Full text of DOI:10.1097/00005537-200206000-00032

The Laryngoscope
Lippincott Williams & Wilkins, Inc., Philadelphia
©
2002 The American Laryngological,
Rhinological and Otological Society, Inc.
Functional Impairment of the Vestibular
End Organ Resulting From Impulse
Noise Exposure
Ronen Perez, MD; Sharon Freeman, PhD; David Cohen, MD; Haim Sohmer, PhD
Objectives/Hypothesis: To assess the effect of expo-
sure to impulse noise, known to cause damage to the
cochlea, on the vestibular part of the inner ear using
short latency vestibular evoked potentials (VsEPs),
which is a direct and objective test for evaluating the
function of the vestibular end organs. Study Design:
Prospective animal study. Methods: Sand rats (Psammo-
mys obesus) underwent baseline measurements of
VsEPs in response to linear and angular acceleration
stimuli and measurement of the auditory nerve and
brainstem evoked response (ABR). The animals were
then exposed to 10 gunshots generating impulse noise
at an intensity of approximately 160 dB sound pressure
level (SPL). Repeat measurements of the evoked poten-
tials were conducted 2 to 4 hours, 1 week, and 6 weeks
after the exposure. The amplitude and latency of the
first wave of VsEPs in response to linear and angular
acceleration stimuli, reflecting the function of the oto-
lith organs and semicircular canals respectively, were
compared between baseline and post-exposure mea-
surements, as were ABR thresholds. Results: The ampli-
tude of the first wave of the VsEPs in response to linear
acceleration was significantly (P <.001) reduced and
the latency significantly (P <.005) prolonged 2 to 4
hours after the exposure in comparison to baseline mea-
surements. The latency prolongation persisted in
follow-up measurements, whereas the amplitude
showed partial recovery. The first wave of VsEPs in
response to angular acceleration was unchanged long-
term. ABR thresholds were elevated in the long-term
by 60 dB. Conclusion: It seems that impulse noise
not only damages the cochlea, but also causes clear
functional impairment to the vestibular end or-
gans, mainly the otolith organs. Key Words: Impulse
noise, vestibular end organs, vertigo, vestibular
evoked potentials, auditory brainstem evoked
response.
Laryngoscope, 112:1110–1114, 2002
INTRODUCTION
Individuals exposed to intense impulse noise occa-
sionally experience vertigo. Over the years, a number of
studies have been directed to the possible association be-
1–7
tween exposure to noise and vestibular dysfunction.
A
small proportion of these studies focused specifically on
4–7
the effect of intense impulse noise.
reports were contradictory; while some studies concluded
The results in these
4
,5,7
that impulse noise damaged the vestibular system,
6
others concluded that it did not have any effect. The
assessment of possible functional vestibular damage in
the studies on humans was conducted using indirect clin-
ical tests such as electronystagmography and posturogra-
7
phy. In addition, an animal study described varying de-
grees of histologic damage to vestibular end organs of
guinea pigs after exposure to impulse noise. Thus, func-
tional damage to the vestibular end organ resulting from
exposure to impulse noise has not yet been confirmed,
probably because of the lack of an objective and direct test
for assessment of vestibular function.
In recent years, an objective method for directly eval-
uating vestibular function has been developed: short la-
tency vestibular evoked potentials (VsEPs). These evoked
8,9
potentials have been recorded in response to angular
10–12
and linear
acceleration impulses to the head of both
8–13 14,15
laboratory animals
and human subjects.
This
electrical activity, in the form of 5 to 6 waves, reflects the
neural activity of the vestibular pathways similar to the
1
6
auditory nerve and brainstem evoked response (ABR)
reflecting the auditory pathways. It has been shown that
the first wave of the VsEPs is the compound action poten-
tial of the vestibular nerve fibers synchronously activated
From the Department of Otolaryngology and Head and Neck Sur-
gery (R.P., D.C.), Shaare Zedek Medical Center, and the Department of
Physiology (S.F., H.S.), Hebrew University-Hadassah Medical School,
Jerusalem, Israel.
