Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jul 17.
Published in final edited form as: Ear Hear. 2020 Jul 17;42(2):355–363. doi: 10.1097/AUD.0000000000000925

Bone Conduction Vibration Vestibular Evoked Myogenic Potential (VEMP) Testing: Reliability in Children, Adolescents, and Young Adults

Nicole, L Greenwalt 1,2, Jessie N Patterson 2, Amanda I Rodriguez 2,3, Denis Fitzpatrick 2, Katherine R Gordon 2, Kristen L Janky 2
PMCID: PMC7855257  NIHMSID: NIHMS1607194  PMID: 32701728

Abstract

Objectives:

Bone conduction vibration (BCV) vestibular evoked myogenic potentials (VEMP) are clinically desirable in children for multiple reasons. However, no accepted standard exists for stimulus type and the reliability of BCV devices has not been investigated in children. The objective of the current study was to determine which BCV VEMP method (B-71, impulse hammer, or Mini-shaker) yields the highest response rates and reliability in a group of adults, adolescents, and children. It was hypothesized that the Mini-shaker would yield the highest response rates and reliability because it provides frequency specificity, higher output levels without distortion, and the most consistent force output as compared to the impulse hammer and B-71.

Design:

Participants included 10 child (ages 5–10), 11 adolescent (ages 11–18), and 11 young adult (ages 23–39) normal controls. Cervical (cVEMP) and ocular VEMP (oVEMP) were measured in response to suprathreshold air-conducted, 500 Hz tone bursts and 3 types of BCV (B-71, impulse hammer, and Mini-shaker) across two test sessions to assess reliability.

Results:

For cVEMP, response rates were 100% for all methods in all groups with the exception of the adult group in response to the impulse hammer (95%). For oVEMP, response rates varied by group and BCV method. For cVEMP, reliability was highest in adults using the Mini-shaker, in adolescents using the impulse hammer, and in children using the B-71. For oVEMP, reliability was highest in adults using the Mini-shaker, in adolescents using the Mini-shaker or impulse hammer, and in children using the impulse hammer. Age positively correlated with air-conducted oVEMP amplitude, but not cVEMP amplitude or cVEMP corrected amplitude. Age negatively correlated with all BCV VEMP amplitudes with the exception of cVEMP corrected amplitude in response to the Mini-shaker.

Conclusions:

All BCV methods resulted in consistent cVEMP responses (response rates 95 – 100%) with at least moderate reliability (ICC ≥ 0.5) for all groups. Similarly, all BCV methods resulted in consistent oVEMP responses (89 – 100%) with at least moderate reliability (ICC ≥ 0.5) except for the B-71 in adults.

Keywords: vestibular, vestibular evoked myogenic potential, bone conduction, children, adolescents

1. Introduction

Vestibular evoked myogenic potentials (VEMPs) are fundamental to the comprehensive vestibular assessment because function of each otolith organ (utricle and saccule) and vestibular nerve branch (superior and inferior) can be uniquely evaluated. While the majority of clinical research in VEMP testing is focused on adults, recent research supports the use of VEMP in children (Picciotti et al. 2007; Chou et al. 2012; Wang et al. 2013; Zhou et al. 2014; Kuhn et al. 2018). The objective, non-invasive, time-saving, and well-tolerated nature of the VEMP test makes it especially appropriate for use in the pediatric population.

The VEMP is based on the acoustic and vibratory sensitivity of the otolith organs. Cervical VEMP (cVEMP) utilizes high-intensity, short-duration stimulation to evoke an inhibitory electromyogram (EMG) response from the contracted ipsilateral sternocleidomastoid muscle (SCM) and is mediated by the saccule and inferior portion of the vestibular nerve (Colebatch and Halmagyi, 1992; Colebatch et al. 1994). Ocular VEMP (oVEMP) utilizes similar stimulation to evoke an excitatory EMG response from the contralateral inferior oblique extraocular muscle and is mediated by the utricle, superior portion of the vestibular nerve, and some saccular afferents (Todd et al. 2000; Curthoys et al. 2006; Iwasaki et al. 2007; Iwasaki et al. 2009).

VEMPs can be evoked using air conduction stimulation (ACS) or bone conduction vibration (BCV); however, ACS consistently demonstrates lower response rates compared to BCV in adults (Nguyen et al., 2010; Wang et al. 2010; Curthoys et al., 2011; Chou et al. 2012; Wackym et al. 2012). A difference in the stimulus transduction mechanism between ACS and BCV may explain this dissimilarity in response rates (Curthoys et al., 2011). Linear accelerometers on the skulls of guinea pigs show that BCV causes small linear accelerations, which are not detectable during ACS (Curthoys et al. 2011). It is hypothesized that during ACS the pumping action of the stapes causes hair cell cilia of the otoliths to be deflected, resulting in movement of the cilia relative to the stationary cell body. In contrast, specific to utricular maculae which are situated on a flexible membrane, BCV may cause the cell body to move relative to the stationary cilia (Curthoys et al. 2011).

The use of BCV stimuli in children is clinically desirable for three reasons: (1) VEMP responses cannot be observed using ACS in the presence of conductive hearing loss with air-bone gaps greater than 20 dB (Halmagyi et al. 1994) and are significantly attenuated with air-bone gaps as small as 9 dB (Bath et al. 1999). Therefore, using ACS is problematic in the pediatric population where middle ear pathology is common. BCV stimulation allows for evaluation of the otoliths and vestibular nerve in the presence of middle ear pathology. (2) BCV VEMP can be obtained at significantly lower stimulation levels compared to ACS, thus making it a safer and more comfortable stimulus (Curthoys et al. 2011, Rodriguez et al. 2018). (3) Lastly, when BCV is delivered to the midline (Fz), the labyrinths are stimulated simultaneously and almost equally in magnitude, as opposed to individually with ACS (Halmagyi et al. 1995; Iwasaki et al. 2007; Yang et al. 2016). This method effectively cuts test time in half, which is crucial when testing children.

