- Email: [email protected]
Ocular vestibular-evoked myogenic potentials via bone-conducted vibration in children
Clinical Neurophysiology 123 (2012) 1880–1885
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Ocular vestibular-evoked myogenic potentials via bone-conducted vibration in children Chen-Han Chou, Wei-Chung Hsu, Yi-Ho Young ⇑ Department of Otolaryngology, National Taiwan University Hospital, Taipei, Taiwan
a r t i c l e
i n f o
Article history: Accepted 6 February 2012 Available online 3 March 2012 Keywords: Acceleration magnitude Bone-conducted vibration Children Ocular vestibular evoked myogenic potential
h i g h l i g h t s The characteristic parameters (latencies and amplitude) of oVEMPs did not significantly differ between children >3 years and adults. The mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, a non-significant difference. When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters unaffected by structural factors.
a b s t r a c t Objective: This study utilized bone-conducted vibration (BCV) stimuli for eliciting ocular vestibularevoked myogenic potentials (oVEMPs) to assess the vestibulo-ocular reflex (VOR) system in healthy children and adults. Methods: Fifteen healthy children aged 3–14 years and 18 healthy adults aged 24–28 years underwent oVEMP testing. Structural factors such as body weight, body height and body mass index were measured for each healthy subject. Results: All healthy children and adults presented clear oVEMPs, bilaterally. The characteristic parameters (latencies and amplitude) of oVEMPs did not significantly differ between children and adults. The mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, a non-significant difference. As stimulation intensity increased stepwise, interaural acceleration magnitude increased correspondingly, leading to early nI latency and large nI– pI amplitude of oVEMPs. However, no structural factor was statistically correlated with interaural acceleration magnitude. Conclusion: When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters (latencies and amplitude) unaffected by structural factors. Significance: Establishing the norm of oVEMP is essential for diagnosing VOR deficit in children aged >3 years. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction The vestibular system is anatomically developed and functionally responsive at birth (Jeffery and Spoor, 2004). Although a serial test of postural control is used to assess the balance function for small children, poor postural control is not specific for vestibular ⇑ Corresponding author. Address: Department of Otolaryngology, National Taiwan University Hospital, 1 Chang-te St., Taipei, Taiwan. Tel.: +886 2 23123456x65221; fax: +886 2 2394 6674. E-mail address: [email protected] (Y.-H. Young).
deficits. Alternatively, caloric and rotational chair tests have been applied to evaluate the vestibulo-ocular reflex (VOR) system in small children, which normalizes at 2 months and matures further during the first 2 years of life (Fife et al., 2000). However, the caloric test is poorly tolerated by small children, as a rotational chair is not always available in each laboratory. Additionally, both tests fail to evaluate the otolithic (utricular and saccular) function. Over the last decade, the cervical vestibular evoked myogenic potential (cVEMP) test using air-conducted sound (ACS) stimulation has been widely applied to investigate saccular disorders (Welgampola and Colebatch, 2005; Young, 2006). The cVEMP is
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2012.02.059
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
generated from the saccular macula to the sternocleidomastoid muscle via the inferior vestibular nerve. Children with benign paroxysmal vertigo, the so-called migraine equivalent (Fenichel, 1967; Lewis, 2007), may present delayed cVEMPs, indicating a retrolabyrinthine lesion along the sacculo-collic reflex pathway, which descends via the lower brainstem (Lin et al., 2010). Likewise, the recently developed ocular VEMP (oVEMP) test has been applied to assess the utricular function predominantly and the crossed VOR pathway, and to investigate vestibular disorders, i.e. vestibular neuritis, Menieres’ disease, and vestibular schwannoma (Iwasaki et al., 2008; Huang et al., 2011; Curthoys, 2010). Previously, we have reported that the characteristic parameters of oVEMP using ACS stimulation in children aged >3 years did not differ significantly from those of adults (Hsu et al., 2009). However, ACS stimuli demonstrate lower response rates than boneconducted vibration (BCV) stimuli (Curthoys et al., 2011; Wang et al., 2010). Hence, this study conducted oVEMP testing using BCV stimulation coupled with triaxial accelerometry to compare oVEMPs between healthy children and adults. 2. Subjects and methods 2.1. Subjects Fifteen healthy children (11 males and 4 females; aged 3– 14 years; mean, 9 years) and 18 healthy adults (15 males and 3 females; aged 24–28 years; mean, 26 years) were enrolled in this study. All subjects denied previous ear disorders, and were further checked with otoscope. Each subject underwent oVEMP test via BCV stimuli combined with triaxial accelerometry. Structural factors such as body weight, body height and body mass index (BMI) were also measured. This study was approved by the institutional review board of the College of Medicine, National Taiwan University Hospital, and each child’s parent and each adult signed the informed consent to participate. 2.2. oVEMP test The subject was placed in a sitting position. Surface potentials, predominantly electromyographic (EMG) activities, were recorded (Smart EP 3.90, Intelligent Hearing Systems, Miami, FL, USA). Two active electrodes were placed around 1 cm below the center of the two lower eyelids. The other two reference electrodes were positioned about 1–2 cm below the active ones, and one ground electrode was placed on the sternum. During recording, the subject was instructed to look upward at a small fixed target 2 m high from the eyes, with a vertical visual angle of approximately 30° above horizontal. The EMG signals were amplified and bandpass filtered between 1 and 1000 Hz. The stimulation rate was 5 Hz. The duration of analysis of each response was 50 ms, and 30 responses were averaged for each run. BCV stimuli were delivered using a hand-held electromechanical vibrator (V201 shaker, Ling Dynamic Systems, Royston, England) that was fitted with a short M4 bolt (2 cm in length) terminated in a bakelite cap (1.6 cm in diameter). The input signal was one half cycle of 500-Hz sine wave driven by a custom amplifier. The initial peak driving voltage was 2 dB re 10 V, approximately 128 dB force level (FL). The stimulation voltage was then decreased in steps of 5 dB until no oVEMP waveform was present. The operator held the vibrator by hand and supported most of its weight, maintaining constant coupling force during each session, such that the axis of the connected bakelite cap perpendicularly delivered a repeatable tap with little pressure on the subject’s skull at Fz (the midline forehead at the hairline). Bilateral
