Ocular and cervical vestibular evoked myogenic potentials produced by air- and bone-conducted stimuli: Comparative properties and effects of age

Clinical Neurophysiology 122 (2011) 2282–2289

Contents lists available at ScienceDirect

Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph

Ocular and cervical vestibular evoked myogenic potentials produced by air- and bone-conducted stimuli: Comparative properties and effects of age Sally M. Rosengren ⇑, Sendhil Govender, James G. Colebatch Prince of Wales Clinical School and Neuroscience Research Australia, University of New South Wales, Randwick, Sydney, NSW 2031, Australia

a r t i c l e

i n f o

Article history: Accepted 6 April 2011 Available online 6 May 2011 Keywords: Vestibular oVEMP VEMP Vibration Age Sound Gender

h i g h l i g h t s  oVEMPs and cVEMPs evoked by air-conducted sounds, bone-conducted vibration and lateral accelerations were compared in 61 normal subjects.  oVEMP response rates varied widely (59–96%) for the different stimulus modalities, while cVEMP rates were more consistent (91–100%).  Significant declines with age were present only for the AC stimuli and BC mastoid 500 Hz vibration, but not forehead taps or lateral accelerations.

a b s t r a c t Objective: To compare amplitudes, latencies, symmetry and the effects of age for both ocular and cervical vestibular evoked myogenic potentials (oVEMPs and cVEMPs) produced by different types of air- (AC) and bone-conducted (BC) stimuli. Methods: Sixty-one normal subjects aged 18–80 years participated. Both reflexes were recorded in response to AC clicks, AC and BC 500 Hz tone bursts, forehead taps and lateral mastoid accelerations. Results: AC tone bursts, clicks and BC tone bursts evoked oVEMPs in 81%, 59% and 65% of ears, respectively. The AC stimuli had higher thresholds for oVEMPs than for cVEMPs and all three stimuli produced higher asymmetry for the oVEMP than for the cVEMP. Forehead taps and lateral pulses evoked oVEMPs in 96% and 92% of cases. AC click- and BC tone burst-evoked oVEMPs showed a significant decline with age. Conclusions: AC stimulation and BC tone bursts delivered to the mastoid are less effective in evoking oVEMPs than in evoking cVEMPs, have high degrees of asymmetry in normals and appear to decline with age. Forehead taps and lateral accelerations produce more symmetrical effects and showed no significant decline with age. Significance: Stimulus properties need to be considered when deciding the most appropriate way to investigate vestibular function using oVEMPs. Ó 2011 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction Vestibular evoked myogenic potentials (VEMPs) have become an accepted test of vestibular function. Recorded from the sternocleidomastoid (SCM) muscles in the neck (cervical VEMPs, cVEMPs), they are a measure of the short-latency vestibulo-collic reflex (Colebatch et al., 1994). In recent years VEMPs recorded from the extraocular muscles (ocular VEMPs, oVEMPs) have also gained interest in the field of vestibular neurophysiology. Although they are measured by surface electrodes placed near the eyes, they are not eye movements (i.e. shift of the corneoretinal dipole), but ⇑ Corresponding author. Address: Neurologische Klinik, UniversitätsSpital Zürich, Frauenklinikstrasse 26, 8091 Zürich, Switzerland. Tel.: +41 44 255 3996; fax: +41 44 255 4533. E-mail address: [email protected] (S.M. Rosengren).

represent the activity of the eye muscles preceding an eye movement (Todd et al., 2007). Although cVEMPs and oVEMPs share many similarities, there are also significant differences between the reflexes. While the dominant cVEMP is seen in the ipsilateral SCM and has an initial positive (inhibitory) peak at approximately 13 ms (i.e. the p13 or p1; Colebatch et al., 1994), the primary oVEMP projection appears to be crossed and is usually an initial negative peak at about 10 ms (i.e. n10 or n1). For both reflexes bilateral responses are sometimes present, due to a bilateral projection and/or because both sides are stimulated simultaneously. While the myogenic origin of the cVEMP is known to be the SCM muscle, the extraocular muscles involved in the oVEMP are less clear, but are thought to usually reflect inferior oblique muscle activity (Rosengren et al., 2005; Iwasaki et al., 2007; Govender et al., 2009).

1388-2457/$36.00 Ó 2011 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2011.04.001