Unfortunately, little is known about appropriate methods for BCV VEMP testing in adults and children. First, there is no accepted standard for the BCV stimulus, as different stimulus generators, such as the Mini-shaker, B-71, and impulse hammer devices have been tested in adults (Halmagyi et al. 1995; Rosengren et al. 2005; Iwasaki et al. 2007; Iwasaki et al., 2008; Chou et al. 2012) and in children (Chou et al. 2012; Zhou et al. 2014; Fuemmeler et al, in press). Second, only one study has investigated BCV VEMP reliability in children and found reliable cVEMP, but not oVEMP responses using the impulse hammer (Fuemmeler et al, in press). In adults, the B-71 bone oscillator is less reliable and produces less robust responses compared to reflex hammer and Mini-shaker stimuli (Iwasaki et al. 2007). Although the B-71 bone oscillator has been successfully used in the pediatric population (Zhou et al. 2014; Maes et al. 2016), the device has not been compared to the reflex hammer and Mini-shaker in children. In adults, reliability also varies between c- and oVEMP. In response to reflex hammer and Mini-shaker stimuli, oVEMP amplitudes demonstrate excellent reliability, while cVEMP amplitudes demonstrate fair reliability (Nguyen et al. 2010).

Therefore, the objective of the current study was to determine which BCV VEMP method (B-71, impulse hammer, or Mini-shaker) yields the highest response rates and reliability in children and adolescents.

2. Methods

2.1. Participants

Thirty-two typically developing control participants were recruited from the local community and the Human Subjects Research Core, a database of individuals who have indicated interest in participating in research at Boys Town National Research Hospital (BTNRH). All participants provided informed consent through the Institutional Review Board at BTNRH. Participants were separated into the following age groups:

  • Adults: 23 – 39 years (6 females, 5 males; mean 29.5; n=11)

  • Adolescents: 11 – 18 years (6 females, 5 males; mean 13.9; n=11)

  • Children: 5 – 10 years (4 females, 6 males; mean 6.8; n=10)

Participants were characterized by having normal hearing sensitivity and middle ear function and no history of balance disorder, dizziness, or neurologic involvement per self and parental report when appropriate. To rule out the presence of middle ear pathology, all subjects completed 226 Hz probe tone tympanometry with the following inclusion criteria: ear canal volumes (ECV) ≤ 2.0 ml in adults and adolescents and ≤ 1.5 ml in children (British Society of Audiology, 2013), compliance ≥ 0.2 mmhos, and tympanic peak pressure within the range of −100 to 30 daPa. To rule out the presence of hearing loss, all subjects completed a hearing screening at 25 dB HL for 1000, 2000, and 4000 Hz (ASHA, 1997).

2.2. Testing Sessions

Each subject participated in two separate test sessions to assess the reliability of each BCV method. The average time between the two sessions was 12.8 days (range = 1 to 26 days) in adults and 4.95 days (range = 1 to 14 days) in adolescents and children.

2.3. cVEMP Recordings

Participants laid in the supine position and were instructed to lift their heads, nose pointed upward, to simultaneously activate bilateral SCM muscles. The active (non-inverting) electrode was placed on the SCM muscle belly with a reference (inverting) electrode on the sternoclavicular notch and a ground electrode on the right inner canthus (Figure 1A). For recordings using ACS, the B-71, and the Mini-shaker, EMG monitoring electrodes were integrated just below each active electrode to ensure recordings remained within 100 – 300 μV (Bogle et al., 2013). For recordings using the impulse hammer, EMG was recorded from the active SCM electrode. Electrode impedances were maintained below 10 kΩ. Outcome parameters were latency (p13/n23), peak-to-peak amplitude, and corrected amplitude (raw peak-to-peak amplitude/raw EMG).

Figure 1.

Figure 1.

Open in a new tab

Depiction of electrode montage for: A) cVEMP, and B) oVEMP.

2.4. oVEMP Recordings

Participants laid in the supine position and were instructed to gaze upward at a sticker placed on the ceiling at 30 degrees above eye level. Child and adolescent participants who required a distractor to maintain attention gazed upward at an iPod (Apple Inc., Cupertino, CA) cartoon presentation placed in lieu of a sticker. The active electrode was placed mediolaterally below the eye, over the contralateral inferior oblique muscle with a reference electrode on the right inner canthus and a ground electrode on the sternoclavicular notch (Figure 1B). Electrode impedances were maintained below 10 kΩ. Outcome parameters were latency (n10/p16) and peak-to-peak amplitude.

2.5. Stimuli and Recording Parameters

ACS and BCV stimuli from the B-71 and Mini-shaker consisted of 500 Hz tone bursts (Blackman gated; 2 ms rise/fall time, 0 ms plateau, condensation polarity) presented at a rate of 5.1 Hz. The stimulus frequency was selected based on physiologic data showing otolith neurons are sensitive to a 500 Hz stimulus (Curthoys et al., 2006). Responses were deemed morphologically acceptable if they met latency criteria (p13/n23 for cVEMP and n10/p16 for oVEMP) and adequately increased out of the noise floor. Two trials were completed to ensure replicability. Responses were considered absent if not replicated over two trials. Responses were coded with a 1 if present, a 0 if absent and a 0.5 if present only on the right or left side across test sessions. Each recording represented an average of 75 sweeps per trial. Artifact rejection was turned off. EMG signals were amplified 5000x and band-pass filtered from 5 to 500 Hz.

For recordings using the impulse hammer, VEMP waveforms were imported into Matlab (Mathworks, Natick, MA; Version 2016a) for analysis. cVEMP amplitudes were normalized using a 20 ms (−30 to −10 ms) pre-stimulus window, which was rectified and averaged to create an estimated pre-stimulus tonic EMG level. This mean root-mean-square (RMS) value was divided by the averaged cVEMP waveform (McCaslin et al., 2014; McCaslin et al., 2013; Bogle et al., 2013). Due to differences in manufacturer and EMG sensitivity, EMG was limited between 100 and 400 μV for 30 of the total 33 participants. The maximum for EMG limiting software was increased to 450 or 500μV to accommodate naturally high EMG levels in three subjects (1 adult, 1 adolescent, and 1 child).