1881
oVEMPs were recorded simultaneously. The initial negative–positive biphasic waveform comprised peaks nI and pI. The latencies of peaks nI and pI, and the nI–pI amplitude were measured. The asymmetry ratio (%) was defined as the difference of the amplitude nI–pI on each ear divided by the sum of amplitude nI–pI of both ears, that is, (larger amplitude smaller amplitude/larger amplitude + smaller amplitude) 100. The lowest stimulus intensity that could elicit clear oVEMPs was determined (Wang et al., 2010). 2.3. Triaxial accelerometry One miniature (10 mm3, 5 gm) triaxial accelerometer (Endevco model 65–100, San Juan Capistrano, CA, USA) accompanied by a tight elastic bandage surrounding the head girth was fixed to each subject’s left mastoid area just behind the auricle (Tseng et al., 2012). Thus, the accelerometer was attached to the mastoid bone tightly. The accelerometer was used to measure the three-dimensional linear acceleration along the x-axis (anterior– posterior), y-axis (inter-aural), and z-axis (rostro-caudal) simultaneously. Forward, outward, and upward were defined as the positive x, y, and z directions, respectively. The voltage sensitivity was 100 mV/g and the frequency response was 1.5–6000 Hz (±1 dB). The triaxial accelerometer signals were digitalized at 10 kHz using a dynamic signal analyzer (NI USB-4431, National Instruments, Austin, Texas, USA), which provided signal conditioning and constant current supply to the accelerometer. The initial peak magnitudes of linear acceleration were analyzed using a customized program (LabVIEW 8.5, National Instruments, Austin, Texas, USA) during oVEMP testing. 2.4. Statistical methods Since the highest acceleration magnitude along the y-axis was observed, as compared with that along the x- and z-axis, the acceleration magnitudes along the y-axis in response to various driving voltages of the vibrator were compared by one-way repeated-measures analysis of variance (ANOVA) and Bonferroni-adjusted t-test. The difference in the acceleration magnitudes between the children and adult groups were compared by non-paired t-test. The relationships between the acceleration magnitude and oVEMP parameters, between the age and oVEMP parameters, and between the acceleration magnitude and structural factors, were examined by Pearson correlation analysis. A difference of p < 0.05 is regarded as significant. 3. Results 3.1. oVEMPs: healthy children vs. healthy adults All healthy children and adults presented clear oVEMPs bilaterally when stimulated by a vibrator with a driving voltage of 2 dB re 10 V (Fig. 1). Mean nI and pI latencies, nI–pI amplitude and asymmetry ratio in children were 8.0 ± 0.7 (mean ± SD) ms, 12.2 ± 1.5 ms, 16.1 ± 9.0 lV, and 12 ± 14%, respectively. These data did not differ significantly from those in adults (p > 0.05, non-paired t-test, Table 1). Mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, an insignificant difference (p > 0.05, Table 1). 3.2. Acceleration magnitude vs. oVEMP parameters Mean interaural acceleration magnitudes in both children and adults were 0.37 g for a driving voltage of 2 dB re 10 V, 0.18 g for 7 dB re 10 V, and 0.08 g for 12 dB re 10 V, showing a significant difference (p < 0.05, one way ANOVA and Bonferroni-adjusted
1882
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
Fig. 1. Clear oVEMPs are observed in an 8-year-old boy and a 28-year-old man.
t-test). Restated, as stimulation intensity decreased stepwise, interaural acceleration magnitude declined correspondingly, resulting in a serial reduction in nI–pI amplitude (Fig. 2). Additionally, the lowest acceleration magnitude for evoking oVEMPs in children was 0.14 ± 0.08 g, which did not significantly differ from 0.16 ± 0.08 g for adults (p > 0.05, non-paired t-test). Analyses of acceleration magnitude and characteristic parameters indicated that the nI latency and interaural acceleration magnitude were negatively correlated (r = 0.426, p = 0.003, Pearson correlation analysis, Fig. 3A), and the nI–pI amplitude and interaural acceleration magnitude were positively correlated (r = 0.315, p = 0.029, Fig. 3C). However, pI latency was not correlated with interaural acceleration magnitude (r = 0.184, p = 0.210, Fig. 3B). Thus, as interaural acceleration magnitude increased, early nI latency and large nI–pI amplitude were anticipated. 3.3. Acceleration magnitude vs. structural factors Structural factors such as the mean body height, body weight, and BMI were 127 ± 18 cm, 31 ± 14 kg, and 18 ± 4 kg/m2 for the child group, and 173 ± 6 cm, 65 ± 9 kg, and 22 ± 2 kg/m2 for the
adult group, respectively. Although significant differences existed in the structural factors between the children and adults, none of them was statistically correlated with interaural acceleration magnitude (p > 0.05, Pearson correlation analysis). 3.4. Effects of age and gender on oVEMPs To investigate the effect of age on oVEMPs, data of oVEMP responses from both adults and children were pooled together. Analyses of the age factor and oVEMP parameters revealed no correlations between the age and nI latency (r = 0.133, p = 0.313, Pearson correlation analysis), pI latency (r = 0.015, p = 0.909), and nI–pI amplitude (r = 0.067, p = 0.614, Fig. 4). Since only seven females were included in the study sample, another seven age-matched male subjects were also enrolled for comparison. The mean age of each group (including four children and three adults) was 15 ± 8 years. The mean BMI for the male group was 21.6 ± 4.0 kg/m2, which did not significantly differ from 20.5 ± 0.5 kg/m2 for the female group (p > 0.05, unpaired t -test). The mean nI latency, pI latency, and nI–pI amplitude in male group were 8.0 ± 0.7 ms, 12.6 ± 1.6 ms, and 18.6 ± 7.3 lV, respectively,
Table 1 Comparison of characteristic parameters of oVEMPs between children and adults.