17
by the stimulus and therefore reliably reflects the func-
18
tion of the vestibular end organs. It has also been spe-
cifically shown that the first wave of VsEPs in response to
impulses of angular acceleration stimuli is initiated in the
semicircular canals and that the same wave in response to
impulses of linear acceleration stimuli is generated in the
Supported by a grant from the Israel Defense Forces Medical Corps.
Editor’s Note: This Manuscript was accepted for publication January
6, 2002.
1
Send Correspondence to Prof. Haim Sohmer, Department of Physi-
ology, POB 12272, Jerusalem 91120, Israel. E-mail: sohmer@md2.huji.ac.il
11
otolith organs.
Laryngoscope 112: June 2002
Perez et al.: Impulse Noise and Vestibular Impairment
1
110

The purpose of the present study was to assess the
effect of exposure to impulse noise, known to cause dam-
age to the cochlea, on the vestibular part of the inner ear
using vestibular evoked potentials (VsEPS), which is a
direct and objective test for evaluating the function of the
vestibular end organs.
measured with a Bruel and Kjaer 4393 accelerometer mounted on
the sliding device. The stimuli were accelerations of 3 g with a
short rise time of 1 to 1.5 ms and displacements of approximately
5
0 ␮m. This 3-g intensity is in the midregion of the VsEP input–
11
output function. After each stimulus, the head was returned to
its original position using a much lower acceleration. Acceleration
impulse stimuli were given at a rate of 2.06 per second.
Angular vestibular evoked potentials. A complete tech-
nical description of the stimulator has also been previously re-
METHODS
8,9
ported. Generally, the apparatus consisted of a drum, which
was accelerated by a stepper motor. The animal was placed in the
drum and the acceleration stimuli were transferred to the ani-
mal’s head by a head holder, which firmly gripped the upper jaw
in a plane that is optimal for maximal stimulation of the lateral
semicircular canals (the head flexed down at approximately 15°).
The resultant stimuli were clockwise and counterclockwise accel-
General Description
The study was conducted on 18 adult sand rats (Psammo-
mys obesus) with a mean weight Ϯ standard deviation (SD) of 190
Ϯ 20 g. The sand rat is a rodent species found in the deserts of the
Middle East and Northern Africa, and this laboratory has exten-
sive experience in induction and recording of short latency audi-
11,13,19
tory and vestibular evoked potentials in this animal.
For
2
eration impulses at a peak intensity of 15,000°/second with a rise
experimentation, the animals were anesthetized using an intra-
peritoneal injection of 25 mg/kg pentobarbital and additional
doses were given intraperitoneally as needed. While under anes-
thesia, rectal temperature was monitored using a thermistor
probe (Yellow Spring Instruments, Yellow Springs, OH) and
maintained at 37 Ϯ 0.5°C (using heating pads).
Initially, the animals underwent baseline measurements of
all three types of evoked potentials. All 18 animals underwent
ABR recordings, 16 of these animals underwent VsEPs in re-
sponse to linear acceleration stimuli (L-VsEPs), and 10 of them
underwent VsEPs in response to angular acceleration stimuli
time of 1–3 ms (8 stimuli in one direction and then 8 stimuli in the
other direction). Acceleration stimuli were given at a rate of 2.06
per second.
Auditory evoked potentials. The response was elicited by
alternating polarity click stimuli at a rate of 20.6 per second from
an intensity of 120 dB pe SPL down to threshold in 5-dB steps.
Threshold was defined as the lowest intensity which elicited
repeatable responses in at least three repeat measurements. If no
response could be recorded at 120 dB pe SPL, an intensity of 135
dB pe SPL was used. The earphone was placed 0.5 cm from the
left ear without deflecting the pinna, which could obstruct the
external meatus.
Recording apparatus. The electrical activity in response
to the different stimuli was recorded by needle electrodes (Grass
Instruments, Astro-Med, Inc., RI) inserted subdermally into the
vertex referred to the left pinna with the right pinna serving as
ground. The activity was band-pass filtered (300–1500 Hz), am-
plified, and averaged (n ϭ 128) by standard evoked potential
equipment (Microshev 4000, Efrath, Israel) and displayed “vertex
positive up.” Each response was obtained at least three times to
ensure reproducibility.