2.6. VEMP Testing

All participants completed ACS VEMP testing using ER-3A insert earphones at a presentation level of 94 dBnHL (125 dB pSPL) for ECVs > 0.8 ml or 89 dBnHL (120 dB pSPL) for ECVs ≤ 0.8 ml. Levels based on equivalent ECV were chosen as a result of previous research showing peak equivalent sound pressure level (peSPL) sound recordings in children’s ears are significantly higher (~ 3dB) than adults in response to high-intensity ACS VEMP 500 Hz tone burst stimuli (Rodriguez et al., 2018). BCV VEMPs were completed using a B-71 bone oscillator (Radioear Corporation, New Eagle, PA, USA), 4810 Mini-shaker (Bruel & Kjaer, Denmark), and Piezotronics impulse hammer with integrated ICP quartz force sensor (Model 086C01, sensitivity of 11.2 millivolts/Newton; PCB Corporation, Depew, New York, USA), which quantified the force of each tap. ACS elicited using ER-3A insert earphones and BCV VEMPs elicited using the B-71 bone oscillator and Mini-shaker were recorded and analyzed using the Otometrics Chartr EP 200 system (Natus Medical, Denmark). BCV VEMPs elicited using the impulse hammer were recorded using the Intelligent Hearing Systems 1.30 Opti-Amp differential amplifier (Intelligent Hearing Systems, Miami, FL, USA).

The B-71 bone oscillator was applied to the mastoid using a standard bone conduction headset (5.4 N force) with output set to 59 dBnHL. B-71 placement was not standardized, but rather placed on the mastoid where it was felt to remain stable throughout testing. The Mini-shaker was manually applied to Fz via a cylindrical perspex rod (2.54cm in diameter, 8.89cm in length). Impulse hammer taps were manually presented at Fz at a rate of approximately two taps per second for 30 taps per trial. To reduce variability in impulse hammer tap placement, a bandage was adhered to Fz and used as a target. Data collection software provided concurrent force feedback (in Newtons) to ensure that only taps between 10 – 30 N were accepted for analysis. This is based on previous work in this lab suggesting 10 – 30 Newtons is a comfortable stimulus for both adults and children (Rodriguez, in press). The epoch began when the hammer force exceeded a 2.0 Newton threshold. Using a miniature accelerometer (Model 166 352A24; PCB Corporation, Depew, New York, USA) an ~1 ms delay was measured, thus all latencies were adjusted accordingly.

The peak force level (pFL) produced by each BCV method was measured using an oscilloscope and an Artificial Mastoid (Larson Davis AMC493B) fixed to a sound level meter (Larson Davis model 824). Peak dB FL was calculated using the following equation:

dBpFL=20log(pV/Vcal)+SPLcal+dB conversion, where pV = the peak RMS voltage for each transient stimulus (B-71: 480μv, Mini-shaker: 3.2V; impulse hammer: 2.05V), Vcal = RMS voltage in response to the calibration tone (92 μv), SPLcal = measured output of the calibration tone (113.5 dB SPL) and the dB conversion = the SPL to FL conversion for the artificial mastoid at 500 Hz (4.4 dB) resulting in the following dB pFL for each stimulus: B-71: 132.3 dB pFL, Mini-shaker: 148.7 dB pFL, impulse hammer: 144.8 dB pFL. The impulse hammer also quantifies the force of each tap in Newtons. Force level was further verified using the following equation:

dBFL=20log(Force/1micronewton)

Impulse hammer taps were limited between 10 and 30 Newtons which equates to a range of ~140 – 150 dB pFL, suggesting good agreement between both methods of dB pFL calculation.

2.8. Statistical Analysis

Statistical analyses were performed using SPSS 16.0 (Chicago, IL). Analysis of variance (ANOVA) was applied to investigate differences in response parameters between the BCV methods. The intraclass correlation coefficient (ICC) was calculated to indicate the amount of relative consistency and average agreement between test and retest sessions, while accounting for subject variability and measurement error (ICC = subject variability/ (subject variability + variability in repetition + measurement error)). As described by Koo & Li (2016), an ICC value > 0.9 indicated excellent reliability, values between 0.9 and 0.75 indicated good reliability, values between 0.75 and 0.5 indicated moderate reliability, and values below 0.5 indicated poor reliability. Correlation analyses were completed to determine the relationship between age and VEMP amplitude.

3. Results

3.1. Response Rates Adults, Adolescents, and Children

Response rates for each of the three groups for ACS and each BCV method are shown in Table 1. Of the 10 subjects in the child group, one did not tolerate the B-71 bone oscillator. Therefore, within the child group, nine children remained in the final group analysis for B-71 and 10 children remained in the final group analysis for ACS, impulse hammer and Mini-shaker stimuli.

Table 1.

c- and oVEMP Response Rates Across Age Group.

cVEMP
ACS MS IH B-71
Adult 100% 100% 95% 100%
Adolescent 100% 100% 100% 100%
Child 100% 100% 100% 100%
All 100% 100% 98% 100%
oVEMP
ACS MS IH B-71
Adult 100% 89% 95% 89%
Adolescent 100% 100% 100% 100%
Child 85% 100% 100% 94%
All 95% 96% 98% 94%
Open in a new tab

BCV force levels: MS (148.7 dB pFL), IH (~144.8 dB pFL), B-71 (132.3 dB pFL)

ACS = air conduction stimulation, IH = impulse hammer, MS = Mini-shaker

For cVEMP, response rates were 100% for all methods in all groups with the exception of the adult group in response to the impulse hammer. For oVEMP, response rates varied by group and method. In the adult group, although response rates for ACS were 100%, no BCV method achieved 100% response rates. In the adolescent group, all BCV methods achieved 100% response rates. In the child group, while ACS testing yielded an 85% response rate and the B-71 yielded a 94% response rate, the Mini-shaker and impulse hammer achieved 100% response rates. Across BCV methods for oVEMP, the impulse hammer had the highest response rate (98%) for all groups combined.