Children Adults
N (ears)
nI latency (ms)
pI latency (ms)
nI–pI interval (ms)
nI–pI amplitude (lV)
Asymmetry ratio (%)
Acceleration magnitude (g)
30 36
8.0 ± 0.7 8.4 ± 0.5 (NS)
12.2 ± 1.5 12.7 ± 1.2 (NS)
4.3 ± 1.1 4.3 ± 1.0 (NS)
16.1 ± 9.0 15.2 ± 8.3 (NS)
12 ± 14 13 ± 14 (NS)
0.37 ± 0.12 0.41 ± 0.20 (NS)
Data are expressed as mean ± SD; NS: non-significant difference, p > 0.05, non-paired t-test.
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
1883
Fig. 2. Clear oVEMPs in a 28-year-old man in response to BCV stimuli. The nI–pI amplitude declines as the interaural acceleration magnitude reduces from 0.37 to 0.09 g.
Fig. 3. Significant correlations exist between interaural acceleration magnitude and nI latency (A), nI–pI amplitude (C), but not pI latency (B).
exhibiting non-significant difference when compared with the respective 8.4 ± 0.6 ms, 12.3 ± 1.1 ms, and 15.3 ± 7.4 lV in female group (p > 0.05, unpaired t-test). 4. Discussion We previously reported that abnormal (prolonged or absent) cVEMPs were elicited in 60% of newborns aged 2–5 days, indicating that great variation exists in the maturation of the sacculo-collic reflex in early life (Chen et al., 2007; Young et al., 2009). As a child grows, neck length and cVEMP latencies become strongly correlated, suggesting that structural factors affect the cVEMPs (Chang et al., 2007). Thus, each laboratory should establish a cVEMP norm for children of different ages. Likewise, before oVEMP test can be applied to children for investigating VOR function, one must optimize its methodology and establish the norms for children. 4.1. oVEMPs: children vs. adults The characteristic parameters (e.g. latencies and amplitude) of oVEMPs via BCV stimuli for children aged 3–14 years were not
significantly different from those of adults aged 24–28 years (Fig. 1, Table 1). Further, no correlation exists between the age factor and oVEMP parameters (Fig. 4), which is in agreement with our previous report that oVEMPs by ACS stimuli have developed at 3 years of age without statistical difference in oVEMP parameters between children and adults (Hsu et al., 2009). Opposed to our previous report that gender difference exists in oVEMP amplitude, which is attributed to variance in the muscle bulk between male and female adults (Sung et al., 2011), this study showed no gender difference in oVEMP parameters. The reason is probably because no significant difference in BMI between boys and girls, leading to no effect of gender on oVEMP parameters in children. As the ACS stimuli primarily activate the saccule (Murofushi et al., 1995), whereas both saccule and utricle are stimulated by BCV stimuli (Curthoys, 2010), indicating that BCV mode has relatively higher stimulus intensity than ACS mode in clinical testing. However, measurements of vibratory force from a vibrator cannot reflect acceleration magnitude because the same force level may generate different acceleration outputs for subjects. Hence, one must measure acceleration magnitude during oVEMP testing.
1884
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
Fig. 4. Analyses of the age factor and oVEMP parameters reveal no correlations between the age and nI latency (A), pI latency (B), and nI–pI amplitude (C).
4.2. Acceleration magnitude vs. oVEMP parameters Acceleration magnitudes along three perpendicular axes can be measured by a triaxial accelerometer. Among three axes, interaural (y-axis) acceleration magnitude correlates strongly with nI latency and nI–pI amplitude (Fig. 3), and is therefore utilized in subsequent analysis in this study. Since oVEMP is an excitatory response, increased BCV stimulation intensity may evoke oVEMPs of early latencies and large amplitude (Fig. 2). However, under the same stimulation intensity, mean interaural acceleration magnitude between children and adults did not differ significantly (Table 1). Even the lowest interaural acceleration magnitude required to elicit oVEMPs is similar for children and adults, likely explaining why characteristic parameters of oVEMPs did not significantly differ between these two groups. However, one may question why acceleration magnitude did not differ given that the structure of children and adults is different.
