assuming a calibration of 10 μV/deg, these movements would be expected to produce an EOG signal of about 0.5–0.6 μV. Given the magnitude of the observed responses (∼23.5 μV), it is clear that displacement of the corneo-retinal dipole does not contribute significantly. Although EOG recordings in this context cannot be interpreted as an eye movement, they are still capable of differentiating patients with SCD from normal subjects.16 However, as OVEMPs reflect the extraocular muscle activity that generates the observed transient sound-evoked eye movements, this type of montage should now become the preferred method for recording extraocular muscle potentials.

It is generally accepted that both the sound-evoked nystagmus and short-duration eye movements in SCD are generated by the superior semicircular canal, as they align with the superior canal plane.1 15 However, it is less clear which end-organ/s are responsible for producing the enhanced reflexes in the neck and extraocular muscles. Sound-evoked VEMPs and OVEMPs in normal subjects are thought to be mediated by the saccule, as animal studies have shown selective activation of the saccule in response to AC sound.6 7 17 The enhanced VEMP in SCD is also thought to be predominantly due to increased saccular stimulation; however, there is evidence of additional receptor involvement. With increasing sound intensity in SCD, the response in the ipsilateral SCM increases in size and a crossed excitatory response appears in the contralateral SCM.12 Animal research has shown an inhibitory projection to the ipsilateral SCM mediated by the saccule, but no excitatory projection to the contralateral SCM.13 In contrast, inhibitory ipsilateral and excitatory contralateral projections to the SCM have been reported for the utricle13 and the three semicircular canals.18 19 Given that the dehiscence in this condition is in the temporal bone overlying the superior canal, whose afferents are activated at short latency,20 it is likely that activation of the superior canal contributes to the enhanced vestibulocollic reflexes.

Similar considerations apply to the n10 component of the OVEMP. The saccule excites the superior recti and inhibits the inferior recti bilaterally,21 and may have a weak connection to the oblique muscles.21 22 In contrast, although the superior canal also projects to the vertical recti, it additionally excites the contralateral inferior oblique,23 thus potentially contributing to the selective enhancement of the n10 response. Stimulation of both sets of afferents would produce upwards eye movements, but the canal additionally evokes distinct torsional movements.24 Both receptors may also contribute to the eye movements in SCD, with the predominant effect being due to the superior canal. Evidence for the involvement of more than just the superior canal comes from Aw et al.,9 who showed that, although the ipsilateral sound-evoked eye movements in SCD were aligned closely with the superior canal, the contralateral eye movements showed significant differences.

As both VEMPs and OVEMPs in normal subjects rely on the integrity of the saccule, OVEMPs can be considered to be an additional test of saccular function. However, OVEMPs and VEMPs test different vestibular pathways: VEMPs assess the vestibulo-collic pathway, whereas OVEMPs test the vestibulo-ocular pathway, a distinction that is relevant when central disorders of vestibular function are the subject of investigation. OVEMPs in patients with SCD are characterised by a large amplitude and low threshold, similar to the well-established VEMPs. Both the OVEMPs and the induced eye movements in SCD are likely to be the result of the combination of intense saccular stimulation and excitation of superior canal afferents.


This research is supported by a grant from the National Health and Medical Research Council of Australia, a UNSW Goldstar Award, the Garnett Passe and Rodney Williams Memorial Foundation and the UK Royal Society.



  • Competing interests: None declared.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.