3.2. Test-retest Reliability

ICC values were calculated comparing amplitude (cVEMP and oVEMP) and corrected amplitude (cVEMP) for Session 1 to Session 2 for each age group for ACS and each BCV method. This is reflected in Tables 2 (cVEMP) and 3 (oVEMP). Based on the ICC value alone, each BCV method demonstrated good-to-excellent reliability for the subject groups combined except for the B-71 for oVEMP which indicated moderate reliability. The bone conduction method with the greatest reliability and highest response rates varied by age group and VEMP type. For cVEMP, reliability was highest in adults using the Mini-shaker, in adolescents using the impulse hammer, and in children using the B-71. Response rates for each of these BCV methods in these groups were also 100%. For oVEMP, impulse hammer and Mini-shaker demonstrated good-to-excellent reliability for the subject groups combined; however, reliability was highest in the adults and adolescents using the Mini-shaker and in children using the impulse hammer. Response rates for each of these BCV methods in the adolescent and child groups were also 100%, while response rates for the Mini-shaker in the adult group were 89%.

Table 2.

Means (SD), Mean Session Difference p-values, and ICC values of cVEMP Amplitudes across sessions

Amplitude (μv) Corrected Amplitude
Method Session 1 Session 2 p-value ICC value Session 1 Session 2 p-value ICC value
Adult ACS 315.32 (101.82) 332.76 (126.52) 0.198 0.902 2.17 (0.63) 2.14 (0.75) 0.696 0.871
MS 344.10 (140.23) 338.96 (117.34) 0.789 0.871* 2.39 (0.93) 2.31 (0.86) 0.504 0.883*
IH 260.59 (176.89) 262.42 (159.24) 0.964 0.568 0.99 (0.60) 0.96 (0.52) 0.812 0.497
B-71 227.37 (88.51) 229.01 (114.69) 0.935 0.723 1.47 (0.52) 1.43 (0.68) 0.716 0.709
Adolescent ACS 482.63 (159.05) 441.09 (163.21) 0.029 0.915 4.17 (4.45) 2.84 (0.82) 0.177 0.744
MS 539.94 (238.75) 532.66 (250.80) 0.779 0.939 3.57 (1.45) 3.28 (1.09) 0.214 0.788
IH 625.04 (412.36) 635.48 (490.94) 0.784 0.962* 2.10 (1.09) 2.14 (1.18) 0.75 0.925*
B-71 505.41 (225.16) 490.30 (238.70) 0.565 0.929 3.28 (1.30) 3.06 (1.27) 0.295 0.836
Child ACS 331.21 (188.71) 295.21 (157.44) 0.158 0.877 1.90 (1.09) 1.74 (0.96) 0.192 0.928
MS 425.81 (187.86) 476.55 (156.09) 0.113 0.801 2.29 (1.01) 2.70 (0.97) 0.059 0.698
IH 648.08 (297.20) 558.82 (268.34) 0.182 0.639 2.27 (0.97) 2.10 (0.82) 0.332 0.779
B-71 474.81 (212.68) 440.60 (182.54) 0.319 0.854* 2.49 (1.16) 2.43 (1.02) 0.679 0.899*
All ACS 377.80 (166.21) 358.26 (159.92) 0.084 0.913 2.77 (2.85) 2.26 (0.95) 0.126 0.894
MS 436.95 (207.25) 448.54 (199.55) 0.432 0.91 2.76 (1.28) 2.76 (1.04) 0.993 0.824
IH 506.96 (355.16) 483.28 (370.74) 0.396 0.898 1.77 (1.06) 1.72 (1.03) 0.551 0.887
B-71 397.87 (221.72) 383.15 (217.06) 0.327 0.919* 2.41 (1.28) 2.30 (1.22) 0.237 0. 897*
Open in a new tab

Bolded p-value = significance (p < .05); Bolded ICC value = good-to-excellent reliability (.75 – 1.0); Starred (*) ICC value = BCV method with greatest reliability across subject groups

ACS = air conduction stimulation, IH = impulse hammer, MS = Mini-shaker

Table 3.

Means (SD), Mean Differences and ICC Values of oVEMP amplitude across sessions

Amplitude (μv)
Method Session 1 Session 2 p-value ICC value
Adult ACS 23.49 (14.18) 21.91 (16.81) 0.319 0.93
MS 21.86 (14.90) 21.88 (15.91) 0.98 0.986*
IH 28.79 (16.75) 27.81 (20.26) 0.712 0.881
B-71 10.80 (8.08) 14.80 (14.63) 0.233 0.277
Adolescent ACS 20.01 (10.51) 20.12 (11.30) 0.902 0.96
MS 34.20 (19.58) 35.45 (15.00) 0.62 0.878*
IH 36.81 (26.50) 41.53 (18.40) 0.161 0.868
B-71 15.67 (7.05) 20.69 (10.89) 0.007 0.711
Child ACS 12.55 (10.62) 11.62 (9.77) 0.266 0.967
MS 41.87 (13.33) 42.26 (12.78) 0.906 0.54
IH 58.91 (20.38) 52.82 (20.83) 0.081 0.837*
B-71 21.29 (9.89) 21.83 (10.61) 0.791 0.797
All ACS 18.87 (12.59) 18.08 (13.61) 0.239 0.951
MS 32.35 (17.99) 32.91 (16.77) 0.678 0.895*
IH 40.96 (24.76) 40.34 (22.03) 0.736 0.894*
B-71 15.46 (9.24) 19.24 (12.33) 0.008 0.628
Open in a new tab

Bolded p-value = significance (p < .05); Bolded ICC value = good-to-excellent reliability (.75 – 1.0); Starred (*) ICC value = BCV method with greatest reliability across subject groups

ACS = air conduction stimulation, IH = impulse hammer, MS = Mini-shaker

3.3. Relationship Between Age and VEMP Amplitude.

For ACS, age was positively correlated with oVEMP amplitude and was not correlated with cVEMP amplitude or cVEMP corrected amplitude (Figures 2 – 4, Table 4). For BCV, age was negatively correlated with all BCV VEMP amplitudes (B-71, mini-shaker, and impulse hammer) except for cVEMP corrected amplitude in response to the Mini-shaker (Figures 2 – 4, Table 4).