4.3. Structural factor vs. acceleration magnitude Through Pearson correlation analysis, the structural factor, i.e. body height, body weight, and BMI, did not correlate with interaural acceleration magnitudes, which may explain why oVEMP parameters did not differ significantly between children (aged >3 years) and adults at the same stimulus intensity. Two hypotheses are proposed to account for this phenomenon. First, a child’s skull is complex. This complexity is magnified by numerous sutures, synchondroses, and various degrees of ossification. As age increases, there is a progressive rise in the stage of sutural closure. Most children do not develop ossification of suture before the age of 15 years, and suture closure completes at a mean age of 33 years (Harth et al., 2009). Thus, a wide cranial suture in children may hinder vibration propagation. Second, the human skull consists of a layer of cancellous bone sandwiched between two layers of compact bone. Compared with adults, the skulls of children have more cancellous bone and less compact bone. The cancellous bone markedly attenuates energy propagation via the scattering effect at the interface between its two components – bony spicules and soft tissues. Furthermore, propagation time differs when energy transits in these two components and the resulting interference may have a cancelling effect (White et al., 1978). That is, the wide cranial suture and large
amount of cancellous bone in a child’s skull may cancel increased acceleration as related to a small head size and mass. 4.4. Clinical relevance Since the inferior oblique muscle is the most superficial extraocular muscle that transverses to the electrode recording site, oVEMPs can be obtained easily from the skin surface beneath the eyes (Iwasaki et al., 2008). Thus, the amplitude of oVEMPs increases when subjects gaze upward. However, small children are usually intolerant to maintain their upward gaze. Alternatively, they usually elevate the head instead, which deteriorates the oVEMP waveform or even generates no response (Hsu et al., 2009). To ensure the stability of upward gaze, small child was instructed to look at a toy with some clicking sounds held by the parent about 1 m high from the child’s eyes, while another parent gently fixed the child’s head. When pediatric subjects underwent oVEMP testing, the results can be interpreted based on the norm of nI latency (8.0 ± 0.7 ms) for those aged >3 years. Accordingly, children with nI latency >9.4 ms are defined as having delayed oVEMPs, indicating a brainstem lesion, as evidenced by a recent report of delayed oVEMPs in a 5 year-old boy with metastatic neuroblastoma on the left cerebellopontine angle (Chen et al., 2012). However, when an adult aged >60 years, prolonged nI and pI latencies and decreased nI–pI amplitude of oVEMPs were observed (Tseng et al., 2010). Thus, establishing the norm of oVEMP is essential for diagnosing VOR deficit in children aged >3 years. For children aged < 3 years, modified methodologies for testing oVEMPs are warranted, because small children typically fail to maintain their upward gaze, which may deteriorate the oVEMP waveform or even generating no response. 5. Conclusion When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters (latencies and amplitude) unaffected by structural factors. References Chang CH, Yang TL, Wang CT, Young YH. Measuring neck structures in relation to vestibular evoked myogenic potentials. Clin Neurophysiol 2007;118:1105–9.
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885 Chen CM, Chang HH, Lin KS, Lin DT, Peng SF, Young YH. Functional recovery of cranial nerves VII and VIII after hypofractionated CyberKnife radiosurgery in a neuroblastoma patient with cerebellopontine angle metastasis - Case report. Clin Neurol Neurosurg 2012;114:50–3. Chen CN, Wang SJ, Wang CT, Hsieh WS, Young YH. Vestibular evoked myogenic potentials in newborns. Audiol Neurotol 2007;12:59–63. Curthoys IS. A critical review of the neurophysiological evidence underlying clinical vestibular testing using sound, vibration and galvanic stimuli. Clin Neurophysiol 2010;121:132–44. Curthoys IS, Iwasaki S, Chihara Y, Ushio M, McGarvie LA, Burgess AM. The ocular vestibular-evoked myogenic potential to air-conducted sound; probable superior vestibular nerve origin. Clin Neurophysiol 2011;122:611–6. Fenichel GM. Migraine as a cause of benign paroxysmal vertigo of childhood. J Pediatr 1967;71:114–5. Fife TD, Tusa RJ, Furman JM, Zee DS, Frohman E, Baloh RW, et al. Assessment: vestibular testing techniques in adults and children. Neurology 2000;55:1431–41. Harth S, Obert M, Ramsthaler F, Reuss C, Traupe H, Verhoff MA. Estimating age by assessing the ossification degree of cranial sutures with the aid of Flat-Panel-CT. Legal Med 2009;11(Suppl 1):186–9. Hsu YS, Wang SJ, Young YH. Ocular vestibular-evoked myogenic potentials in children using air conducted sound stimulation. Clin Neurophysiol 2009;120:1381–5. Huang CH, Wang SJ, Young YH. Localization and prevalence of hydrops formation in Meniere’s disease using a test battery. Audiol Neurotol 2011;16:41–8. 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;119:2135–47.
1885
Jeffery N, Spoor F. Prenatal growth and development of the modern human labyrinth. J Anat 2004;204:71–92. Lewis DW. Headache in children and adolescents. Curr Probl Pediatr Health Care 2007;37:207–46. Lin KY, Hsu YS, Young YH. Brainstem lesion in benign paroxysmal vertigo children: evaluated by a combined ocular and cervical vestibular-evoked myogenic potential test. Int J Pediatr Otorhinolaryngol 2010;74:523–7. Murofushi T, Curthoys IS, Topple AN, Colebatch JG, Halmagyi GM. Responses of guinea pig primary vestibular neurons to clicks. Exp Brain Res 1995;103:174–8. Sung PH, Cheng PW, Young YH. Effect of gender on ocular vestibular-evoked myogenic potentials via various stimulation modes. Clin Neurophysiol 2011;122:183–7. Tseng CL, Chou CH, Young YH. Aging effect on the ocular vestibular-evoked myogenic potential results. Otol Neurotol 2010;31:959–63. Tseng CC, Wang SJ, Young YH. Comparison of bone-conducted vibration for eliciting ocular vestibular-evoked myogenic potentials: forehead versus mastoid tapping. Otolaryngol Head Neck Surg 2012;146:289–94. Wang SJ, Weng WJ, Jaw FS, Young YH. Ocular and cervical vestibular-evoked myogenic potentials: a study to determine whether air- or bone-conducted stimuli are optimal. Ear Hear 2010;31:283–8. Welgampola MS, Colebatch JG. Characteristics and clinical applications of vestibular-evoked myogenic potentials. Neurology 2005;64:1682–8. White DN, Curry GR, Steveson RJ. The acoustic characteristics of the skull. Ultrasound Med Biol 1978;4:225–52. Young YH. Vestibular evoked myogenic potentials: optimal stimulation and clinical application. J Biomed Sci 2006;13:745–51. Young YH, Chen CN, Hsieh WS, Wang SJ. Development of vestibular evoked myogenic potentials in early life. Eur J Pediatr Neurol 2009;13:235–9.