(
A-VsEPs). Subsequently, the awake animals in a cage 50 cm
from the rifle were exposed to 10 gunshots at an intensity of
approximately 160 dB per sound pressure level (SPL). The exper-
imenter was equipped with ear protectors during the exposure.
The intensity of the impulse noise was measured using a Bruel &
Kjaer, type 2218, precision integrating sound level meter
(
Naerum, Denmark). It was necessary to extend the range of the
sound level meter. A cover for the microphone was fashioned from
Mack’s earplugs (McKeon Products, Inc., Madison Heights, MI)
material, providing a 20-dB sound attenuation. Recordings of all
of the evoked potentials were conducted again in the laboratory 2
to 4 hours, 1 week, and 6 weeks after the exposure. At the
conclusion of the final recording session, a lethal dose of pento-
barbital was injected intraperitoneally and 5 minutes after respi-
ratory arrest; postmortem recordings of VsEPs were performed to
rule out possible electromagnetic or electromechanically induced
artifacts in the measurements.
The amplitude and latency of the first wave of VsEPs in
response to linear and angular acceleration stimuli, reflecting the
function of the otolith organs and semicircular canals, respec-
tively, were compared between the baseline and post-exposure
measurements using a non-parametric test (Wilcoxon paired
rank test). ABR thresholds were also compared between the dif-
ferent recordings. P Ͻ.05 was taken to be statistically significant.
All experimental procedures were authorized by the Hebrew
University–Hadassah Medical School Animal Care and Use
Committee.
RESULTS
Linear Vestibular Evoked Potentials
The mean (Ϯ SD) amplitude of the first wave of
L-VsEPs before exposure to impulse noise (baseline) was
1.61 Ϯ 0.48 ␮V. Two to 4 hours after exposure, it was
significantly reduced (P Ͻ.001) to 1.06 Ϯ 0.23 ␮V. The
significantly reduced amplitude persisted in the measure-
ment conducted 1 week after the exposure (P Ͻ.005) and
showed partial recovery after 6 weeks (Table I).
The mean latency (Ϯ SD) of the first wave of the
L-VsEPs before exposure was 2.03 Ϯ 0.32 ms. Two to 4
hours after the exposure, it was significantly prolonged (P
Ͻ.005) to 2.29 Ϯ 0.36 ms. The prolongation of the latency
persisted in the following measurements and was statis-
tically significant even after 1 week (P Ͻ.02) and 6 weeks
Techniques for Induction and Recording of
Evoked Potentials
(
P Ͻ.005) (Table I). Figure 1 demonstrates typical record-
Linear vestibular evoked potentials. A detailed descrip-
tion of the stimulating apparatus has been reported in previous
ings from a sand rat before, 2 hours, and 6 weeks after
exposure.
10,11
publications.
Briefly, the linear acceleration stimulator con-
sisted of a solenoid that repeatedly delivered acceleration im-
pulses to a sliding device restricted to moving the head of the
animal in one axis. The head of the animal was attached to the
moving sliding device by a head holder that firmly gripped the
upper jaw in a plane that is optimal for utricle stimulation in
Angular Vestibular Evoked Potentials
The mean amplitude of the first wave of A-VsEPs
before the exposure was 0.59 Ϯ 0.32 ␮V. It was unchanged
in the measurement 2 to 4 hours after the exposure. There
20
rodents (head forward). The magnitude of the acceleration was
Laryngoscope 112: June 2002
Perez et al.: Impulse Noise and Vestibular Impairment
1111

TABLE I.
Comparison of First Wave Peak-to-Peak Amplitudes and Peak Latencies (average Ϯ SD) of Linear and Angular VsEPs and ABR Thresholds
Between Baseline and Following Measurements.