Figure 2.

Figure 2.

Open in a new tab

Scatterplot of cVEMP amplitude versus age in response to ACS and all BCV methods. Age was not related to ACS cVEMP amplitude; however, age negatively correlated with all BCV VEMP amplitudes (B-71, Mini-shaker, and impulse hammer).

Figure 4.

Figure 4.

Open in a new tab

Scatterplot of oVEMP amplitude versus age in response to ACS and all BCV methods. Age positively correlated with ACS oVEMP amplitude, but negatively correlated with all BCV oVEMP amplitudes.

Table 4.

Age and VEMP Amplitude Correlation Analyses

Method Correlation (r)
oVEMP Amp ACS 0.323
MS −0.503
IH −0.463
B-71 −0.471
cVEMP Amp ACS −0.070
MS −0.293
IH −0.492
B-71 −0.564
cVEMP C-Amp ACS 0.029
MS −0.127
IH −0.578
B-71 e−0.507
Open in a new tab

Significant correlations are noted in bold. ACS = air-conducted stimuli, MS = mini-shaker, IH = impulse hammer, Amp = amplitudes, C = corrected.

4. Discussion

4.1. Response Rates and Reliability

While studies have evaluated VEMP in children (e.g., Chou et al., 2012; Zhou et al., 2014), only one study assessed reliability of the impulse hammer (Fuemmeler et al, in press) and none have compared BCV methods for VEMP testing in children. Additionally, few studies (e.g., Iwasaki et al., 2007; Iwasaki et al., 2008; Nguyen et al., 2010) have compared the reliability of BCV methods for VEMP testing in adults. These studies show variable reliability for VEMP testing across BCV methods (Iwasaki et al., 2007; Iwasaki et al., 2008; Nguyen et al., 2010). Higher reliability for oVEMP as compared to cVEMP (Nguyen et al., 2010), and similar reliability for BCV as compared to ACS (Nguyen et al., 2010).

When pediatric patients undergo ACS VEMP testing and the response is absent, it is unclear if the absent response is due to stimulus method, poor test-retest reliability, middle ear pathology, or a true vestibulopathy. To address these potential confounds, the main objective of the current study was to compare response rates and reliability of three BCV methods.

Overall, results varied by age group, VEMP type, and BCV method. This is not surprising as we speculated there would be differences in response rates and reliability between the B-71, impulse hammer, and Mini-shaker due to differences in the frequency specificity and output (i.e., intensity) of these devices. VEMP testing is traditionally completed between 250 and 750 Hz due to tuning properties of the otolith organs (Janky et al. 2009; Park et al. 2010). While the B-71 and Mini-shaker devices allow for frequency specificity within the 250 to 750 Hz range, the impulse hammer lacks frequency specificity and is reported to produce a broadband stimulus between 100 to 450 Hz with the soft tip utilized in the current study (PCB Piezotronics, Inc., 2007). Additionally, because the impulse hammer is manually applied, the level of force delivered is variable per tap. While our recording system did control for force, the accepted force range was still broad (i.e., 140 – 150 dB pFL), thereby variability in the response is still a risk. Regarding output, the maximum output of the B-71 at 500 Hz is approximately 59 dBnHL (132 dB pFL) compared to 128 to 147 dB FL (Nguyen et al. 2010; Rosengren et al. 2011; Chou et al. 2012; Tseng et al. 2012; Janky et al. 2013; Kinoshita et al. 2013; Manzari et al. 2013; Taylor et al. 2014; Holmeslet et al. 2015; Govender et al., 2016; Govender & Colebatch, 2017) and 148.7 dB pFL (current study) for the Mini-shaker. For these reasons, it was hypothesized that the Mini-shaker would yield the highest response rates and reliability because it provides frequency specificity, higher output levels without distortion as compared to the B-71, and the most consistent force output as compared to the impulse hammer.

Many have reported greater efficiency of BCV compared to ACS for eliciting the VEMP response (Welgampola et al., 2003; Wang et al., 2010; Curthoys et al., 2011; Chou et al., 2012; Wackym et al., 2012). For cVEMP, response rates were the same and ICC values were comparable between ACS and all BCV methods across all age groups. For oVEMP, Mini-shaker and impulse hammer response rates were higher in children and ACS response rates were higher in adults. ICC values were comparable across groups with the exception of the child group.

4.2. Recommendations for Adults

In the adult group, reliability varied between c- and oVEMP responses. cVEMP reliability was highest using the Mini-shaker and B-71, while reliability for impulse hammer stimulation was moderate-to-poor (Table 2). We speculate this is a result of greater variability for manually applied stimulation. The impulse hammer recording software was designed to decrease variability associated with a manually applied stimulus, however, stimulus intensity varied across a range of 10 dB pFL. This variability was not present in the B-71 or Mini-shaker.

oVEMP reliability in adults was excellent using the Mini-shaker, good using the impulse hammer, and considerably poor using the B-71 (Table 3). These results are consistent with Nguyen et al. (2010), which reported good reliability of oVEMP amplitudes in response to impulse hammer and Mini-shaker stimuli in adults, and Iwasaki et al. (2007), which reported poorer reliability for oVEMP amplitudes in response to B-71 stimulation as compared to impulse hammer and Mini-shaker stimulation in adults.