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Ocular vestibular-evoked myogenic potentials via bone-conducted vibration in children Chen-Han Chou, Wei-Chung Hsu, Yi-Ho Young ⇑ Department of Otolaryngology, National Taiwan University Hospital, Taipei, Taiwan
a r t i c l e
i n f o
Article history: Accepted 6 February 2012 Available online 3 March 2012 Keywords: Acceleration magnitude Bone-conducted vibration Children Ocular vestibular evoked myogenic potential
h i g h l i g h t s The characteristic parameters (latencies and amplitude) of oVEMPs did not significantly differ between children >3 years and adults. The mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, a non-significant difference. When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters unaffected by structural factors.
a b s t r a c t Objective: This study utilized bone-conducted vibration (BCV) stimuli for eliciting ocular vestibularevoked myogenic potentials (oVEMPs) to assess the vestibulo-ocular reflex (VOR) system in healthy children and adults. Methods: Fifteen healthy children aged 3–14 years and 18 healthy adults aged 24–28 years underwent oVEMP testing. Structural factors such as body weight, body height and body mass index were measured for each healthy subject. Results: All healthy children and adults presented clear oVEMPs, bilaterally. The characteristic parameters (latencies and amplitude) of oVEMPs did not significantly differ between children and adults. The mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, a non-significant difference. As stimulation intensity increased stepwise, interaural acceleration magnitude increased correspondingly, leading to early nI latency and large nI– pI amplitude of oVEMPs. However, no structural factor was statistically correlated with interaural acceleration magnitude. Conclusion: When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters (latencies and amplitude) unaffected by structural factors. Significance: Establishing the norm of oVEMP is essential for diagnosing VOR deficit in children aged >3 years. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction The vestibular system is anatomically developed and functionally responsive at birth (Jeffery and Spoor, 2004). Although a serial test of postural control is used to assess the balance function for small children, poor postural control is not specific for vestibular ⇑ Corresponding author. Address: Department of Otolaryngology, National Taiwan University Hospital, 1 Chang-te St., Taipei, Taiwan. Tel.: +886 2 23123456x65221; fax: +886 2 2394 6674. E-mail address: [email protected] (Y.-H. Young).
deficits. Alternatively, caloric and rotational chair tests have been applied to evaluate the vestibulo-ocular reflex (VOR) system in small children, which normalizes at 2 months and matures further during the first 2 years of life (Fife et al., 2000). However, the caloric test is poorly tolerated by small children, as a rotational chair is not always available in each laboratory. Additionally, both tests fail to evaluate the otolithic (utricular and saccular) function. Over the last decade, the cervical vestibular evoked myogenic potential (cVEMP) test using air-conducted sound (ACS) stimulation has been widely applied to investigate saccular disorders (Welgampola and Colebatch, 2005; Young, 2006). The cVEMP is
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2012.02.059
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
generated from the saccular macula to the sternocleidomastoid muscle via the inferior vestibular nerve. Children with benign paroxysmal vertigo, the so-called migraine equivalent (Fenichel, 1967; Lewis, 2007), may present delayed cVEMPs, indicating a retrolabyrinthine lesion along the sacculo-collic reflex pathway, which descends via the lower brainstem (Lin et al., 2010). Likewise, the recently developed ocular VEMP (oVEMP) test has been applied to assess the utricular function predominantly and the crossed VOR pathway, and to investigate vestibular disorders, i.e. vestibular neuritis, Menieres’ disease, and vestibular schwannoma (Iwasaki et al., 2008; Huang et al., 2011; Curthoys, 2010). Previously, we have reported that the characteristic parameters of oVEMP using ACS stimulation in children aged >3 years did not differ significantly from those of adults (Hsu et al., 2009). However, ACS stimuli demonstrate lower response rates than boneconducted vibration (BCV) stimuli (Curthoys et al., 2011; Wang et al., 2010). Hence, this study conducted oVEMP testing using BCV stimulation coupled with triaxial accelerometry to compare oVEMPs between healthy children and adults. 2. Subjects and methods 2.1. Subjects Fifteen healthy children (11 males and 4 females; aged 3– 14 years; mean, 9 years) and 18 healthy adults (15 males and 3 females; aged 24–28 years; mean, 26 years) were enrolled in this study. All subjects denied previous ear disorders, and were further checked with otoscope. Each subject underwent oVEMP test via BCV stimuli combined with triaxial accelerometry. Structural factors such as body weight, body height and body mass index (BMI) were also measured. This study was approved by the institutional review board of the College of Medicine, National Taiwan University Hospital, and each child’s parent and each adult signed the informed consent to participate. 2.2. oVEMP test The subject was placed in a sitting position. Surface potentials, predominantly electromyographic (EMG) activities, were recorded (Smart EP 3.90, Intelligent Hearing Systems, Miami, FL, USA). Two active electrodes were placed around 1 cm below the center of the two lower eyelids. The other two reference electrodes were positioned about 1–2 cm below the active ones, and one ground electrode was placed on the sternum. During recording, the subject was instructed to look upward at a small fixed target 2 m high from the eyes, with a vertical visual angle of approximately 30° above horizontal. The EMG signals were amplified and bandpass filtered between 1 and 1000 Hz. The stimulation rate was 5 Hz. The duration of analysis of each response was 50 ms, and 30 responses were averaged for each run. BCV stimuli were delivered using a hand-held electromechanical vibrator (V201 shaker, Ling Dynamic Systems, Royston, England) that was fitted with a short M4 bolt (2 cm in length) terminated in a bakelite cap (1.6 cm in diameter). The input signal was one half cycle of 500-Hz sine wave driven by a custom amplifier. The initial peak driving voltage was 2 dB re 10 V, approximately 128 dB force level (FL). The stimulation voltage was then decreased in steps of 5 dB until no oVEMP waveform was present. The operator held the vibrator by hand and supported most of its weight, maintaining constant coupling force during each session, such that the axis of the connected bakelite cap perpendicularly delivered a repeatable tap with little pressure on the subject’s skull at Fz (the midline forehead at the hairline). Bilateral