Linear VsEPs (n ϭ 16)
Angular VsEPs (n ϭ 10)
ABR (n ϭ 18)
P1 Amplitude
␮V)
P1 Latency
(msec)
P1 Amplitude
P1 Latency
(msec)
Threshold
(dB pe SPL)
(
(␮V)
Before exposure
1.61 Ϯ 0.48
1.06 Ϯ 0.23*
1.16 Ϯ 0.43*
1.42 Ϯ 0.68
2.03 Ϯ 0.32
2.29 Ϯ 0.36*
2.23 Ϯ 0.28*
2.25 Ϯ 0.38*
0.59 Ϯ 0.32
0.67 Ϯ 0.23
0.46 Ϯ 0.26
0.86 Ϯ 0.51
3.02 Ϯ 0.36
3.30 Ϯ 0.23*
3.13 Ϯ 0.45
3.14 Ϯ 0.39
55 Ϯ 3.5
Ͼ135*
2
1
6
–4 hr after
wk after
wk after
115 Ϯ 9*
112 Ϯ 8*
*
Statistically significant (P Ͻ .05) in comparison to baseline measurements.
SD ϭ standard deviation; VsEPs ϭ vestibular evoked potentials; ABR ϭ auditory nerve and brainstem evoked response.
were also no significant changes in amplitude in the fol-
lowing measurements (Table I).
The mean latency of the first wave of A-VsEPs before
exposure was 3.02 Ϯ 0.36 ms. Two to 4 hours after expo-
sure it was significantly prolonged (P Ͻ.02) to 3.30 Ϯ 0.23
115 Ϯ 9 dB and 6 weeks after the exposure it was 112 Ϯ 8
dB (Table I). Figure 2 demonstrates ABR recordings from
a sand rat before, 2 hours, and 6 weeks after exposure
(stimulus intensity 120 dB per SPL).
␮s. This statistically significant prolongation did not per-
DISCUSSION
sist and there was recovery in the following measure-
ments (Table I).
This study has shown that exposure of sand rats to
intense impulse noise causes clear functional damage to
their vestibular end organs, mainly the otolith organs. This
is demonstrated by the significant decrease of the amplitude
and irreversible prolongation of the latency of the first wave
of L-VsEPs, which reflects the function of the otolith organs
after exposure to impulse noise. In contrast, the amplitude
and latency of first wave of A-VsEPs, which reflects the
function of the semicircular canals, was not affected in the
Auditory Brainstem Response
The ABR thresholds before exposure (baseline) were
between 50 to 60 dB pe SPL (mean, 53 Ϯ 3.5 dB). Two to
4
hours after the exposure, ABR could not be recorded
even in response to clicks at an intensity of 135 dB pe SPL.
One week after exposure, the mean ABR threshold was
Fig. 1. Recordings of linear vestibular
evoked potentials from a typical sand
rat. The upper traces were recorded be-
fore exposure (baseline), the middle
traces 2 hours after, and the bottom
traces 6 weeks after the exposure (the
decreased amplitude and prolonged la-
tency in the traces after the exposure are
evident). The stimulus trigger was at the
onset of the trace and P1 shows the first
wave of the response.
Laryngoscope 112: June 2002
Perez et al.: Impulse Noise and Vestibular Impairment
1
112

Fig. 2. Recordings of ABR in response
to 120 dB pe SPL clicks from the same
typical sand rat. The upper traces were
recorded before exposure (baseline), the
middle traces 2 hours after, and the bot-
tom traces 6 weeks after exposure.
long-term by the exposure. As expected, ABR thresholds
were significantly elevated after the noise exposure.
A number of clinical and animal studies examined
the effect of impulse noise on the vestibular end organ. As
mentioned earlier, the conclusions of these studies were
When clinically applying the results of the present
animal study to human subjects, one must be careful.
First, it seems that the sand rats’ inner ears are more
vulnerable to the impulse noise than those of humans. The
evidence for this is the dramatic effect of the impulse noise
on their cochlear response. It seems that during the first
hours after the exposure, the animals suffered total hear-
ing loss. This is usually not the case in humans. Secondly,
we would rarely expect clinical vestibular symptoms in
humans because of the existence of compensation mecha-
nisms. While this study has shown clear damage to the
vestibular end organ of the sand rat, the range of appear-
ance of vestibular symptoms in humans exposed to im-
pulse noise is probably variable and might depend on
individual vulnerability and compensation ability.