The lower oVEMP response rates across groups suggest that oVEMPs may require a higher stimulus level for BCV stimulation (Table 1). Comparing the same three BCV methods utilized in the current study in adults only, Rosengren et al. (2011) reported a higher threshold for BCV oVEMP than cVEMP, specifying that B-71 stimulation delivered at the mastoid was least effective in evoking oVEMPs compared to cVEMPs. This would also explain the lower reliability using the B-71. In addition, oVEMP response rates for BCV and ACS varied widely, while cVEMP rates were more consistent (Rosengren et al., 2011). This is consistent with the current study as response rates varied more widely for oVEMP than cVEMP (Table 3).

Overall, all BCV methods resulted in consistent c- and oVEMP responses (response rates 89 – 100%) with at least moderate reliability (ICC ≥ 0.5) except for oVEMP testing using the B-71 on the mastoid. Given that response rates were lower, reliability was exceptionally poor, and there was a consistent relationship with lower amplitudes and age with the B-71, use of the B-71 on the mastoid is not recommended for oVEMP testing in adults. While the current study did not assess B-71 response rates and reliability at various sites on the head, forehead placement also does not provide adequate stimulus intensity to produce reliable VEMP responses (Iwasaki et al., 2007). This finding is problematic, as most clinics have greater access to a B-71. However, it is also worth noting that ACS achieved excellent (100%) response rates and excellent reliability for both cVEMP and oVEMP in the adult group and remains a reliable VEMP method for this age group.

4.3. Recommendations for Adolescents and Children

Within the adolescent and child groups ages 5 – 18, response rates were 100% for each BCV method across c- and oVEMP with the exception of the B-71 oVEMP in the child group where response rates were 94%. Additionally, all c- and oVEMP responses had at least moderate reliability (ICC ≥ 0.5). These results suggest that all BCV methods are sufficient to elicit a cVEMP response and that ACS and BCV VEMP can be reliably performed in the pediatric population.

Similar to the adult group, reliability in the adolescent and child groups for the various BCV methods varied between c- and oVEMP responses. In the adolescent group, cVEMP reliability was highest using the impulse hammer and oVEMP reliability was highest using the Mini-shaker or impulse hammer. In the child group, cVEMP reliability was highest using the B-71 and oVEMP reliability was highest using the impulse hammer. In addition, ACS remained a reliable method across age groups (Tables 2 and 3).

The advantage of using either an impulse hammer or Mini-shaker is that right and left responses can be obtained simultaneously, reducing testing time. For the child group (5 –10 years) the B-71 yielded the highest ICC value for BCV cVEMP. However, the ICC value for Mini-shaker (cVEMP raw amplitude) and impulse hammer (cVEMP corrected amplitude) were also in the good-to-excellent range and elicited 100% response rates, suggesting any method of BCV is recommended for cVEMP in children. For oVEMP in children, the impulse hammer elicited the best response rates and reliability. Similar to adults, it is worth noting that ACS achieved excellent (100%) response rates and good-to- excellent reliability for both cVEMP and oVEMP in the adolescent and child groups and remains a reliable VEMP method for all age groups.

4.4. Relationship Between Age and VEMP Amplitude.

Historically, there has been a trend that VEMP amplitudes (Welgampola and Colebatch, 2001; Su et al., 2004; Nguyen et al., 2010; Rosengren et al., 2011) and response rates (Janky & Shepard, 2009) decrease with age. However, effects of age on VEMP amplitudes have been most pronounced in adults greater than 40 years of age (Taylor et al., 2012). Therefore, the significant downward trend of BCV VEMP amplitude with age in our young population (range 5 – 39 years) was surprising. Additionally, the opposite effect of age on ACS oVEMP amplitude was observed. Age was positively correlated with ACS oVEMP amplitude and was not correlated with ACS cVEMP amplitude or cVEMP corrected amplitude (Figures 2– 4). We hypothesize the differences in relationship between BCV and ACS with age could be related to differences in skull bone density, changes in mechanical impedance, or smaller skull size noted in children. However, further work is needed to explore the relationship between bone density, skull size, and BCV VEMP amplitudes.

5. Conclusion

The current study suggests BCV methods for VEMP testing are reliable in young adults, adolescents, and children. Based on the results of the current study, all BCV methods resulted in consistent cVEMP responses (response rates 95 – 100%) with at least moderate reliability (ICC ≥ 0.5). Similarly, all BCV methods resulted in consistent oVEMP responses (89 – 100%) with at least moderate reliability (ICC ≥ 0.5) except for the B-71 in adults.

Figure 3.

Figure 3.

Open in a new tab

Scatterplot of cVEMP corrected amplitude versus age in response to ACS and all BCV methods. Age was not related to ACS cVEMP corrected amplitude; however, age negatively correlated with B-71 and impulse hammer cVEMP corrected amplitude, but not Mini-shaker.

Footnotes

Financial disclosures/conflicts of interest: Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under award number P20GM109023 and by the National Institute on Deafness and Other Communication Disorders under award numbers T35 DC 008757, 5T32DC00013-36, and R03DC015318