1881
oVEMPs were recorded simultaneously. The initial negative–positive biphasic waveform comprised peaks nI and pI. The latencies of peaks nI and pI, and the nI–pI amplitude were measured. The asymmetry ratio (%) was defined as the difference of the amplitude nI–pI on each ear divided by the sum of amplitude nI–pI of both ears, that is, (larger amplitude smaller amplitude/larger amplitude + smaller amplitude) 100. The lowest stimulus intensity that could elicit clear oVEMPs was determined (Wang et al., 2010). 2.3. Triaxial accelerometry One miniature (10 mm3, 5 gm) triaxial accelerometer (Endevco model 65–100, San Juan Capistrano, CA, USA) accompanied by a tight elastic bandage surrounding the head girth was fixed to each subject’s left mastoid area just behind the auricle (Tseng et al., 2012). Thus, the accelerometer was attached to the mastoid bone tightly. The accelerometer was used to measure the three-dimensional linear acceleration along the x-axis (anterior– posterior), y-axis (inter-aural), and z-axis (rostro-caudal) simultaneously. Forward, outward, and upward were defined as the positive x, y, and z directions, respectively. The voltage sensitivity was 100 mV/g and the frequency response was 1.5–6000 Hz (±1 dB). The triaxial accelerometer signals were digitalized at 10 kHz using a dynamic signal analyzer (NI USB-4431, National Instruments, Austin, Texas, USA), which provided signal conditioning and constant current supply to the accelerometer. The initial peak magnitudes of linear acceleration were analyzed using a customized program (LabVIEW 8.5, National Instruments, Austin, Texas, USA) during oVEMP testing. 2.4. Statistical methods Since the highest acceleration magnitude along the y-axis was observed, as compared with that along the x- and z-axis, the acceleration magnitudes along the y-axis in response to various driving voltages of the vibrator were compared by one-way repeated-measures analysis of variance (ANOVA) and Bonferroni-adjusted t-test. The difference in the acceleration magnitudes between the children and adult groups were compared by non-paired t-test. The relationships between the acceleration magnitude and oVEMP parameters, between the age and oVEMP parameters, and between the acceleration magnitude and structural factors, were examined by Pearson correlation analysis. A difference of p < 0.05 is regarded as significant. 3. Results 3.1. oVEMPs: healthy children vs. healthy adults All healthy children and adults presented clear oVEMPs bilaterally when stimulated by a vibrator with a driving voltage of 2 dB re 10 V (Fig. 1). Mean nI and pI latencies, nI–pI amplitude and asymmetry ratio in children were 8.0 ± 0.7 (mean ± SD) ms, 12.2 ± 1.5 ms, 16.1 ± 9.0 lV, and 12 ± 14%, respectively. These data did not differ significantly from those in adults (p > 0.05, non-paired t-test, Table 1). Mean interaural (y-axis) acceleration magnitudes for eliciting oVEMPs were 0.37 ± 0.12 g for children and 0.41 ± 0.20 g for adults, an insignificant difference (p > 0.05, Table 1). 3.2. Acceleration magnitude vs. oVEMP parameters Mean interaural acceleration magnitudes in both children and adults were 0.37 g for a driving voltage of 2 dB re 10 V, 0.18 g for 7 dB re 10 V, and 0.08 g for 12 dB re 10 V, showing a significant difference (p < 0.05, one way ANOVA and Bonferroni-adjusted
1882
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
Fig. 1. Clear oVEMPs are observed in an 8-year-old boy and a 28-year-old man.
t-test). Restated, as stimulation intensity decreased stepwise, interaural acceleration magnitude declined correspondingly, resulting in a serial reduction in nI–pI amplitude (Fig. 2). Additionally, the lowest acceleration magnitude for evoking oVEMPs in children was 0.14 ± 0.08 g, which did not significantly differ from 0.16 ± 0.08 g for adults (p > 0.05, non-paired t-test). Analyses of acceleration magnitude and characteristic parameters indicated that the nI latency and interaural acceleration magnitude were negatively correlated (r = 0.426, p = 0.003, Pearson correlation analysis, Fig. 3A), and the nI–pI amplitude and interaural acceleration magnitude were positively correlated (r = 0.315, p = 0.029, Fig. 3C). However, pI latency was not correlated with interaural acceleration magnitude (r = 0.184, p = 0.210, Fig. 3B). Thus, as interaural acceleration magnitude increased, early nI latency and large nI–pI amplitude were anticipated. 3.3. Acceleration magnitude vs. structural factors Structural factors such as the mean body height, body weight, and BMI were 127 ± 18 cm, 31 ± 14 kg, and 18 ± 4 kg/m2 for the child group, and 173 ± 6 cm, 65 ± 9 kg, and 22 ± 2 kg/m2 for the
adult group, respectively. Although significant differences existed in the structural factors between the children and adults, none of them was statistically correlated with interaural acceleration magnitude (p > 0.05, Pearson correlation analysis). 3.4. Effects of age and gender on oVEMPs To investigate the effect of age on oVEMPs, data of oVEMP responses from both adults and children were pooled together. Analyses of the age factor and oVEMP parameters revealed no correlations between the age and nI latency (r = 0.133, p = 0.313, Pearson correlation analysis), pI latency (r = 0.015, p = 0.909), and nI–pI amplitude (r = 0.067, p = 0.614, Fig. 4). Since only seven females were included in the study sample, another seven age-matched male subjects were also enrolled for comparison. The mean age of each group (including four children and three adults) was 15 ± 8 years. The mean BMI for the male group was 21.6 ± 4.0 kg/m2, which did not significantly differ from 20.5 ± 0.5 kg/m2 for the female group (p > 0.05, unpaired t -test). The mean nI latency, pI latency, and nI–pI amplitude in male group were 8.0 ± 0.7 ms, 12.6 ± 1.6 ms, and 18.6 ± 7.3 lV, respectively,
Table 1 Comparison of characteristic parameters of oVEMPs between children and adults.