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289

There is also uncertainty concerning which vestibular end organ is responsible for each reflex under different stimulation conditions. Many animal studies suggest that saccular afferents are activated by air-conducted (AC) sound (McCue and Guinan, 1994; Young et al., 1977), although some utricular afferents will likely also be activated (Murofushi and Curthoys, 1997). Similarly, irregular afferents from both otoliths are sensitive to bone-conducted (BC) vibration at 500 Hz (Curthoys et al., 2006). However, the relative sensitivities of the saccule and utricle to sound and vibration have not been determined. Human studies support both cVEMPs and oVEMPs evoked by different forms of BC stimulation being mediated at least in part by utricular afferents (i.e. including forehead 500 Hz vibration and taps as well as taps delivered to the mastoid: Brantberg and Mathiesen, 2004; Govender et al., 2011; Iwasaki et al., 2009; Manzari et al., 2010). The cVEMP evoked by AC sound appears to be mostly saccular in origin, with a contribution from other (probably utricular) afferents (Welgampola and Colebatch, 2001). By contrast, the AC oVEMP follows a similar pattern to the BC oVEMP in vestibular neuritis, suggesting that it is also mediated by afferents coursing through the superior nerve (Curthoys et al., 2011; Govender et al., 2011). Although there has been recent debate about the origins of the cVEMP and oVEMP to AC and BC stimulation, there has been relatively little comparison of the reflexes evoked using different types of stimulation. Many studies use one stimulus for the cVEMP (usually AC sound) and another for the oVEMP (usually BC stimulation). Many papers on BC cVEMPs used a B-71 clinical bone conductor, but this stimulator has not been taken up for the oVEMP (Iwasaki et al., 2007). Evidence suggests that the properties of the stimulus are critical determinants of both cVEMP and oVEMP characteristics. In particular, for BC stimulation the direction, frequency and site of stimulation have a significant effect on the polarity and latency of the reflexes (Cai et al., 2010). Finally, it is known that age can have significant effects on cVEMPs, which differ for different stimulus types (Welgampola and Colebatch, 2001). As such, it is important to understand the characteristic responses evoked by these stimuli, their variability and the effects of age. We therefore compared the cVEMPs and oVEMPs evoked by a range of effective vestibular stimuli in a large sample of normal subjects aged between 18 and 80 years of age. 2. Methods 2.1. Subjects Sixty-one normal, community-dwelling volunteers participated, with 10 in each decade from 20 to 80 years of age (28 females, 33 males; age range 18–80 years). We also tested an 18 year old male who was included in the 20–29 year group. Exclusion criteria included conductive hearing loss, middle ear disease or surgery, a diagnosis of vestibular disease, vertigo that lasted for more than a day or required hospitalization or neurological disease. Screening audiograms were conducted in all subjects using 500 Hz AC and BC tones and subjects with more than a 20 dB airbone gap were excluded. This value incorporated the error due to the acoustically unshielded room. All participants gave informed consent according to the Declaration of Helsinki and the study was approved by the local ethics committee. 2.2. Stimuli All subjects were stimulated with AC clicks and tone bursts, BC tone bursts delivered to the mastoids and head taps delivered to the forehead. Impulsive acceleration applied to the mastoid was added slightly later to the study (N = 51). The clicks and tone bursts

2283

were matched for sound energy (i.e. in dB LAeq (1s)). This allowed accurate comparison between stimuli as the differing energy content was taken into account (Rosengren et al., 2009b). The clicks were 0.2 ms square waves of 105 dB LAeq (1s) (2.72 V input, 135 dB peak SPL) and the tone bursts were 500 Hz, 2 ms unshaped sine waves of 105 dB LAeq (3.83 V peak to peak input, 132 dB peak SPL) delivered with calibrated headphones (TDH 49, Telephonics Corp., Farmingdale, USA). The BC tone bursts were 500 Hz, 4 ms unshaped sine waves of 136 dB peak FL (6.3 N peak) and alternating polarity delivered by a B-71 bone conductor placed over the mastoid (B-71, Radioear Corp., New Eagle, PA, USA). Taps were delivered near the hairline using a reflex tendon hammer with an electronic trigger (model 842-116700, Nicolet Biomedical Inc, Madison, USA). Impulsive head acceleration was produced by a hand-held ‘minishaker’ (model 4810, Brüel & Kjaer, Denmark). The minishaker input was a third order gamma distribution with 4 ms rise time and peak amplitude of 131 dB FL (3.55 N peak; Todd et al., 2008). The minishaker was held behind the external auditory meatus by the experimenter with 1 kg force. The stimulus had positive polarity, i.e. the rod moved the head away from the motor, similar to a tap to the mastoid. The pulses produced 0.13 g head acceleration and the taps 0.45 g (Rosengren et al., 2009a). A total of 256 stimuli were delivered at a rate of 5 Hz, except for the forehead taps, for which 40 stimuli were delivered at 2 Hz. 2.3. oVEMP recordings Subjects sat upright and directed their gaze to a target located 3 m away at an elevation of 20 deg. In some trials subjects were also asked to adopt maximal up-gaze (see below). The active electrode was placed on the orbital margin below the eye and referred to an electrode 15 mm below it on the cheek, with the earth placed on the sternum. Our recording parameters have been reported in detail previously (Govender et al., 2009). Amplitudes were measured from baseline to peak (for the n1) and also peak to peak (n1-p1). 2.4. cVEMP recordings Subjects reclined to 30 deg above horizontal and lifted their heads to activate the SCM muscles. The active electrode was placed over the muscle belly, the reference over the medial clavicle, and the earth on the sternum. Rectified EMG was monitored and recorded, and care was taken to ensure similar background activation of the SCM between sides of the neck and trials. Standard recording parameters were used and have been reported in detail previously (Rosengren et al., 2009a). Amplitudes were measured from peak to peak (pp: p1-n1), and expressed as the ratio of pp amplitude to the background mean rectified activity measured over the 20 ms pre-stimulus period. For the acceleration pulses, the ipsilateral cVEMP was measured peak to peak, while the contralateral cVEMP was measured from baseline to p1 peak, as only the first peak is vestibular-dependent (Rosengren et al., 2009a). 2.5. Procedure Screening audiograms were conducted before the VEMPs. The order of VEMP tests was counterbalanced. Half of the subjects underwent cVEMP testing first and half oVEMP testing first. AC stimuli were delivered first (counterbalanced for clicks vs tone bursts), and one ear of each subject was chosen a priori for threshold testing with both types of AC stimuli (half left and half right in each decade). Thresholds were measured using a 10–5 dB bracketing technique and responses were accepted when present on two

2284

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289

trials. For the AC cVEMP trials, subjects who had no response to 105 dB LAeq stimuli were also tested at 110 dB. The same procedure was followed for AC oVEMPs, but if no response was present at 110 dB, the test was repeated at 110 dB with maximal up-gaze. Ocular VEMP trials were repeated if n1 responses were absent or smaller than 0.5 lV. Stimulation with BC tone bursts, pulses and taps followed.

3. Results 3.1. AC stimulation 3.1.1. 500 Hz, 2 ms tone bursts Ocular VEMPs evoked by 105 dB LAeq tone bursts were present in the contralateral eye with stimulation of most ears (99/122 ears, 81%). In 5 subjects, oVEMPs were absent with stimulation of both ears and in 13 with stimulation of one ear. The proportion present increased to 93% when stimulus intensity was increased to 110 dB and to 97% when subjects were allowed to adopt maximal up-gaze. Stimulation with 105 dB tone bursts produced cVEMPs in significantly more subjects (117/122, 96% of ears, v2(1) = 13.1, P < 0.005). This proportion increased to 99% when the 5 ears with absent responses were stimulated at 110 dB. Mean amplitudes and latencies of the contralateral n1 oVEMP at 105 dB were 1.75 lV and 9.9 ms. Additional amplitude, latency and inter-side latency difference values for the oVEMP and all values for the cVEMP for each stimulus are shown in Tables 1 and 2. Traces from two representative subjects are shown in Figs. 1 and 2, while amplitudes from all subjects are shown in Figs. 3 and 4. Ocular VEMPs were present in the ipsilateral eye in 68/122 cases (56%) at 105 dB and had a mean amplitude of 0.82 ± 0.50 lV at 11.6 ± 1.5 ms. Although the contralateral responses were larger on average (t(65) = 6.1, P < 0.001), the ipsilateral response was larger in 9 cases. Mean amplitude asymmetry values are shown in Tables 3 and 4. In terms of threshold, oVEMP n1 threshold was 100 ± 6 dB LAeq, significantly higher than cVEMP threshold for the same stimulus (94 ± 7 dB, t(59) = 6.3, P < 0.001). The mean threshold difference between the oVEMP and cVEMP was 6.3 ± 7.7 dB (Fig. 5). Below the oVEMP threshold for the n1 peak, later potentials were still present, suggesting that the threshold for the entire oVEMP response may be lower.

2.6. Data analysis Responses were coded as originating from the eye or SCM ipsilateral or contralateral to the stimulus and considered together for the left and right sides (i.e. usually 122 sides of stimulation). For the responses evoked by forehead taps the right and left sides were also considered together. For oVEMPs, the main comparisons were of contralateral responses and for cVEMPs of ipsilateral responses, reflecting the dominant projections for these reflexes. For comparison of amplitudes across stimuli, 0-values were excluded, as large numbers of missing values in some conditions led to very small mean values, which were not representative of responses when they were present. For asymmetry and age analyses, 0-values were included. For asymmetry analysis subjects with bilaterally absent responses were excluded. Two asymmetry values were calculated for the AC stimuli, one at 105 dB and one including only subjects with ‘true’ 100% asymmetry (i.e. excluding subjects who had absent responses at 105 dB but present responses at 110 dB). Asymmetry values and inter-side latency differences were calculated using the larger-smaller method (i.e. 100  (larger response smaller response)/(larger response + smaller response)). Response prevalence was examined using chi-square, while amplitude, latency and asymmetry were compared using repeated-measures ANOVA and t-tests, where appropriate (for the BC stimuli this reduced the N to 51 as the pulses were used in less subjects). The Bonferroni correction for multiple comparisons was used to control for family-wise error. Correlations were used to investigate age and gender effects. The coefficient of variation (CV) was calculated for each stimulus and compared using Wilcoxon’s signed ranks test. Values reported in the text are mean ± SD. Statistics were performed with PASW Statistics (ver. 18 SPSS Inc, Chicago, Illinois).

3.1.2. 0.2 ms clicks Ocular VEMPs evoked by clicks were present with stimulation of 72/122 ears (59%). In 15 subjects oVEMPs were absent with stimulation of both sides and in 20 with stimulation of one side. The proportion increased to 84% with the 110 dB stimulus and to

Table 1 Ocular VEMP amplitudes and latencies from the ipsilateral and contralateral eyes. Ipsilateral eye n1 amplitude

Contralateral eye pp amplitude

n1 latency

p1 latency

Latency diff.

AC stimulation Tone burst Mean SD N Click

Mean SD N

BC stimulation Tone burst Mean SD N Tap

Mean SD N

Pulse

Mean SD N

1.00 0.72 75

4.12 3.10 94

2.43 1.67 75

10.26 6.68 94

11.6 1.5 75

15.5 1.2 94

16.8 1.7 75

21.1 1.4 94

1.1 1.3 26

0.9 1.0 45

n1 amplitude

pp amplitude

n1 latency

pl latency

Latency diff.

1.75 1.40 99

3.78 3.06 99

9.9 1.0 99

15.4 1.3 99

0.7 0.7 43

1.20 0.84 72

2.35 1.50 72

9.1 0.9 72

14.4 1.4 72

0.7 0.9 26

1.04 0.74 71

2.90 2.43 71

10.9 1.7 71

16.3 2.2 71

0.7 0.7 27

5.78 3.10 117

13.11 7.39 117

7.2 0.7 117

13.1 1.2 117

0.5 0.5 57

3.00 2.24 97

7.36 5.58 97

11.8 0.8 97

17.1 1.3 97

0.5 0.4 57

Amplitudes are in lV and latencies in ms. Only subjects with responses present are included. For AC stimuli, only the dominant, contralateral projection is shown. For BC tone burst stimuli, responses that began with initial positivities were not included (n = 4 for each eye). For taps the stimulus was presented in the midline, and all responses are therefore shown together. Abbreviations: pp amplitude = peak-to-peak amplitude, AC = air-conducted, BC = bone-conducted, latency diff. = inter-side latency difference.

2285

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289 Table 2 Cervical VEMP amplitudes and latencies from the ipsilateral and contralateral SCM muscles. Ipsilateral SCM

Contralateral SCM

pp amplitude

p1 latency

n1 latency

Latency diff.

Mean SD N

1.41 0.59 117

14.9 1.9 117

22.5 2.2 117

1.2 0.9 56

Mean SD N

0.90 0.39 111

13.9 2.0 111

21.1 2.2 111

1.6 1.2 53

Mean SD N

1.25 0.64 113

15.5 1.8 113

24.1 2.5 113

1.2 1.0 54

Tap

Mean SD N

1.84 0.79 119

12.5 1.9 119

21.0 10.2 119

1.2 0.9 57

Pulse

Mean SD N

1.13 0.42 108

16.3 1.8 108

24.1 1.9 108

1.6 1.3 54

AC stimulation Tone burst

Click

BC stimulation Tone burst

pp amplitude

p1 latency

n1 latency

Latency diff.

0.86 0.58 102

16.2 2.6 102

25.0 3.4 102

2.2 2.5 47

0.84 0.36 108

21.7 1.9 108

1.6 1.6 54

Amplitudes were corrected for background activity (see text) and latencies are in ms. Only subjects with responses present are included. For AC stimuli, only ipsilateral responses were present. For taps the stimulus was presented in the midline, and all responses are therefore shown together. Abbreviations: pp amplitude = peak-to-peak corrected amplitude, SCM = sternocleidomastoid, AC = air-conducted, BC = bone-conducted, latency diff. = inter-side latency difference.

Fig. 1. oVEMP traces from 2 single subjects, one a 35 year female the other a 70 year old male. All stimuli were given at time 0 and simultaneous recordings from below the eye ipsilateral and contralateral to the stimuli are shown. For taps the traces are from the right (‘‘ipsilateral side’’) and left eyes (‘‘contralateral side’’), as the stimulus was presented in the midline. Note the lower gain for the forehead tap and lateral pulse responses.

Fig. 2. cVEMP traces from the same 2 subjects illustrated in Fig. 1. All stimuli were given at time 0 and recordings from the SCM ipsilateral and contralateral to the stimuli are shown. For taps the traces from the right SCM are shown as ipsilateral and from the left SCM as contralateral, as the stimulus was presented in the midline. For AC tone bursts and clicks (A and B), p1 and n1 are more commonly known as p13 and n23.

2286

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289

Fig. 3. oVEMP amplitudes across stimuli for all subjects. For the BC tone bursts (part C) and lateral pulses (part E) the black diamonds indicate responses recorded from the eye contralateral to the stimulus and the grey diamonds those ipsilateral to the stimulus. There was a significant correlation between age and amplitude for the AC click oVEMP (part B) and contralateral BC tone burst oVEMP (part C).

89% with maximal up-gaze. In contrast, cVEMPs evoked by 105 dB clicks were more prevalent (v2(1) = 33.4, P < 0.005) and were present with stimulation of 111/122 (91%) ears (increasing to 97% with stimulation at 110 dB). The mean amplitude and latency of the contralateral n1 at 105 dB were 1.20 lV and 9.1 ms. Responses were seen in the ipsilateral eye in 46/122 cases (38%) at 105 dB and had a mean amplitude of 0.64 ± 0.27 lV at 11.1 ± 1.7 ms. Threshold was 103 ± 5 dB LAeq, significantly higher than the cVEMP threshold for the same click stimulus (99 ± 6 dB, t(53) = 4.6, P < 0.001). The mean threshold difference between the oVEMP and cVEMP for clicks was 3.9 ± 6.4 dB. 3.2. BC stimulation 3.2.1. 500 Hz, 4 ms tone bursts at mastoid Ocular VEMPs evoked by 500 Hz BC vibration were variable. Responses were therefore accepted as present if the initial peak occurred between 8 and 15 ms and was either very clear on a single trial or reproducible over 2 trials. BC oVEMPs were present in 79/122 (65%) of cases in the eye ipsilateral to stimulation and 75/122 (62%) of cases in the contralateral eye. Responses were bilateral in 49% of subjects, ipsilateral only in 16%, contralateral only in 13% and bilaterally absent in 20%. In many cases the n1 peak was small, not reproducible or absent, but later peaks were clear, suggesting that the stimulus was near threshold for the n1 response. In contrast, cVEMPs evoked by the same stimulus were present in the ipsilateral SCM in 113/122 cases (93%) and in the contralateral SCM in 102/122 cases (84%), significantly more frequent than the oVEMP (88% vs 63%, v2(1) = 41.3, P < 0.005). In four

Fig. 4. cVEMP amplitudes across stimuli for all subjects. In parts C and E the black diamonds indicate responses recorded from the SCM ipsilateral to the stimulus and the grey diamonds those contralateral to the stimulus. There were significant correlations between age and amplitude for AC tone bursts and clicks (parts A and B), BC tone bursts on both sides (part C) and pulses (part E, contralateral side only).

Table 3 Ocular VEMP asymmetry values for the ipsilateral and contralateral eyes. Ipsilateral eye n1 amplitude

Contralateral eye pp amplitude

AC stimulation Tone Mean burst SD N Click

Mean SD N

BC stimulation Tone Mean burst SD N Tap

Mean SD N

Pulse

Mean SD N

n1 amplitude

pp amplitude

44 (30)

45 (30)

36 (25) 56 (45)

35 (23) 56 (45)

59 (32) 38 (24) 45 (27)

56 (27) 39 (21) 45 (27)

61

54

51

49

38 46

43 46

38 42

39 42

19 21 60

18 21 60

28 26 50

27 27 50

23 23 47

20 21 47

Asymmetry values (all in%) exclude bilaterally absent responses. For AC stimuli, the values in brackets represent asymmetry values after removal of subjects who had absent responses at 105 dB but present responses at 110 dB. AC clicks and tone bursts and BC tone bursts were all associated with high average levels of asymmetry. Abbreviations: n1 amplitude = asymmetry calculated using n1 amplitudes, pp amplitude = calculations based upon the peak-to-peak amplitude, AC = air-conducted, BC = bone-conducted.

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289 Table 4 Cervical VEMP asymmetry values for the ipsilateral and contralateral SCM muscles. Ipsilateral SCM pp amplitude AC stimulation Tone burst

Mean SD N

18 17 57

Mean SD N

15 20 55

Mean SD N

21 27 59

Tap

Mean SD N

15 16 60

Pulse

Mean SD N

13 11 54

Click

BC stimulation Tone burst

Contralateral SCM pp amplitude

2287

eral eye peaked late at 15.5 ms with amplitude 4.12 lV (Todd et al., 2008). The pattern was the opposite for the cVEMP, with an early response in the ipsilateral SCM and late response on the contralateral side (Table 2; Rosengren et al., 2009a). 3.3. Comparison of stimuli

31 32 59

13 12 54

Asymmetry values (%) exclude bilaterally absent responses. Abbreviations: pp amplitude = calculations based upon peak-to-peak amplitude, AC = air-conducted, BC = bone-conducted.

Fig. 5. Mean thresholds for oVEMPs and cVEMPs in response to the 500 Hz AC tone burst and the click stimuli for all subjects who had responses. The mean and standard errors are shown. On average oVEMPs had a higher threshold than cVEMPs and tone bursts had a significantly lower threshold than clicks.

subjects, at least one oVEMP began with an initial positivity instead of a negativity (8/244 possible cases, 4 in the ipsilateral eye, 4 contralateral, mean amplitude 1.02 ± 0.56 lV and latency 10.9 ± 1.1 ms). 3.2.2. Forehead taps Ocular VEMPs produced by forehead taps were present in both eyes in 58/62 subjects (in 117/122 cases, 96%). One subject had absent responses bilaterally and 3 had tap oVEMPs in one eye only. The prevalence was similar for the cVEMP (119/122 cases, 98%). The mean amplitude and latency for the oVEMP n1 were 5.78 lV and 7.2 ms (trigger delay was 4.5 ms). 3.2.3. Acceleration pulses at mastoid Ocular VEMPs evoked by lateral acceleration pulses were present in the ipsilateral eye in 94/102 cases (92%) and the contralateral eye in 97/102 cases (95%). Five subjects had absent responses, and two subjects had a response with unusual morphology, which were classed as absent. All subjects had cVEMPs on both sides of the neck when stimulated with lateral pulses. Both pulse oVEMPs and cVEMPs had different properties when recorded from the eye or neck ipsilateral versus contralateral to the stimulus. The oVEMP in the contralateral eye peaked early at 11.8 ms with a mean amplitude of 3.00 lV, while the oVEMP in the ipsilat-

Compared to the AC tone bursts, oVEMPs were significantly less prevalent when evoked by clicks at the same relative intensity of 105 dB LAeq (59% vs 81%, v2(1) = 14.3, P < 0.005), while there was no significant difference in cVEMP prevalence (91% vs 96%). The amplitudes of both reflexes were also larger when evoked by tone bursts compared to clicks (oVEMP F(1,22) = 26.1, P < 0.001, cVEMP F(1,52) = 218.9, P < 0.001). Tone burst thresholds were significantly lower for both reflexes (oVEMP t(53) = 6.7, P < 0.001, cVEMP t(57) = 9.9, P < 0.001). In terms of BC stimuli, BC tone burst oVEMPs were significantly less prevalent than tap and pulse evoked responses (v2(1) = 46.5 and 58.7, P < 0.005). When contralateral oVEMP amplitudes were compared across BC stimuli (i.e. the dominant projection), there was a significant effect (F(2,44) = 50.2, P < 0.001). Tap oVEMPs were significantly larger than pulse oVEMPs (t(46) = 7.5, P < 0.001), which in turn were larger than tone burst oVEMPs (t(30) = 4.5, P < 0.001). For the ipsilateral (dominant) cVEMP, the BC tone burst produced significantly fewer responses than the lateral pulse (v2(1) = 9.7, P < 0.005). Tap cVEMPs were significantly larger than those evoked by tone bursts or pulses (F(2,88) = 28.3, P < 0.001). Reflexes evoked by AC clicks and tone bursts were well correlated with each other (oVEMP amplitude r = 0.65 and 0.77 for right and left ears, P < 0.001; cVEMP r = 0.84–0.86, P < 0.001), but not well correlated with those evoked by the BC stimuli (r = 0.05–0.60, P = 0.001–0.72). In contrast, reflexes evoked by BC stimuli were only moderately correlated with each other (oVEMP amplitude r = 0.26–0.53, P = 0.001–0.06, cVEMP r = 0.20–0.50, P = 0.001–0.15). The correlation between cVEMPs and oVEMPs was also only moderate (r = 0.11–0.42, P = 0.001–0.40 over all stimuli). The asymmetry values for the oVEMP were significantly higher than for the cVEMP overall (F(1,12) = 6.8, P = 0.023), but there were no significant differences between stimuli for either reflex. Asymmetry values were significantly correlated between the taps and lateral pulses for the oVEMP (r = 0.56, P < 0.001) and between the AC stimuli and BC tone bursts for the cVEMP (r = 0.47–0.63, P < 0.001). The coefficient of variation for oVEMP amplitude ranged from 0.5 to 0.8 and for cVEMP amplitude from 0.4 to 0.5. There was little difference in variability between stimuli, but the oVEMPs were significantly more variable than the cVEMPs (P = 0.042). The opposite pattern was seen for latency, with the CV ranging from 0.11 to 0.15 for the cVEMP and from only 0.07 to 0.10 for the oVEMP (excluding the BC tone burst stimuli, for which the oVEMP CV was 0.16). 3.4. The effects of age and gender There was a significant, moderate correlation between age and amplitude for the AC click oVEMP and contralateral BC tone burst oVEMP only, indicating that the responses became smaller with increasing age (clicks r = 0.33, P = 0.009; BC tone bursts r = 0.41, P = 0.001). For the cVEMP, there were significant correlations between age and amplitude when evoked by AC clicks and tone bursts, BC tone bursts (on both sides of the neck) and pulses (contralateral side only) (clicks r = 0.61, P < 0.001; AC tone bursts r = 0.58, P < 0.001; BC tone bursts ipsilateral r = 0.62, P < 0.001; BC tone bursts contralateral r = 0.56, P < 0.001; pulses contralateral r = 0.30, P = 0.027). Compatible with this, the thresholds for both the cVEMP and oVEMP increased with increasing age for both

2288

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289

clicks and tone bursts (r = 0.24–0.42, P = 0.001–0.06). In terms of latency, for the oVEMPs evoked by AC clicks and tone bursts, the n1 peaked significantly later with increasing age (clicks r = 0.37, P = 0.012; tone bursts r = 0.42, P = 0.001). There were no effects of age on latency for the cVEMP. There were no significant effects of gender on oVEMP or cVEMP amplitude or latency for any stimulus.

4. Discussion We have shown that a range of vestibular stimuli are sufficient to produce cVEMPs in most subjects across a large age range, while the same stimuli differed in their ability to evoke oVEMPs. Our data confirm that the overwhelming majority of normal people under 60 years of age should have a cVEMP to both AC and BC stimuli (Welgampola and Colebatch, 2001). In contrast, only the lateral pulses and forehead taps were capable of producing oVEMPs in over 90% of all cases. Our data demonstrate the importance of stimulus parameters for VEMPs. We have confirmed that tone bursts are superior to clicks for producing both cVEMPs and oVEMPs (Chihara et al., 2007; Rosengren et al., 2009b). We matched the stimuli using dB LAeq, which takes the energy content of each stimulus into account and is a better method to compare stimuli than simple measures of intensity such as nHL or SPL. Despite this matching, the tone bursts produced responses that were larger, more frequent and had a lower threshold than the clicks. This is most likely due to the documented tuning of the AC cVEMP to frequencies between 400 and 1000 Hz (e.g. Akin et al., 2003; Todd et al., 2000). For the oVEMP to AC sound in particular, the stimulus intensity and angle of gaze were important factors. Increasing the intensity by 5 dB or changing the angle of gaze from 20 deg upwards to maximal up-gaze increased the frequency of responses. For small responses which are close to threshold, optimizing the stimulus and recording conditions is therefore critical. However, although maximal up-gaze is recommended for most clinical purposes (i.e. to avoid false positive errors), for comparison of responses between test sessions or groups of patients, the angle of gaze should also be controlled. We found a significant threshold difference between the cVEMP and oVEMP when evoked by both AC stimuli. Although our AC stimulus intensity was sufficient for the cVEMP, this was not the case for the oVEMP, particularly for clicks. Similar threshold differences in normal subjects have been reported by others (Welgampola et al., 2008; Park et al., 2010). The pattern of responses to the BC tone bursts suggests that the small oVEMPs might also be the result of a threshold difference, though threshold was not explicitly measured. In contrast, the lateral pulses and forehead taps did not appear to show a threshold difference. The threshold difference was independent of the overall size of the cVEMPs, suggesting that it is instead specific to the particular stimuli. It is possible that these effects are due to variation in the strength of the reflex pathways activated by these stimuli. Although the AC cVEMP is thought to reflect predominantly saccular (i.e. inferior vestibular nerve) activation, recent evidence suggests that the AC oVEMP requires activation of fibres travelling in the superior vestibular nerve (Curthoys et al., 2011; Govender et al., 2011). Given that all utricular afferents and some afferents from the anterior portion of the saccule course through the superior vestibular nerve (Lorente de No, 1933), the smaller oVEMPs to AC sound could therefore be the result of fewer saccular fibres being activated and/or less efficient activation of utricular fibres by AC sound. In contrast, BC vibration at 500 Hz has been shown to activate irregular otolith afferents from both end organs (Curthoys et al., 2006), though a contribution from canal afferents with lower frequency stimulation has not been ruled out. We found that the oVEMP evoked by 500 Hz BC vibration also appeared to have a higher threshold than

the cVEMP. However, similar stimulation delivered to the forehead instead of the mastoids can produce large oVEMPs (albeit with higher intensity stimulation: Iwasaki et al., 2008), but may not be consistently effective for cVEMPs (Rosengren SM, personal observations). The reason for this differential effect based on stimulus direction is not clear, however it suggests that the position of a vibration stimulus on the head, and therefore the direction of the induced head acceleration, is a critical determinant of VEMP properties. The variation in the initial acceleration produced by the B71 bone conductor might explain the presence of an initial oVEMP positivity in some subjects (Cai et al., 2010). The presence of a threshold effect means that it is important for clinical studies in which cVEMPs and oVEMPs are compared to demonstrate that an absent oVEMP is not the consequence of a simple threshold difference. We also found differential age effects across the stimulus types. The age effect was independent of the overall size of the reflexes, as it occurred with the stimuli which produced the smallest oVEMPs but some of the largest cVEMPs. Decline in amplitude with increasing age has previously been reported for the AC-evoked cVEMP (Basta et al., 2007; Brantberg et al., 2007; Welgampola and Colebatch, 2001). A similar age effect has also been reported for the oVEMP for AC sound and some forms of BC vibration (Iwasaki et al., 2008; Nguyen et al., 2010; Tseng et al., 2010). Taps and lateral pulse oVEMPs and cVEMPs were less affected by age. This extends previous reports (Welgampola and Colebatch, 2001), and is consistent with these stimuli activating a different group of vestibular afferents than the AC sound. Although the effects of age are not uniform across stimuli, it is an important factor to consider in older patients. We also found that the asymmetry and amplitude variability were higher for the oVEMP than the cVEMP. Our values were similar to those reported by Nguyen et al. (2010) for both reflexes and to previous cVEMP studies using similar stimuli (e.g. Welgampola and Colebatch, 2001). In contrast, values reported by Iwasaki et al. (2008) and Tseng et al. (2010) were lower, possibly due to higher stimulus intensities being used in their studies. Latencies for the cVEMP had low variability, similar to previous studies (e.g. Basta et al., 2005; Welgampola and Colebatch, 2001). We found even less variation in oVEMP latency, with the standard deviation across stimuli being 1 ms or less (apart from the BC tone burst). This supports results from other large oVEMP studies (Iwasaki et al., 2008; Nguyen et al., 2010; Tseng et al., 2010). Although normal values will differ between centres, our data show that the latency range for the oVEMP is very narrow across a range of stimuli. This is likely to be due to the strong connectivity of the threeneuron arc underling the VOR and reflects the importance of rapid stabilization of gaze. Care should therefore be taken in accepting responses with peaks later than a couple of milliseconds beyond the normal mean. Delayed responses can be caused by demyelination of central vestibular pathways (Rosengren and Colebatch, 2010), however late peaks can also be due to failure to control gaze angle (Govender et al., 2009) or inadequate stimulus intensity (e.g. when a stimulus is below threshold for the n1). The results of this study demonstrate that stimulus parameters should be chosen carefully. All of our stimuli were suitable for eliciting cVEMPs, while the forehead taps and lateral pulses produced the largest oVEMPs. An important consideration in stimulus selection is the specificity of the stimulus (or reflex pathway) for a particular set of vestibular afferents. Given that cVEMPs evoked by AC sound are mainly produced by activation of the saccule, an AC tone burst is the most suitable stimulus for examining the sacculo-collic pathway. Recent data has shown that cVEMPs elicited by lateral pulses are likely produced by utricular activation (Govender et al., 2011; Rosengren et al., 2009a), making the lateral pulse a good stimulus to examine the utriculo-collic pathway. For the oVEMPs, the origin of the AC-evoked response has not been confirmed,

S.M. Rosengren et al. / Clinical Neurophysiology 122 (2011) 2282–2289

though it appears to be produced by otolith fibres coursing through the superior vestibular nerve. The small amplitudes, high thresholds and high asymmetry associated with oVEMPs evoked by the AC and B-71 stimuli reduce the utility of these modalities, but AC stimulation appears to be better for the detection of superior canal dehiscence (Rosengren et al., 2008; Welgampola et al., 2008). Ocular VEMPs evoked by specific types of BC stimulation such as lateral pulses (Todd et al., 2008) and 500 Hz vibration delivered to the forehead (at higher intensities than used here: Iwasaki et al., 2009; Manzari et al., 2010) appear to evoke larger and more symmetrical responses and may thus prove to have an important role as stimuli for oVEMPs. Acknowledgements This work was supported by the Garnett Passe and Rodney Williams Memorial Foundation and the National Health and Medical Research Council of Australia. References Akin FW, Murnane OD, Proffitt TM. The effects of click and tone-burst stimulus parameters on the vestibular evoked myogenic potential (VEMP). J Am Acad Audiol 2003;14(9):500–9. Basta D, Todt I, Ernst A. Normative data for P1/N1-latencies of vestibular evoked myogenic potentials induced by air- or bone-conducted tone bursts. Clin Neurophysiol 2005;116:2216–9. Basta D, Todt I, Ernst A. Characterization of age-related changes in vestibular evoked myogenic potentials. J Vestib Res 2007;17(2–3):93–8. Brantberg K, Granath K, Schart N. Age-related changes in vestibular evoked myogenic potentials. Audiol Neurootol 2007;12(4):247–53. Brantberg K, Mathiesen T. Preservation of tap vestibular evoked myogenic potentials despite resection of the inferior vestibular nerve. J Vestib Res 2004;14(4):347–51. Cai KY, Rosengren SM, Colebatch JG. Cervical and ocular vestibular evoked myogenic potentials are sensitive to stimulus phase. Audiol Neurootol 2010;16(5):277–88. Chihara Y, Iwasaki S, Ushio M, Murofushi T. Vestibular-evoked extraocular potentials by air-conducted sound: another clinical test for vestibular function. Clin Neurophysiol 2007;118(12):2745–51. Colebatch JG, Halmagyi GM, Skuse NF. Myogenic potentials generated by a clickevoked vestibulocollic reflex. J Neurol Neurosurg Psychiatry 1994;57(2):190–7. Curthoys IS, Iwasaki S, Chihara Y, Ushio M, McGarvie LA, Burgess AM. The ocular vestibular-evoked myogenic potential to air-conducted sound; probable superior nerve origin. Clin Neurophysiol 2011;122:611–3. Curthoys IS, Kim J, McPhedran SK, Camp AJ. Bone conducted vibration selectively activates irregular primary otolithic vestibular neurons in the guinea pig. Exp Brain Res 2006;175(2):256–67. Govender S, Rosengren SM, Colebatch JG. The effect of gaze direction on the ocular vestibular evoked myogenic potential produced by air-conducted sound. Clin Neurophysiol 2009;120(7):1386–91. Govender S, Rosengren SM, Colebatch JG. Vestibular neuritis has selective effects on air- and bone-conducted cervical and ocular vestibular evoked myogenic potentials. Clin Neurophysiol 2011;122:1246–55.

2289

Iwasaki S, Chihara Y, Smulders YE, Burgess AM, Halmagyi GM, Curthoys IS, et al. The role of the superior vestibular nerve in generating ocular vestibular-evoked myogenic potentials to bone conducted vibration at Fz. Clin Neurophysiol 2009;120(3):588–93. Iwasaki S, McGarvie LA, Halmagyi GM, Burgess AM, Kim J, Colebatch JG, et al. Head taps evoke a crossed vestibulo-ocular reflex. Neurology 2007;68(15):1227–9. Iwasaki S, Smulders YE, Burgess AM, McGarvie LA, MacDougall HG, Halmagyi GM, et al. Ocular vestibular evoked myogenic potentials to bone conducted vibration of the midline forehead at Fz in healthy subjects. Clin Neurophysiol 2008;119(9):2135–47. Lorente de No R. Anatomy of the eighth nerve. I. The central projection of nerve endings of the internal ear. Laryngoscope 1933;43:1–38. Manzari L, Tedesco A, Burgess AM, Curthoys IS. Ocular vestibular-evoked myogenic potentials to bone-conducted vibration in superior vestibular neuritis show utricular function. Otolaryngol Head Neck Surg 2010;143(2):274–80. McCue MP, Guinan JJ. Acoustically responsive fibers in the vestibular nerve of the cat. J Neurosci 1994;14(10):6058–70. Murofushi T, Curthoys IS. Physiological and anatomical study of click-sensitive primary vestibular afferents in the guinea pig. Acta Otolaryngol 1997;117(1):66–72. Nguyen KD, Welgampola MS, Carey JP. Test–retest reliability and age-related characteristics of the ocular and cervical vestibular evoked myogenic potential tests. Otology Neurotol 2010;31:793–802. Park HJ, Lee I-S, Shin JE, Lee YJ, Park MS. Frequency-tuning characteristics of cervical and ocular vestibular evoked myogenic potentials induced by air-conducted tone bursts. Clin Neurophysiol 2010;121(1):85–9. Rosengren SM, Aw ST, Halmagyi GM, Todd NP, Colebatch JG. Ocular vestibular evoked myogenic potentials in superior canal dehiscence. J Neurol Neurosurg Psychiatry 2008;79(5):559–68. Rosengren SM, Colebatch JG. Ocular vestibular evoked myogenic potentials are abnormal in internuclear ophthalmoplegia. Clin Neurophysiol 2011;122:1264– 67. Rosengren SM, Govender S, Colebatch JG. The relative effectiveness of different stimulus waveforms in evoking VEMPs: significance of stimulus energy and frequency. J Vest Res 2009b;19(1–2):33–40. Rosengren SM, Todd NPM, Colebatch JG. Vestibular-evoked extraocular potentials produced by stimulation with bone conducted sound. Clin Neurophysiol 2005;116:1938–48. Rosengren SM, Todd NPM, Colebatch JG. Vestibular evoked myogenic potentials evoked by brief interaural head acceleration: properties and possible origin. J Appl Physiol 2009a;107(3):841–52. Todd NPM, Cody FWJ, Banks JR. A saccular origin of frequency tuning in myogenic vestibular evoked potentials? Implications for human responses to loud sounds. Hear Res 2000;141(1–2):180–8. Todd NPM, Rosengren SM, Aw ST, Colebatch JG. Ocular vestibular evoked myogenic potentials (OVEMPs) produced by air- and bone-conducted sound. Clin Neurophysiol 2007;118(2):381–90. Todd NPM, Rosengren SM, Colebatch JG. Ocular vestibular evoked myogenic potentials (OVEMPs) produced by impulsive transmastoid accelerations. Clin Neurophysiol 2008;119(7):1638–81. Tseng C-L, Chou C-H, Young Y-H. Aging effect on the ocular vestibular-evoked myogenic potentials. Otol Neurotol 2010;31:959–63. Welgampola MS, Colebatch JG. Vestibulocollic reflexes: normal values and the effect of age. Clin Neurophysiol 2001;112(11):1971–9. Welgampola MS, Myrie OA, Minor LB, Carey JP. Vestibular-evoked myogenic potential thresholds normalize on plugging superior canal dehiscence. Neurology 2008;70(6):464–72. Young ED, Fernandez C, Goldberg JM. Responses of squirrel monkey vestibular neurons to audio-frequency sound and head vibration. Acta Otolaryngol 1977;84(5–6):352–60.