REFERENCES

  1. American Speech-Language-Hearing Association. (1997). Guidelines for Audiologic Screening. Retrieved from doi: 10.1044/policy.GL1997-00199. [DOI] [Google Scholar]
  2. Bath A, Harris N, McEwan J. (1999). Effect of conductive hearing loss on the vestibulo-collic response. Clinical Otolaryngology, 24: 181–3. [DOI] [PubMed] [Google Scholar]
  3. Bogle J, Zapala D, Criter R, Burkard R. (2013). The effect of muscle contraction level on the cervical vestibular evoked myogenic potential (cVEMP): usefulness of amplitude normalization. Journal of the American Academy of Audiology, 24(2):77–88. [DOI] [PubMed] [Google Scholar]
  4. British Society of Audiology (2013) Recommended Procedure: Tympanometry. Retrieved from http://www.thebsa.org.uk/wp-content/uploads/2014/04/BSA_RP_Tymp_Final_21Aug13_Final.pdf
  5. Chou C, Hsu W, Young Y. (2012). Ocular vestibular-evoked myogenic potentials via bone conducted vibration in children. Clinical Neurophysiology, 123: 1880–1885. [DOI] [PubMed] [Google Scholar]
  6. Colebatch J, Halmagyi G. (1992). Vestibular evoked potentials in human neck muscles before and after unilateral vestibular deafferentation. Neurology, 42: 1635–1636. [DOI] [PubMed] [Google Scholar]
  7. Colebatch J, Halmagyi G, Skuse N. (1994). Myogenic potentials generated by a click-evoked vestibulocollic reflex. Journal of Neurology, Neurosurgery, & Psychiatry, 57:190–197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Curthoys I, Vulovic V, Burgess A, Cornell E, Mezey L, MacDougall H, Manzari L, McGarvie L. (2011). The basis for using bone-conducted vibration or air-conducted sound to test otolithic function. Annals of The New York Academy of Sciences, 1233:231–241. [DOI] [PubMed] [Google Scholar]
  9. Curthoys I, Kim J, McPhedran S, Camp A. (2006). Bone conducted vibration selectively activates irregular primary otolithic vestibular neurons in the guinea pig. Experimental Brain Research, 175(2):256–67. [DOI] [PubMed] [Google Scholar]
  10. Fuemmeler E, Rodriguez AI, Thomas MLA, Creutz T, Fitzpatrick D, Janky KL (in press). Vestibular evoked myogenic potential (VEMP) test-retest reliability in children. Otology & Neurotology. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Govender S, Cheng P, Dennis D, Colebatch J. (2016). Electrode montage and gaze effects on ocular vestibular evoked myogenic potentials (oVEMPs). Clinical Neurophysiology, 127(8): 2846–2854. [DOI] [PubMed] [Google Scholar]
  12. Govender S & Colebatch J. (2017) Location and phase effects for ocular and cervical vestibular-evoked myogenic potentials evoked by bone-conducted stimuli at midline skull sites. Journal of Neurophysiology, 119: 1045–1056. [DOI] [PubMed] [Google Scholar]
  13. Halmagyi G, Colebatch J, Curthoys I. (1994). New tests of vestibular function. Clinical Neurology, 3(3): 485–500. [PubMed] [Google Scholar]
  14. Halmagyi G, Yavor R, Colebatch J. (1995). Tapping the head activates the vestibular system: a new use for the clinical reflex hammer. Neurology, 45:1927–1929. [DOI] [PubMed] [Google Scholar]
  15. Holmeslet B Foss O, Bugten V, Brantberg K. (2015). Ocular vestibular-evoked myogenic potentials (oVEMPs) in response to bone-conducted vertex vibration. Clinical Neurophysiology, 126(3): 608–613. [DOI] [PubMed] [Google Scholar]
  16. Iwasaki S, Chihara Y, Smulders Y, Burgess A, Halmagyi G, Curthoys I. (2009). The role of the superior vestibular nerve in generating ocular vestibular-evoked myogenic potentials to bone conducted vibration at Fz. Clinical Neurophysiology, 120(3):588–93. [DOI] [PubMed] [Google Scholar]
  17. Iwasaki S, McGarvie L, Halmagyi G, Burgess A, Kim J, Colebatch J. (2007). Head taps evoke a crossed vestibulo-ocular reflex. Neurology, 68(15):1227–1229. [DOI] [PubMed] [Google Scholar]
  18. Iwasaki S, Smulders Y, Burgess A, McGarvie L, MacDougall H, Halmagyi G. (2008). Ocular vestibular evoked myogenic potentials to bone conducted vibration of the midline forehead at Fz in healthy subjects. Clinical Neurophysiology, 119: 2135–47. [DOI] [PubMed] [Google Scholar]
  19. Janky K & Shepard N. (2009). Vestibular evoked myogenic potential (VEMP) testing: normative threshold response curves and effects of age. Journal of the American Academy of Audiology, 20(8): 514–522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Janky K, Nguyen K, Welgampola M, Zuniga M, Carey J. (2013). Air-conducted oVEMPs provide the best separation between intact and superior canal dehiscent labyrinths. Otology & Neurotology, 34(1): 127–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kinoshita M, Iawasaki I, Fujimoto C, Inoue A, Egami N, Chihara Y, Ushio M, Yamasoba T. (2013). Ocular vestibular evoked myogenic potentials in response to air-conducted and bone-conducted vibration in vestibular schwannoma. Otology & Neurotology, 34(7): 1342–1348. [DOI] [PubMed] [Google Scholar]
  22. Koo T, Li M. (2016). A guideline of selecting and reporting intraclass correlation coefficients for reliability research. Journal of Chiropractic Medicine, 15(2): 155–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kuhn J, Lavendar V, Hunter L, McGuire S, Meinzen-Derr J, Keith R, Greinwald J. (2018). Ocular vestibular evoked myogenic potentials: normative findings in children. The Journal of the American Academy of Audiology, 29(5): 443–450. [DOI] [PubMed] [Google Scholar]
  24. Maes L, De Kegel A, Waelvelde H, De Leenheer E, Van Hoecke H, Goderis J, Dhooge I. (2016). Comparisons of the Motor Performance and Vestibular Function in Infants with a Congenital Cytomegalovirus Infection or a Connexin 26 Mutation: A Preliminary Study. Ear & Hearing, X(XX): 1–7. [DOI] [PubMed] [Google Scholar]