Children Adults
N (ears)
nI latency (ms)
pI latency (ms)
nI–pI interval (ms)
nI–pI amplitude (lV)
Asymmetry ratio (%)
Acceleration magnitude (g)
30 36
8.0 ± 0.7 8.4 ± 0.5 (NS)
12.2 ± 1.5 12.7 ± 1.2 (NS)
4.3 ± 1.1 4.3 ± 1.0 (NS)
16.1 ± 9.0 15.2 ± 8.3 (NS)
12 ± 14 13 ± 14 (NS)
0.37 ± 0.12 0.41 ± 0.20 (NS)
Data are expressed as mean ± SD; NS: non-significant difference, p > 0.05, non-paired t-test.
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
1883
Fig. 2. Clear oVEMPs in a 28-year-old man in response to BCV stimuli. The nI–pI amplitude declines as the interaural acceleration magnitude reduces from 0.37 to 0.09 g.
Fig. 3. Significant correlations exist between interaural acceleration magnitude and nI latency (A), nI–pI amplitude (C), but not pI latency (B).
exhibiting non-significant difference when compared with the respective 8.4 ± 0.6 ms, 12.3 ± 1.1 ms, and 15.3 ± 7.4 lV in female group (p > 0.05, unpaired t-test). 4. Discussion We previously reported that abnormal (prolonged or absent) cVEMPs were elicited in 60% of newborns aged 2–5 days, indicating that great variation exists in the maturation of the sacculo-collic reflex in early life (Chen et al., 2007; Young et al., 2009). As a child grows, neck length and cVEMP latencies become strongly correlated, suggesting that structural factors affect the cVEMPs (Chang et al., 2007). Thus, each laboratory should establish a cVEMP norm for children of different ages. Likewise, before oVEMP test can be applied to children for investigating VOR function, one must optimize its methodology and establish the norms for children. 4.1. oVEMPs: children vs. adults The characteristic parameters (e.g. latencies and amplitude) of oVEMPs via BCV stimuli for children aged 3–14 years were not
significantly different from those of adults aged 24–28 years (Fig. 1, Table 1). Further, no correlation exists between the age factor and oVEMP parameters (Fig. 4), which is in agreement with our previous report that oVEMPs by ACS stimuli have developed at 3 years of age without statistical difference in oVEMP parameters between children and adults (Hsu et al., 2009). Opposed to our previous report that gender difference exists in oVEMP amplitude, which is attributed to variance in the muscle bulk between male and female adults (Sung et al., 2011), this study showed no gender difference in oVEMP parameters. The reason is probably because no significant difference in BMI between boys and girls, leading to no effect of gender on oVEMP parameters in children. As the ACS stimuli primarily activate the saccule (Murofushi et al., 1995), whereas both saccule and utricle are stimulated by BCV stimuli (Curthoys, 2010), indicating that BCV mode has relatively higher stimulus intensity than ACS mode in clinical testing. However, measurements of vibratory force from a vibrator cannot reflect acceleration magnitude because the same force level may generate different acceleration outputs for subjects. Hence, one must measure acceleration magnitude during oVEMP testing.
1884
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885
Fig. 4. Analyses of the age factor and oVEMP parameters reveal no correlations between the age and nI latency (A), pI latency (B), and nI–pI amplitude (C).
4.2. Acceleration magnitude vs. oVEMP parameters Acceleration magnitudes along three perpendicular axes can be measured by a triaxial accelerometer. Among three axes, interaural (y-axis) acceleration magnitude correlates strongly with nI latency and nI–pI amplitude (Fig. 3), and is therefore utilized in subsequent analysis in this study. Since oVEMP is an excitatory response, increased BCV stimulation intensity may evoke oVEMPs of early latencies and large amplitude (Fig. 2). However, under the same stimulation intensity, mean interaural acceleration magnitude between children and adults did not differ significantly (Table 1). Even the lowest interaural acceleration magnitude required to elicit oVEMPs is similar for children and adults, likely explaining why characteristic parameters of oVEMPs did not significantly differ between these two groups. However, one may question why acceleration magnitude did not differ given that the structure of children and adults is different.