4
5
contradictory. Juntunen et al. and Ylikoski et al. re-
ported that subjects exposed to impulse noise showed sig-
nificantly more body sway than controls. They suggested
that exposure to impulse noise causes subclinical vestib-
6
ular disturbances. Pyyko et al., using posturography for
assessment of vestibular damage, reported that no signif-
icant differences in body sway were found, suggesting that
impulse noise may not cause functional changes in the
vestibular system. Of note is a recently published clinical
1
report by Golz et al. who showed that when intense noise
induced asymmetric hearing loss, there was also evidence
of vestibular dysfunction assessed by electronystagmogra7-
phy. In an animal study conducted by Ylikoski et al.,
guinea pigs were exposed to rifle shots and histologic
analysis of their inner ears was conducted. Damage was
seen in their ampullary cristae and utricular and saccular
maculae. It appeared that the damage was primarily me-
chanical. Our study, using a direct test for functional
assessment of the vestibular end organ, correlates with
most of these studies. Nevertheless, there is disagreement
with the reported histologic damage to the ampullary cris-
tae of the guinea pigs described in Ylikoski’s animal study,
because in our study there was little effect on the response
of the semicircular canals (no long-term changes in
A-VsEPs).
The vulnerability of the vestibular end organs to
noise exposure may depend on several additional factors:
3
1) the exposure regimen—in a previous study con-
ducted in this laboratory, rats were exposed to different
regimens of noise duration and intensity (between 113
and 135 dB) and for periods from a few minutes to 3
weeks. These noise exposures did not cause a significant
long-term effect on the VsEPs, in contrast to the clear
affect in the present study (about 160 dB SPL), implying
that the appearance of damage and its extent are related
to the intensity of the noise. Another possibility is that the
impulsive nature of the noise (repeat onset gunshots) to
which the animals were exposed in the present study is
21
more “destructive” than the steady state and may play a
significant role in inflicting vestibular end organ damage.
Laryngoscope 112: June 2002
Perez et al.: Impulse Noise and Vestibular Impairment
1113

2
) The state of the ear—Exposure of the normal ear to
laryngol 1988;105:558–563.
6
7
. Pyykko I, Aalto H, Ylikoski J. Does impulse noise induce
vestibular disturbances? Acta Otolaryngol Suppl 1989;468:
1
13 dB SPL broad band noise for 60 minutes did not have
an effect on the vestibular end organs, whereas after fen-
estration of the semicircular canal, the same noise expo-
sure caused significant vestibular dysfunction. This is
similar to the findings in patients with superior canal
211–216.
. Ylikoski J. Impulse noise induced damage in the vestibular
end organs of the guinea pig. A light microscopic study.
Acta Otolaryngol 1987;103:415–421.
22
23
8. Elidan J, Sohmer H, Nitzan M. Recording of short latency
vestibular evoked responses to acceleration stimuli in rats
by means of skin electrodes. Electroencephalogr Clin Neu-
rophysiol 1982;53:501–505.
9. Elidan J, Sohmer H, Lev S, Gay I. Short latency vestibular
evoked response to acceleration stimuli recorded by skin
electrodes. Ann Otol Rhinol Laryngol 1984;93:257–261.
dehiscence syndrome in whom vertigo or oscillopsia is
seen after exposure to loud noise. It has been suggested
that these findings may be the result of the round window
serving as a pressure release in the cochlear perilym-
phatic channel in the normal ear. Therefore, the sound
pressures induced in the cochlear perilymph by the stapes
footplate are preferentially transmitted to the cochlear
channels and not to vestibular channels. By inducing a
fenestration in the semicircular canal, presumably “a
round window” is created in the vestibular part of the
inner ear, allowing additional sound energy to reach the
vestibular channel.
The finding in the present study that the semicircu-
lar canals are less sensitive to impulse noise than the
otolith organs is interesting and may also have implica-
tions for the possible mechanism of the vestibular damage
by noise. This may be the result of a possible symmetric
spread of energy through the semicircular canals causing
a smaller pressure differential across the cristae, whereas
there is greater pressure differential across the maculae.
1
0. Plotnik M, Elidan J, Mager M, Sohmer H. Short latency
vestibular evoked potentials (VsEPs) to linear acceleration
impulses in rats. Electroencephalogr Clin Neurophysiol
1997;104:522–530.
