Amplitude normalization reduces cervical vestibular evoked myogenic potential (cVEMP) amplitude asymmetries in normal subjects: proof of concept.



The cervical vestibular evoked myogenic potential (cVEMP) is an acoustically synchronized, signal averaged, brief inhibitory response of a contracted muscle usually resulting from an acoustic stimulus. The cVEMP is recorded from the tonically contracted sternocleidomastoid muscle (SCM). The presence and amplitude of the cVEMP is related to both the integrity of the sacculo-collic pathway and magnitude of electromyographic (EMG) activity at the time of recording. Measurement variables include the absolute latency of the primary positive going component (referred to as P13) and interaural (i.e., left versus right) latency differences. Also measured is the peak-to-peak interaural amplitude asymmetry (IAA; percent difference in amplitude, left versus right). It is known that the amplitude of the cVEMP is positively correlated with the magnitude of tonic EMG from which the evoked potential is extracted. Thus, if EMG amplitude is uncontrolled, one cannot determine whether cVEMP asymmetries are occurring due to unilateral end organ disease or asymmetric tonic EMG activity. Two methods have been suggested to control for tonic EMG activity. These include (1) patient self-monitoring of EMG activity with biofeedback and (2) mathematical correction (i.e., amplitude normalization) of the left and right cVEMP waveforms. Currently, it is unknown how effective amplitude normalization techniques are at reducing cVEMP amplitude asymmetry in the presence of varying levels of EMG.


The purpose of this investigation was to determine whether the use of amplitude correction techniques would reduce significantly the P13-N23 IAA data in otologically and neurologically intact adults when the level of EMG was varied between right and left sides.

Research Design

A prospective, repeated measures design was used for three different investigations in which cVEMPs were recorded and then processed using amplitude correction.

Study Sample

Subjects were 20 otologically and neurologically health young adults between 21 and 29 yr of age.


cVEMPs were recorded at four different EMG target levels ranging from 100 to 400 μV. The absolute peak-to-peak amplitude of P13-N23, absolute latency of P13, and the left/right amplitude asymmetry of P13-N23 were measured both with and without the use of EMG amplitude correction techniques. IAAs were calculated using 10 different conditions of varying EMG asymmetry with and without amplitude correction.

Data Collection And Analysis

Data were analyzed using repeated measures analysis of variance (ANOVA) to detect tonic EMG level-dependent differences separately for P13 latency, P13-N23 peak-to-peak amplitude, and mean root mean square (RMS) amplitude cVEMP responses. The amplitude of cVEMP responses from the left and right side were used to calculate IAA for subsequent analyses. Linear regression analyses compared level of tonic EMG with cVEMP amplitude. A one-way multivariate analysis of variance (MANOVA) was used to determine if IAAs were significantly reduced following amplitude correction. Any differences found were investigated using unplanned linear contrasts.


The uncorrected cVEMP amplitude and RMS EMG all increased significantly with increases in the EMG target levels. With amplitude correction, cVEMP amplitude did not change significantly with changes in RMS EMG or EMG target levels.


These findings suggest that the use of amplitude correction techniques represent an effective method of neutralizing the factor of variability in tonic EMG level on the cVEMP that would be otherwise uncontrolled. Indeed when correction is employed in cases of extreme tonic EMG asymmetry, the upper limit of percent IAA is roughly half of that when EMG correction techniques are not used. Our findings are also in agreement with those of Bogle et al (2013) showing that the input/output growth function for P13/N23 amplitude is not linear but, in fact, saturates at supra-maximal stimulation levels. Accordingly, and contrary to what has been published previously, achieving maximum muscle activation may produce a paradoxically inferior signal-to-noise ratio and in some cases result in an artificially small (or undetectable) corrected cVEMP amplitude. cVEMP amplitude either asymptotes (if maximum EMG amplitude saturation occurs at the same stimulus intensity as yields the maximum cVEMP amplitude), or the cVEMP can become smaller if EMG amplitude can increase further beyond the stimulus intensity that yields that largest P1-N1 amplitude. In the latter case the noise increases further to reach maximum and creates a disadvantageous signal (cVEMP) to noise (tonic EMG) ratio.

American Academy of Audiology.