4.3. Structural factor vs. acceleration magnitude Through Pearson correlation analysis, the structural factor, i.e. body height, body weight, and BMI, did not correlate with interaural acceleration magnitudes, which may explain why oVEMP parameters did not differ significantly between children (aged >3 years) and adults at the same stimulus intensity. Two hypotheses are proposed to account for this phenomenon. First, a child’s skull is complex. This complexity is magnified by numerous sutures, synchondroses, and various degrees of ossification. As age increases, there is a progressive rise in the stage of sutural closure. Most children do not develop ossification of suture before the age of 15 years, and suture closure completes at a mean age of 33 years (Harth et al., 2009). Thus, a wide cranial suture in children may hinder vibration propagation. Second, the human skull consists of a layer of cancellous bone sandwiched between two layers of compact bone. Compared with adults, the skulls of children have more cancellous bone and less compact bone. The cancellous bone markedly attenuates energy propagation via the scattering effect at the interface between its two components – bony spicules and soft tissues. Furthermore, propagation time differs when energy transits in these two components and the resulting interference may have a cancelling effect (White et al., 1978). That is, the wide cranial suture and large
amount of cancellous bone in a child’s skull may cancel increased acceleration as related to a small head size and mass. 4.4. Clinical relevance Since the inferior oblique muscle is the most superficial extraocular muscle that transverses to the electrode recording site, oVEMPs can be obtained easily from the skin surface beneath the eyes (Iwasaki et al., 2008). Thus, the amplitude of oVEMPs increases when subjects gaze upward. However, small children are usually intolerant to maintain their upward gaze. Alternatively, they usually elevate the head instead, which deteriorates the oVEMP waveform or even generates no response (Hsu et al., 2009). To ensure the stability of upward gaze, small child was instructed to look at a toy with some clicking sounds held by the parent about 1 m high from the child’s eyes, while another parent gently fixed the child’s head. When pediatric subjects underwent oVEMP testing, the results can be interpreted based on the norm of nI latency (8.0 ± 0.7 ms) for those aged >3 years. Accordingly, children with nI latency >9.4 ms are defined as having delayed oVEMPs, indicating a brainstem lesion, as evidenced by a recent report of delayed oVEMPs in a 5 year-old boy with metastatic neuroblastoma on the left cerebellopontine angle (Chen et al., 2012). However, when an adult aged >60 years, prolonged nI and pI latencies and decreased nI–pI amplitude of oVEMPs were observed (Tseng et al., 2010). Thus, establishing the norm of oVEMP is essential for diagnosing VOR deficit in children aged >3 years. For children aged < 3 years, modified methodologies for testing oVEMPs are warranted, because small children typically fail to maintain their upward gaze, which may deteriorate the oVEMP waveform or even generating no response. 5. Conclusion When children aged >3 years, the simple and quick oVEMP test via BCV stimulation can be used for investigating the integrity of the VOR system, with the characteristic parameters (latencies and amplitude) unaffected by structural factors. References Chang CH, Yang TL, Wang CT, Young YH. Measuring neck structures in relation to vestibular evoked myogenic potentials. Clin Neurophysiol 2007;118:1105–9.
C.-H. Chou et al. / Clinical Neurophysiology 123 (2012) 1880–1885 Chen CM, Chang HH, Lin KS, Lin DT, Peng SF, Young YH. Functional recovery of cranial nerves VII and VIII after hypofractionated CyberKnife radiosurgery in a neuroblastoma patient with cerebellopontine angle metastasis - Case report. Clin Neurol Neurosurg 2012;114:50–3. Chen CN, Wang SJ, Wang CT, Hsieh WS, Young YH. Vestibular evoked myogenic potentials in newborns. Audiol Neurotol 2007;12:59–63. Curthoys IS. A critical review of the neurophysiological evidence underlying clinical vestibular testing using sound, vibration and galvanic stimuli. Clin Neurophysiol 2010;121:132–44. Curthoys IS, Iwasaki S, Chihara Y, Ushio M, McGarvie LA, Burgess AM. The ocular vestibular-evoked myogenic potential to air-conducted sound; probable superior vestibular nerve origin. Clin Neurophysiol 2011;122:611–6. Fenichel GM. Migraine as a cause of benign paroxysmal vertigo of childhood. J Pediatr 1967;71:114–5. Fife TD, Tusa RJ, Furman JM, Zee DS, Frohman E, Baloh RW, et al. Assessment: vestibular testing techniques in adults and children. Neurology 2000;55:1431–41. Harth S, Obert M, Ramsthaler F, Reuss C, Traupe H, Verhoff MA. Estimating age by assessing the ossification degree of cranial sutures with the aid of Flat-Panel-CT. Legal Med 2009;11(Suppl 1):186–9. Hsu YS, Wang SJ, Young YH. Ocular vestibular-evoked myogenic potentials in children using air conducted sound stimulation. Clin Neurophysiol 2009;120:1381–5. Huang CH, Wang SJ, Young YH. Localization and prevalence of hydrops formation in Meniere’s disease using a test battery. Audiol Neurotol 2011;16:41–8. 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;119:2135–47.
1885
Jeffery N, Spoor F. Prenatal growth and development of the modern human labyrinth. J Anat 2004;204:71–92. Lewis DW. Headache in children and adolescents. Curr Probl Pediatr Health Care 2007;37:207–46. Lin KY, Hsu YS, Young YH. Brainstem lesion in benign paroxysmal vertigo children: evaluated by a combined ocular and cervical vestibular-evoked myogenic potential test. Int J Pediatr Otorhinolaryngol 2010;74:523–7. Murofushi T, Curthoys IS, Topple AN, Colebatch JG, Halmagyi GM. Responses of guinea pig primary vestibular neurons to clicks. Exp Brain Res 1995;103:174–8. Sung PH, Cheng PW, Young YH. Effect of gender on ocular vestibular-evoked myogenic potentials via various stimulation modes. Clin Neurophysiol 2011;122:183–7. Tseng CL, Chou CH, Young YH. Aging effect on the ocular vestibular-evoked myogenic potential results. Otol Neurotol 2010;31:959–63. Tseng CC, Wang SJ, Young YH. Comparison of bone-conducted vibration for eliciting ocular vestibular-evoked myogenic potentials: forehead versus mastoid tapping. Otolaryngol Head Neck Surg 2012;146:289–94. Wang SJ, Weng WJ, Jaw FS, Young YH. Ocular and cervical vestibular-evoked myogenic potentials: a study to determine whether air- or bone-conducted stimuli are optimal. Ear Hear 2010;31:283–8. Welgampola MS, Colebatch JG. Characteristics and clinical applications of vestibular-evoked myogenic potentials. Neurology 2005;64:1682–8. White DN, Curry GR, Steveson RJ. The acoustic characteristics of the skull. Ultrasound Med Biol 1978;4:225–52. Young YH. Vestibular evoked myogenic potentials: optimal stimulation and clinical application. J Biomed Sci 2006;13:745–51. Young YH, Chen CN, Hsieh WS, Wang SJ. Development of vestibular evoked myogenic potentials in early life. Eur J Pediatr Neurol 2009;13:235–9.