  25. Manzari L, Burgess A, McGarvie L, Curthoys I. (2013). An Indicator of Probably Semicircular Canal Dehiscence: Ocular Vestibular Evoked Myogenic Potentials to High Frequencies. Otolaryngology – Head and Neck Surgery, XX(X): 1–4. [DOI] [PubMed] [Google Scholar]
  26. McCaslin D, Jacobson G, Hatton K, Fowler A, Delong A. (2013). The effects of amplitude normalization and EMG targets on cVEMP interaural amplitude asymmetry. Ear & Hearing, 1–9. [DOI] [PubMed] [Google Scholar]
  27. McCaslin D, Folwer A, Jacobson G. (2014). Amplitude normalization reduces cervical vestibular evoked myogenic potential (cVEMP) amplitude asymmetries in normal subjects: proof of concept. Journal of the Academy of Audiology, 25(3):268–277. [DOI] [PubMed] [Google Scholar]
  28. Nguyen K, Welgampola M, Carey J. (2010). Test–retest reliability and age-related characteristics of the ocular and cervical vestibular evoked myogenic potential tests. Otology & Neurotology, 31:793–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Park H, Lee I, Shin J, Lee Y, Park M. (2010). Frequency-tuning characteristics of cervical and ocular vestibular evoked myogenic potentials induced by air-conducted tone bursts. Clinical Neurophysiology, 121(1):85–89. [DOI] [PubMed] [Google Scholar]
  30. Piezotronics PCB, Inc. (2007). Model 086C03 ICP Impact Hammer Installation and Operating Manual. Retrieved from: http://www.pcb.com/products.aspx?m=086C03
  31. Picciotti P, Fiorta A, DiNardo W, Calo L, Scarno E, Paludetti G. (2007) Vestibular evoked myogenic potentials in children. International Journal of Pediatric Otorhinolaryngology, 71(1): 29–33. [DOI] [PubMed] [Google Scholar]
  32. Rodriguez AI, Marler E, Fitzpatrick D, Creutz T, Cannon SA, Thomas MLA, Janky KL (in press). Optimization of Cervical and Ocular Vestibular Evoked Myogenic Potential Testing Using an Impulse Hammer in Adults, Adolescents, and Children. Otology & Neurotology. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rodriguez A, Thomas M, Fitzpatrick D, Janky K. (2018). Effects of high sound exposure during air-conducted vestibular evoked myogenic potential testing in children and young adults. Ear & Hearing, 39(2): 269–277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rosengren S, Govender S, Colebatch J. (2011). Ocular and cervical vestibular evoked myogenic potentials produced by air- and bone- conducted stimuli: comparative properties and effects of age. Clinical Neurophysiology, 122: 2282–2289. [DOI] [PubMed] [Google Scholar]
  35. Rosengren S, McAgnus T, Colebatch J. (2005). Vestibular-evoked extraocular potentials produced by stimulation with bone-conducted sound. Clinical Neurophysiology, 116(8): 1938–1948. [DOI] [PubMed] [Google Scholar]
  36. Su H, Huang T, Young Y, Cheng P. (2004). Aging effect on vestibular evoked myogenic potential. Otology & Neurotology, 25(6): 977–980. [DOI] [PubMed] [Google Scholar]
  37. Taylor R, Bradshaw A, Halmagyi G, Welgampola M. (2012). Effects of age on the tuning of the cVEMP and oVEMP. Audiology & Neurotology, 17: 207–218. [DOI] [PubMed] [Google Scholar]
  38. Taylor R, Vlaivie C, Bom A, Holmeslet B, Pansell T, Brantberg K, Welgampola M. (2014). Ocular vestibular-evoked myogenic potential (oVEMP) to skull taps in normal and dehiscent ears: mechanisms and markers of superior canal dehiscence. Experiments in Brain Research, 232: 1073–1084. [DOI] [PubMed] [Google Scholar]
  39. Todd N, Cody F, Banks J. (2000). A saccular origin of frequency tuning in myogenic vestibular evoked potentials? Implications for human responses to loud sounds. Hearing Research, 141(1–2):180–188. [DOI] [PubMed] [Google Scholar]
  40. Tseng C, Wang S, Young Y. (2012). Comparison of bone-conducted vibration for eliciting ocular vestibular-evoked myogenic potentials: forehead versus mastoid tapping. Otolaryngology Head-Neck Surgery, 146:289–94. [DOI] [PubMed] [Google Scholar]
  41. Wackym A, Ratigan J, Birck J, Johnson S, Doornink J, Bottlang M, Gardiner S, Black O. (2012). Rapid cVEMP and oVEMP Response Elicited by a Novel Head Striker and Recording Device. Otology & Neurology, 33(8): 1392–1400. [DOI] [PubMed] [Google Scholar]
  42. Wang S, Hsieh W, Young Y. (2013). Development of ocular vestibular-evoked myogenic potentials in small children. Laryngoscope, 123(2): 512–517. [DOI] [PubMed] [Google Scholar]
  43. Wang S, Weng W, Jaw F, Young Y. (2010). Ocular and cervical vestibular-evoked myogenic potentials: a study to determine whether air- or bone-conducted stimuli are optimal. Ear & Hearing, 31: 283–8. [DOI] [PubMed] [Google Scholar]
  44. Welgampola M & Colebatch J. (2001). Vestibulocollic reflexes: normal values and the effect of age. Clinical Neurophysiology, 112: 1971–1979. [DOI] [PubMed] [Google Scholar]
  45. Welgampola M, Rosengren S, Halmagyi G, Colebatch J. (2003). Vestibular activation by bone conducted sound. Journal of Neurology, Neurosurgery, and Psychiatry, 74(6): 771–778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Yang C, Lavender V, Meinzen-Derr J, Cohen A, Youssif M, Castiglione M, Manickam V, Bachmann K, Greinwald H. (2016). Vestibular pathology in children with enlarged vestibular aqueduct. Laryngoscope, 126: 2344–2350. [DOI] [PubMed] [Google Scholar]
  47. Zhou G, Dargie J, Dornan B, Whittemore K. (2014). Clinical uses of cervical vestibular-evoked myogenic potential testing in pediatric patients. Medicine, 93(4): 1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES