Introduction
Vestibular evoked myogenic potentials are obtained by transmitting acoustic or galvanic stimuli that can stimulate the saccular or utricular macula and trigger the vestibulo-spinal response.Reference Versino, Colnaghi, Callieco, Bergamaschi, Romani and Cosi1 The cervical vestibular evoked myogenic potential response is thought to be part of the vestibulocolic reflex that starts in the saccule and travels via the inferior vestibular nerve branch to the vestibular nuclei, vestibulospinal tract and ipsilateral sternocleidomastoid muscle.Reference Sandhu, Low, Rea and Saunders2,Reference Murofushi3
Cervical vestibular evoked myogenic potential was first described by Colebatch et al. and has been used in the diagnosis of otological diseases, such as superior semicircular canal dehiscence, vestibular nerve diseases, acoustic neurinoma, benign paroxysmal positional vertigo, Ménière's disease and hearing disorders.Reference Colebatch, Halmagyi and Skuse4,Reference Kaplan, Tecellioğlu, Kamışlı, Kamışlı and Özcan5
The cervical vestibular evoked myogenic potential test can be performed with the individual sitting or lying down. The individual is asked to contract the sternocleidomastoideum muscle at a certain level. The contraction of the sternocleidomastoid muscle can be achieved by lifting the head slightly upwards in the lying position or by turning it towards the side opposite to the stimulated side in the sitting position.Reference Hızal, Erbek and Özlüoğlu6 Sustained tonic contraction of the sternocleidomastoid muscle during testing is critical for the generation of responses. Insufficient contraction of the sternocleidomastoid muscle prevents the generation of cervical vestibular evoked myogenic potential responses.Reference Isaradisaikul, Navacharoen, Hanprasertpong and Kangsanarak7,Reference Park, Han, Shim, Shin, Lim and Ahn8
In cervical vestibular evoked myogenic potential results, peaks called P1-N1 or P13-N13 are formed as a positive peak followed by a negative peak. Various methods or parameters, such as latency comparisons in the P1-N1 wave component, peak-to-peak amplitude comparisons of the P1-N1 waveform complex and inter-ear amplitude comparisons are used to determine right and left ear differences.Reference Lee, Kim, Son, Lim, Bang and Kang9
Patient age, stimulus type and stimulus intensity, filter settings, muscle tone and individual patient characteristics all influence the resulting cervical vestibular evoked myogenic potential waveform. Routine clinical evaluation procedures include ensuring good wave reproducibility and comparing the responses of both sides.Reference Hızal, Erbek and Özlüoğlu6,Reference Lee, Il Cha, Jung, Park and Yeo10
The clinical interpretation of cervical vestibular evoked myogenic potential testing focuses primarily on amplitude or threshold asymmetries between the right and left ears, and thus on the identification of the possible side of pathology. Cervical vestibular evoked myogenic potential amplitude is positively correlated with both click or tone-burst stimulus level and electromyogram level, whereas cervical vestibular evoked myogenic potential latency is independent of both factors. For these reasons, tonic electromyogram level is used as a prerequisite for the correct interpretation of interaural cervical vestibular evoked myogenic potential amplitude differences.Reference Akin, Murnane, Panus, Caruthers, Wilkinson and Profiitt11 Electromyogram scaling (amplitude normalisation) is a method used to eliminate asymmetries during cervical vestibular evoked myogenic potential testing when the sternocleidomastoid muscle fatigues, with significant decreases and fluctuations in the amount of sternocleidomastoid muscle contraction present.Reference Shahnaz and David12
It can be very difficult to obtain waves and determine the peaks of the wave (P1-N1) in cervical vestibular evoked myogenic potential, which is an electrophysiological measurement, if the clinical conditions or the patient's internal noise value is high.Reference Rosengren, Colebatch, Young, Govender and Welgampola13 In clinical practice, a 100 Hz low-pass filter is effective in neutralising the electrical artifact, and therefore it would be useful to know the availability and reliability of the filter to get cleaner and more distinct waves.