1
1. Plotnik M, Sichel J-Y, Elidan J, Honrubia V, Sohmer H.
Origins of the short latency vestibular evoked potentials
(
VsEPs) to linear acceleration impulses. Am J Otol 1999;
20:238–243.
12. Jones TA, Jones SM. Short latency compound action poten-
tials from mammalian gravity receptor organs. Hear Res
1
999;136:75–85.
1
3. Perez R, Freeman S, Sohmer H, Sichel J-Y. Vestibular and
cochlear ototoxicity of topical antiseptics assessed by
evoked potentials. Laryngoscope 2000;110:1522–1527.
14. Elidan J, Leibner E, Freeman S, Sela M, Nitzan M, Sohmer
H. Short and middle latency vestibular evoked responses to
acceleration in man. Electroencephalogr Clin Neurophysiol
1991;80:140–145.
1
5. Leibner E, Elidan J, Freeman S, Sela M, Nitzan M, Sohmer
H. Vestibular evoked potentials with short and middle
latencies recorded in humans. Electroencephalogr Clin
Neurophysiol Suppl 1990;41:119–123.
CONCLUSION
It appears that impulse noise not only damages the
cochlea, but also causes clear functional impairment to the
vestibular end organs, mainly the otolith organs. When
clinically applying these conclusions, it should be taken
into consideration that this is an animal study and that
the range of vestibular symptoms in human subjects ex-
posed to impulse noise may be variable.
1
6. Sohmer H. Auditory nerve and brainstem responses (ABR):
physiological basis and clinical uses. In: Desmedt JE, ed.
Neuromonitoring in Surgery. Amsterdam: Elsevier Science
Publishers, 1989:23–47.
17. Elidan J, Langhofer L, Honrubia V. The neural generators of
the vestibular evoked response. Brain Res 1987;423:
385–390.
1
8. Freeman S, Plotnik M, Elidan J, Sohmer H. Differential
effect of the loop diuretic furosemide on short latency au-
ditory and vestibular-evoked potentials. Am J Otol 1999;
20:41–45.
BIBLIOGRAPHY
1
. Golz A, Westerman ST, Westerman LM, et al. The effects of
noise on the vestibular system. Am J Otolaryngol 2001;22:
1
90–196.
19. Perez R, Ziv E, Freeman S, Sichel J-Y, Sohmer H. Vestibular
end-organ impairment in an animal model of type 2 diabe-
tes mellitus. Laryngoscope 2001;111:110–113.
20. Curthoys IS, Oman CM. Dimensions of the horizontal semi-
circular duct, ampulla and utricle in rat and guinea pig.
Acta Otolaryngol 1986;101:1–10.
21. von Bekesy G. Experiments in Hearing. New York: McGraw-
Hill Book Co., Inc., 1960:423–424.
22. Biron A, Freeman S, Sichel J-Y, Sohmer H. Noise exposure in
the presence of canal fenestration causes a reduction in the
amplitude of short-latency vestibular evoked potentials.
Arch Otolaryngol (in press).
2
. Shupak A, Bar-El E, Podoshin I, Spitzer O, Gordon CR,
Ben-David J. Vestibular findings associated with chronic
noise induced hearing impairment. Acta Otolaryngol 1994;
1
14:579–585.
3
4
5
. Sohmer H, Elidan J, Plotnik M, et al. Effect of noise on the
vestibular system-Vestibular evoked potential studies in
rats. Noise & Health 1999;5:41–51.
. Juntunen J, Matikainen E, Ylikoski J, Ylikoski M, Ojala M,
Vaheri E. Postural body sway and exposure to high-energy
impulse noise. Lancet 1987;2:261–264.
. Ylikoski J, Juntunen J, Matikainen E, Ylikoski M, Ojala M.
Subclinical vestibular pathology in patients with noise-
induced hearing loss from intense impulse noise. Acta Oto-
23. Minor LB. Superior canal dehiscence syndrome. Am J Otol
2000;21:9–19.
Laryngoscope 112: June 2002
Perez et al.: Impulse Noise and Vestibular Impairment
1
114