The aim of our study was to determine the precautions that can be taken to increase the reliability of the vestibular evoked myogenic potential test, which is frequently used for vestibular evaluation, without being affected by the asymmetry of the sternocleidomastoid muscle contraction, artifact-induced adverse conditions and the issues that should be considered in the interpretation of vestibular evoked myogenic potential results if these precautions are not taken.
Materials and methods
Individuals
This study included 53 individuals aged 18–65 years. With the decision of the ethics committee (numbered İ7-434-20), the data of the individuals were collected restrospectively among the individuals who applied to our clinic between 2018 and 2021. The anamnesis of the individuals was examined, and those with any middle-ear pathology, vertigo dizziness, neck problems, regular use of muscle relaxants, and muscular or neuromuscular diseases were excluded. Individuals with an asymmetry ratio of more than 0.35 and sternocleidomastoid muscle activity less than 30 μV as a result of the cervical vestibular evoked myogenic potential testing were not included in our study.
Cervical vestibular evoked myogenic potential
The Interacoustics Eclipse EP25 (Assens, Denmark) vestibular evoked myogenic potential test device was used in our study. In our clinical routine, vestibular evoked myogenic potential testing was performed by placing the reference electrode on the middle third of the sternocleidomastoid muscle, the active electrode on the upper part of the sternum where the sternum meets the sternocleidomastoid muscle and the ground electrode on the centre of the forehead. Electrode impedance values were adjusted to be less than 5 kOhm for all electrodes. During the cervical vestibular evoked myogenic potential test in all of the included individuals, the individuals were asked to contract the sternocleidomastoid muscle by turning the head in the opposite direction of the tested side in a sitting position. After each recording, the individual was allowed to rest with the head in the neutral position. In all individuals, the results of the cervical vestibular evoked myogenic potential test protocol were analysed using a 500 Hz tone-burst stimulus at a stimulus rate of 11.1 and 200 sweeps at 100 dBnHL through the airway.
In the evaluation of the individuals, P1-N1 peaks were determined within a time window of 80 ms. In order to determine whether the individuals had normal asymmetry, the waves obtained at 100 dBnHL were matched as right and left, and then only electromyogram scaling (amplitude normalisation) was applied to calculate the amplitude asymmetry ratio, and those with less than 0.35 were included in the study. Then, to examine the effect of sternocleidomastoid muscle contraction asymmetry on the amplitude asymmetry ratio, the amplitude asymmetry ratio was recalculated without electromyogram scaling and filtering. In order to examine the effect of the filter on the results, the change of the real electromyogram-scaled amplitude asymmetry ratio with the application of a 100 Hz filter was examined.
During the analysis of the individuals included in the evaluation, the sternocleidomastoid muscle contractions during the cervical vestibular evoked myogenic potential test were first determined. Muscle contraction values were analysed by looking at the difference of the microvolt values indicated as right or left RMS (root mean square) in the time-microvolts chart in the tab with electromyogram information on the screen.14
With electromyogram scaling, the asymmetry in sternocleidomastoid muscle contraction can be corrected in one of two ways. The first is self-monitoring by biofeedback; the second is electromyogram scaling. In the self-monitoring approach, the individual is provided with a feedback of the current sternocleidomastoid muscle potential.Reference Maes, Vinck, De Vel, D'haenens, Bockstael and Keppler15 However, this method is variable as the individual tires or loses concentration. Electromyogram scaling does not require the individual to maintain consistent muscle tone and instead mathematically corrects for amplitude asymmetry.Reference Noij, Herrmann, Rauch and Guinan16 Vestibular evoked myogenic potential amplitude scales proportionally to tonic electromyogram activity. Therefore, by scaling the sternocleidomastoid muscle contraction asymmetry according to the following equation, the ratio of amplitude asymmetry (amplitude asymmetry ratio) as a result of vestibular evoked myogenic potential can be accurately calculated such that the ratio of amplitude asymmetry is not affected by sternocleidomastoid muscle contraction.
Electromyogram scaling has become a widely used method for amplitude normalisation in cervical vestibular evoked myogenic potential, reducing the need to tightly control muscle tone; however, few studies and normative data using this approach are available in the literature.Reference Shahnaz and David12
In our study, individuals were divided into different groups according to the right-left sternocleıdomastoıd muscle contraction difference. Sternocleıdomastoıd muscle contraction difference of 0–10 μV and more than 10 μV was divided into group 1 and group 2, and 0–20 μV and more than 20 μV was divided into group 3 and group 4. Amplitude asymmetry ratios were compared by using different groupings of individuals with less or more asymmetry according to sternocleidomastoid muscle asymmetry to increase the sensitivity of the study and to determine the limits of the necessity of using electromyogram scaling.
Our knowledge about the positive and negative effects of the application of 100 Hz low-pass filter on the results is very limited. There is no reference on the effect of 100 Hz low-pass filter on the results and its reliability. Especially in vestibular evoked myogenic potential recordings containing artifacts, the 100 Hz filter allows us to obtain smoother traces and to identify P1-N1 peaks more easily. However, we aimed to determine the reliability of the asymmetry ratios obtained with the 100 Hz filter and to investigate the effects of the filter on the results in detail. When the effect of the 100 Hz filter was examined, individuals were divided into 2 groups as 0.12 and below in group 5, and 0.13 and above in group 6 according to the real amplitude asymmetry ratio calculated after the electromyogram scale was applied.
Results
A total of 53 individuals, 41 women and 12 men, were included in the study. The mean age of the women was 39.8 ± 11.4 years, and the mean age of the men was 38.58 ± 15.62 years (Table 1).
Table 1. Right-left sternocleidomastoid muscle contraction amounts and amplitude asymmetry ratios of the individuals included in the study were presented according to gender
Electromyogram scale analysis
Sternocleidomastoid muscle contraction difference 0–10 and more than 10 μV
Individuals were divided into 2 groups according to the right-left sternocleidomastoid muscle contraction difference as less than 10 μV in group 1 and 10 μV and above in group 2.
In Group 1, 29 subjects were included, and the mean amplitude asymmetry ratio was 0.15 ± 0.08 in the absence of electromyogram scaling and 0.14 ± 0.09 in the presence of electromyogram scaling. In group 2, 24 patients were included, and the mean amplitude asymmetry ratio was 0.19 ± 0.1 in the absence of electromyogram scaling and 0.11 ± 0.08 in the presence of electromyogram scaling.
Amplitude asymmetry ratio with electromyogram scoring was compared with amplitude asymmetry ratio without electromyogram scoring. No statistically significant difference was observed when group 1 was compared according to the change in amplitude asymmetry ratio (p > 0.05). When group 2 was compared according to the change in amplitude asymmetry ratio, a statistically significant difference was observed (p = 0.005). If the electromyogram scale was not applied, the amplitude asymmetry ratio tended to be greater than the actual value in the majority of group 2 (70.7 per cent). According to the analyses we obtained from the results, if the sternocleidomastoid muscle contraction difference is less than 10 μV, performing or not performing electromyogram scaling has no effect on the results.
Sternocleidomastoid muscle contraction difference for 0–20 μV and more than 20 μV
Individuals were divided into 2 groups according to the right-left sternocleidomastoid muscle contraction difference as below 20 μV in group 3 and 20 μV and above in group 4. Group 3 included 43 individuals, and the mean amplitude asymmetry ratio was 0.15 ± 0.08 without electromyogram scaling and 0.14 ± 0.09 with electromyogram scaling. Group 4 included 10 individuals, and the mean amplitude asymmetry ratio was 0.24 ± 0.1 without electromyogram scaling and 0.10 ± 0.05 with electromyogram scaling.
Amplitude asymmetry ratio with electromyogram scoring was compared with amplitude asymmetry ratio without scoring. No statistically significant difference was observed in group 3 when compared according to amplitude asymmetry ratio change (p > 0.05). When group 4 was compared according to the change in amplitude asymmetry ratio, a statistically significant difference was observed (p = 0.005). If the electromyogram scale was not applied, it was observed that the amplitude asymmetry ratio was greater than the actual value in all (100 per cent) of group 4.
In addition, when all the data of the sternocleidomastoid muscle contraction difference were examined, it was found that there was no need to use the scale when the sternocleidomastoid muscle contraction difference was less than 20 μV, but in the group where this contraction difference was 20 μV and above, the amplitude asymmetry ratio change was much more pronounced as the muscle contraction difference increased, indicating that the need for electromyogram scale increased in this group.
The 100 Hz filter analysis
Individuals were divided into 2 groups as 0.12 and below in group 5 and 0.13 and above in group 6, according to the real amplitude asymmetry ratio calculated after the electromyogram scale was applied.
In group 5, 29 individuals were included, and the mean amplitude asymmetry ratio was 0.06 ± 0.03 when a 100 Hz filter was not applied and 0.08 ± 0.05 when a 100 Hz filter was applied. No statistically significant difference was observed when amplitude asymmetry ratio values were compared in group 5 with and without 100 Hz filtering (p > 0.05).
In group 6, 24 individuals were included, and the mean amplitude asymmetry ratio was 0.21 ± 0.06 when a 100 Hz filter was applied and 0.23 ± 0.06 when a 100 Hz filter was applied. A statistically significant difference was observed when amplitude asymmetry ratio values were compared in group 6 with and without 100 Hz filtering (p = 0.00).
Statistical analysis
Data were analysed with SPSS® statistical analysis software. The means of the two groups were compared by t-test. Wilcoxon test was used to compare non-parametric data. Data are presented as mean and standard deviations. The effect of the presence of a 100 Hz low pass filter on the amplitude asymmetry ratio was analysed with the Wilcoxon method. Statistical significance level was accepted as p < 0.05.
Discussion
In our study, we investigated the effect of sternocleidomastoid muscle asymmetry and filter use on the amplitude asymmetry ratio obtained from vestibular evoked myogenic potential. We found that the amplitude asymmetry ratio was most affected when the sternocleidomastoid muscle asymmetry was 20 μV and above. We concluded that reliable results can be obtained without affecting the amplitude asymmetry ratio when the sternocleidomastoid muscle asymmetry is 10 μV and below, and thus there is no need to apply the electromyogram scale in this group. We found that the use of a 100 Hz filter should be considered when the amplitude asymmetry ratio is high and that the results are not affected by the use of a filter when the amplitude asymmetry ratio is below 0.12.
Reducing the variability of vestibular evoked myogenic potential waves is an important goal in the clinical application of cervical vestibular evoked myogenic potentials. We found a linear relationship between cervical vestibular evoked myogenic potential amplitude and tonic electromyogram activity.Reference Colebatch, Halmagyi and Skuse4 It has been suggested that differences in activation can be controlled and thus variability reduced, and normalisation has been used in vestibular evoked myogenic potential results, but limited work has been done to assess its usefulness.Reference van Tilburg, Herrmann, Guinan and Rauch17 In contrast to most studies using biofeedback to normalise muscle tone, recent studies have used electromyogram scaling for amplitude normalisation in cervical vestibular evoked myogenic potential and ocular vestibular evoked myogenic potential.Reference Shahnaz and David12,Reference McCaslin, Fowler and Jacobson18 In our study, we aimed to determine the muscle contraction difference values that are necessary and not necessary using the electromyogram scale.
In our study, individuals were divided into different groups twice according to the right and left sternocleidomastoid muscle contraction difference. The amplitude asymmetry ratio obtained as a result of cervical vestibular evoked myogenic potential was compared, and the aim was to increase the specificity of the study by reducing the sternocleidomastoid muscle contraction difference to smaller ranges and to keep the amplitude asymmetry ratio results within a reliable range so that normalisation and/or electromyogram scaling was not required.
The mean age of the individuals included in the study was 39.8 ± 11.4 years for women and 38.58 ± 15.62 years for men. Although studies in the literature reported that the amplitude of the tonic muscle activity of the sternocleidomastoid muscle decreased with increasing age, no statistically significant difference was observed between the age groups 20–40, 41–60 and older than 60 years.Reference Basta, Todt and Ernst19 In another study, it was reported that cervical vestibular evoked myogenic potential responses remained within the normative value range for individuals up to 60 years of age; after 60 years of age, test results were affected.Reference Welgampola and Colebatch20 Since the mean ages of the individuals included in our study were close to each other, and the oldest individual was 54 years old, no statistically significant difference was observed between age and vestibular evoked myogenic potential results. In order to minimise the effect of age on the tonic muscle activity of the sternocleidomastoid muscle, care was taken to ensure that the age ranges of the individuals included in the study were close.
One study reported that electromyogram levels can be recorded in the range of 30 to 50 μV.Reference Akin, Murnane, Panus, Caruthers, Wilkinson and Profiitt11 Young et al. reported that the contraction levels of the active sternocleidomastoid muscle were between 50 and 200 μV.Reference Young21 The contraction levels of the sternocleidomastoid muscle influence the test results. Cervical vestibular evoked myogenic potential wave morphology should be reproducible under controlled conditions and stimulus parameters within the same patient.Reference Isaradisaikul, Strong, Moushey, Gabbard, Ackley and Jenkins22 Therefore, in our study, we included individuals with a minimum sternocleidomastoid muscle contraction range of 30.5 to a maximum of 149.7.
In order to detect a cervical vestibular evoked myogenic potential response, adequate muscle tone of the sternocleidomastoid muscle is required, and many normative datasets in the literature have used biofeedback approaches to achieve consistency of sternocleidomastoid muscle tone in individuals, as individuals do not actively maintain consistency of sternocleidomastoid muscle tone throughout the test.Reference Isaradisaikul, Navacharoen, Hanprasertpong and Kangsanarak7,Reference Janky and Shepard23 In the literature, it has been observed that the sternocleidomastoid muscle tone required for the vestibular evoked myogenic potential test has different values. Studies have reported that a range of 30–75 μV is sufficient, and the minimum value should be taken as 50 μV.Reference Isaradisaikul, Navacharoen, Hanprasertpong and Kangsanarak7 When we look at the mean sternocleidomastoid muscle contraction in our study, we can say that a population compatible with the literature was included in the study with 73.63 μV in females and 88.24 μV in males.
In the vestibular evoked myogenic potential test, which causes the third window effect and is especially critical in the differential diagnosis of superior semicircular canal dehiscence, the diagnosis is made by using the threshold and amplitude data of P1-N1 waves in vestibular evoked myogenic potential.Reference Mau, Kamal, Badeti, Reddy, Ying and Jyung24 Threshold values are lower than normal and amplitude values are too high.Reference Mikulec, McKenna, Ramsey, Rosowski, Herrmann and Rauch25 In particular, numerical observation and recording of amplitude data is only possible when electromyogram scales are not used. If the difference in sternocleidomastoid muscle contraction is less than 10 μV, electromyogram scaling or not has no effect on the results.
In both groups, amplitude asymmetry ratios with electromyogram scaling were lower than those without electromyogram scaling. But this difference is much more pronounced in group 2. In other words, if the difference in sternocleidomastoid muscle contraction is 10 μV and above, if electromyogram scaling is not applied, the amplitude asymmetry ratio will be higher than it should be in most of the group (70.7 per cent), leading to false positive results (i.e. a pathological diagnosis that is not actually present).
Although there is no harm in not performing electromyogram scaling if the sternocleidomastoid muscle contraction difference is less than 20 μV, it was observed that amplitude asymmetry ratio results obtained without electromyogram scaling were not reliable in group 4. We can prevent misdiagnosis (false positive) by applying electromyogram scaling in group 4.
In both groups, amplitude asymmetry ratios with electromyogram scales were lower than those without electromyogram scales. If the sternocleidomastoid muscle contraction difference is 20 μV and above, if the electromyogram scale is not applied, the amplitude asymmetry ratio will be higher than it should be in all (100 per cent) of the group, leading to false positive results, that is, a pathological diagnosis that is not actually present. The role of the electromyogram scale is clearly seen in group 4.
Especially in individuals with borderline amplitude asymmetry ratio (close to 0.35), it has been shown that electromyogram scaling should definitely be applied when the sternocleidomastoid muscle contraction difference is 20 μV and above and that amplitude asymmetry ratio results without scaling would be misleading.
• In the cervical vestibular evoked myogenic potential test, sternocleidomastoid muscle asymmetry of more than 20 mV impairs test reliability, causing the amplitude asymmetry ratio to be higher than it should be
• To obtain reliable results in cervical vestibular evoked myogenic potential testing, sternocleidomastoid muscle asymmetry should not exceed 10 mV
• Caution should be exercised when the sternocleidomastoid muscle asymmetry is in the range of 10–20 mV in cervical vestibular evoked myogenic potential testing
• If the amplitude asymmetry ratio is within the normal range and far from the pathological value, it can be said that the test results are reliable in sternocleidomastoid muscle asymmetry up to 20 mV
• While the use of filters facilitates the interpretation of cervical vestibular evoked myogenic potential waves and removes artifact, caution should be exercised when the amplitude asymmetry ratio is greater than 0.13
• In general, it should be kept in mind that when the sternocleidomastoid muscle asymmetry exceeds 20 mV and the amplitude asymmetry ratio is greater than 0.13, the use of filters may negatively affect the results
According to the results obtained from our study, the presence of sternocleidomastoid muscle asymmetry in the amplitude asymmetry ratio obtained as a result of vestibular evoked myogenic potential testing and whether a method such as electromyogram scaling is used in the presence of asymmetry should be reported for more detailed research in studies. Electromyogram monitoring should be performed in the background of the vestibular evoked myogenic potential testing because in cases where the difference in sternocleidomastoid muscle contraction increases, the amplitude and latency of cervical vestibular evoked myogenic potential waves are likely to be larger than the normal value range. This will cause the right-left amplitude asymmetry ratio to be larger than it should be and cause false positive results.Reference Rosengren, Welgampola and Colebatch26 However, it has also been reported in the literature that if appropriate normative data and minimum sternocleidomastoid muscle contraction difference are obtained, the results (amplitude asymmetry ratio) can be reliably determined in vestibular evoked myogenic potential testing.
In group 5 and group 6, 100 Hz filter application tended to increase the amplitude asymmetry ratio. The increase in amplitude asymmetry ratio with a 100 Hz filter was observed in 48.16 per cent of group 1 and 87.36 per cent of group 6. Especially in group 6, if the amplitude asymmetry ratio is close to the pathological value (0.35), the use of the filter will increase the amplitude asymmetry ratio and cause false positive results, misdiagnosis and treatment. Based on these results, in group 6, that is, in cases where the amplitude asymmetry ratio is 0.13 and above, filter application may change the results negatively and should be used with caution.
Conclusion
Considering all the data, in our study, when performing electromyogram monitoring and/or determining the right-left sternocleidomastoid muscle contraction interval at the beginning of the test, keeping the contraction interval not exceeding 10 μV will maximise the reliability of cervical vestibular evoked myogenic potential waves and amplitude asymmetry ratio. If clinically feasible, it can be concluded from our study that the sternocleidomastoid muscle contraction difference should ideally not exceed 10 μV and should not exceed 20 μV. It is thought that extending the study more comprehensively, especially in different age groups, gender and sternocleidomastoid muscle contraction amounts, and supporting studies that increase the use and reliability of the cervical vestibular evoked myogenic potential test, will benefit clinical evaluations. There is no reference on the effect of filters used because of artifacts in vestibular evoked myogenic potential testing on the reliability of the results, and this study adds a new perspective to the literature in terms of muscle asymmetry and filter reliability.
Competing interests
None declared
