ffeunng Research, 48 (1990) 209-220

Elsevier

209

HEARES 01448

Electrophysiolo~cal aspects of the middle ear muscle reflex in Latency, rise time and effect on sound transmission

H. van den Berge, H. Kingma, C. Kluge and E.H.M.A. Marres Department oj Otorhinolmyngology. Uniwrsity Hospital, Maastricht, The Netherlands

(Received 15 June 1989; accepted 25 April 1990)

the rat:

The latency, the rise time and the influence of the acoustic reflex on sound transmission were investigated in the adult rat during

ketamin anaesthesia. This was done by recordings of the cochlear microphonics (CM) and electromyographic (EMG) recordings of

the reflex responses of the tensor tympani muscle. The acoustic reflex was elicited by contralateral acoustic stimuli of which the

intensity and frequency was varied. Ipsiiaterally, the effect on sound transmission was determined by estimating the change in

amplitude of the CM’s of ipsilateral administered subliminal stimuli. It was shown that both the tensor tympani muscle and the

stapedius muscle contribute in the refIex. The latency as well as the rise time of the reflex determined by CM recordings showed to be

short (minimal values: 12 and 7 ms respectively). The mean latency of the tensor tympani muscle reflex. measured by EMG, was

about 7 ms. The attenuation of 0.25-8 kHz tone bursts upto 115 dB SPL is limited to a mean maximum of 15 dB SPL. The maximal

attenuation was shown to occur at 1 kHz. Frequencies above 2 kHz appeared to be the best elicitor of the middle ear muscle reflex.

Middle ear muscle reflex; Latency: Contraction time; Attenuation; Electromyography

Introduction

The middle ear muscle reflex has been the subject of research for many years. This has led to

various speculations on the function of the reflex. One of the most attractive theories is that the

middle ear muscles protect the inner ear from noise damage. However, several objections to this

theory have been made (for references see Bosatra et al., 1984). Firstly, the latency of the reflex is

thought to be too long to protect effectively against impact noise. Secondly, the attenuation of sound transmission as a result of contraction of the muscles would be limited. Finally, a process of

fatigue, which occurs during exposure to noise over a longer period of time, would restrict the

protective value of the reflex. Experiments on these subjects have been intensive, but often frag- mentary. and they were performed using different

Correspondence too: H. van den Berge, University Hospital

Maastricht, Department of Otorhinolaryngology, Postbox 1918.

6201 BX Maastricht, The Netherlands,

animal species. Since major interspecies dif- ferences exist in both structure and function of the

middIe ear muscles, an integrated mo~holo~cal and electrophysiological study in one animaf

species would be required. Therefore, we initiated such an integrated study in the rat. The morpho-

logical characteristics of both the tensor tympani and stapedius muscles have already been pre-

sented (Berge and Wirtz, 1989a; Berge and Wirtz, 1989b; Berge and Wal, 1990). These studies

showed that both muscles are able to contract fast, that they have a high fatigue resistance and a well developed motor innervation pattern together with a probably absent sensory innervation.

Concerning the noise protection theory im- portant characteristics to be investigated would be the latency, the rise time, the quantitative effect on sound transmission and the fatiguability of the reflex. In the past, the effect on sound transmis- sion was usually estimated in an artificial situation for example during and after recovery of stapedius muscle paralysis (Borg, 1968; Borg and Zakrisson, 1974; Zakrisson, 1979) and by electrical stimula-

0378-5955/90/$03.50 IQ 1990 Elsevier Science Publishers B.V. (Biomedical Division)

210

tion of the muscles or brainstem nuclei (Moller, 1965; Teig, 1973). Furthermore, the quantitative effect on sound transmission has often been de-

rived indirectly from contraction force studies

(Wever and Bray, 1937, 1942; Teig, 1972a) and

masking studies (Borg and Zakrisson, 1974). The latency of the acoustic reflex has mostly been

determined by the relatively indirect method of

acoustic impedance measurements (see for refer- ences Bosatra et al., 1984). Finally not much is known about the rise time of the reflex, which is here defined as the time lapse between the begin- ning and the moment of maximal effect on sound transmission, measured at the level of the cochlea. This characteristic, which is a measure for the

contraction time of the muscles, determines, to- gether with the latency how fast sound transmis-

sion can be influenced. We therefore set out to investigate the latency,

rise time and quantitative effect on sound trans- mission of the acoustic reflex in the rat by direct

methods, leaving the middle ear muscles and their

inne~ation intact.

Materials and Methods

Surgical procedures

Ten twelve week old male Lewis rats were anesthesized with ketamin (Vetalar”), as this drug is assumed not to influence the middle ear muscle reflex (Bosatra et al., 1984). Following administra- tion of 1.3 ml/kg bodyweight of Vetalar@ i.p., the

anaesthesia was maintained by 0.1 ml Vetalar@ i.m. every 30 min. The tympanic bulla was ex-

posed by a retroaural approach, leaving the outer ear canal intact. The rats were placed in a fixator

especially designed for this purpose. Body temper- ature was controlled at 38*C via a feed back

heating system attached to the fixator. For acous- tic stimulation, Nicolet Model TIP-300 trans- ducers with a tube length of 25 cm were used on both sides, the ends of which were inserted till close to the tympanic membrane. During stimula- tion the outer ear canal was completely sealed by the tube. For measuring cochlear microphonics (CM), a silver ball electrode (0 0.8 mm) was inserted through a small hole made in the bulla and placed on the promontory close to the round window niche. This was done using a micromani-

pulator and operation stereo microscope. For se- cording electromyograms (EMG) a silver wire

electrode was placed on the belly of the tensor

tympani muscle. In both cases a reference elec-

trode and a ground electrode were placed in the

muscles near the bulla and in a hind limb, respec- tively.

Experiments were performed in a total number

of ten rats. Control measurements during total muscular relaxation by Pavulon9Vetalar” intuba- tion anesthesia and by tenotomy of the tensor tympani muscle were performed in two rats. Con- traction of the tensor tympani muscle was moni- tored during all experiments by direct inspection using the operation microscope.

Acou.~tic stimulation

Acoustic stimuli were administered contralater- ally (i.e. the nonoperated ear) and ipsilaterally (i.e.

the operated ear). In general, both ears were stimulated simultaneously, the contralateral stimu-

lus to induce a (bilateral~ middle ear muscle con-

traction and the ipsilateral to determine effects of the contraction on sound transmission, the latency

and the rise time of the reflex (see below). The ipsilateral stimuli were generated by a Nicolet CA-2000 and the contralateral stimuli by an Acoustics AC3 audiometer or the Nicolet CA- 2000. The stimulus parameters indicated in this study (intensity, duration, frequency, envelope, repetition rate) refer to the output characteristics

of the transducer tubes (Nicolet Model TIP-300). Acoustic stimuli were calibrated for each trans-

ducer tube by using a 2 cc Bruel and Kjaer DB- Ol38 coupler and expressed in terms of dB re 20

Pa (SPL). At 500 Hz, 118.5 dBSPL, 1.4% distor- tion was measured. In all experiments the Nicolet

CA 2OOO was used for data-acquisition and signal processing. The bandwidth during registration of the CM was 150 Hz-10 kHz (low and high pass filters 12 dB/octave).

Several experiments were performed to de- termine (a) the relation between CM amplitude and stimulus intensity at various frequencies, (b) the latency of the reflex, fc) the rise time of the reflex, (d) the effect of the reflex on sound trans- mission and (e) the most effective acoustic stimu- lus that elicits the middle ear muscle reflex.

211

In order to establish the relation between CM amplitude and stimulus intensity, 50 tone bursts with a plateau of 15 ms and a rise and fall time of

I ms. were given at a repetition rate of 0.3 Hz or 3.1 Hz. The responses were averaged. Measure-

ments were done at various frequencies (0.25. 0.5. 1.0, 1.5, 2.0, 4.0, 8.0 kHz) and with increasing

stimulus intensity (37-112 dB SPL). The amph- tudes of the CM were determined at 4 ms after the

onset of the stimulus. All CM amplitudes in this

study were measured peak to peak.

The latency and rise time of the reflex were determined from the change in the averaged CM

amplitudes following ipsilateral acoustic stimula-

tion above the reflex threshold (i.e. 50 tone bursts of 50 ms. 87-112 dB SPL. 0.25-8 kHz, repetition rate 3.1 or 0.3, rise- and fall time 1 ms). The

latency was estimated from the time between the beginning of the stimulus and beginning of the change in CM amplitude. A change in CM ampli-

tude was considered to be significant when it was larger than 10%. The rise time was estimated from the time between the beginning and the end of the

change in CM amplitude. To determine the effect of the middle ear muscle

reflex on sound transmission, contralateral acous- tic stimuli were given to elicite a (bilateral) reflex.

Ipsilaterally. the effect of the contraction on sound transmission was determined by measuring the

difference of the averaged CM amplitude of a

sub-threshold ipsilateral stimulus (see below) with and without contralateral stimulation. The con- tralateral stimuli were administered manually from

0.5- 1.0 s prior to. till 0.5-1.0 s after the ipsilater- ally administered tone bursts. The ipsilateral

stimulus consisted of 50 tone bursts of 67 dB SPL (i.e. below the reflex threshold) administered with a repetition rate of 3.1 Hz at various frequencies (0.25, 0.5. 1, 1.5. 2, 4 and 8 kHz). An intensity of 67 dB SPL was chosen since this induced a CM with a good signal to noise ratio, the stimulus did not elicit an middle ear muscle contraction, and the stimulus CM amplitude response curve showed to be linear around this intensity (see results). To quantify a possible subthreshold facilitation (Sim-

mons, 1965; Blood and Greenburg, 1981) of the 67 dBSPL stimulus, measurements with a stimuli of 57 dB at 1, 1.5, 2 and 4 kHz were done in two animals (see results). The CM amplitude without

contralateral stimulation was measured before as well as following contralateral stimulation to check for reproducibility and cochlear damage. Only those data sets were used for analysis of which

these two CM amplitudes without contralateral stimulation showed a test-retest variation of less

than 5%. The effectiveness of contralateral stimuli in

eliciting the middle ear muscle reflex was de-

termined from the difference in peak to peak CM amplitude of an ipsilateral administered stimulus

of 67 dB SPL and 1.5 kHz (50 tone bursts of 15

ms, repetition rate 3.1 Hz) with and without the

contralateral eliciting stimulus. Ipsilaterally. 1.5 kHz was chosen as this frequency appeared to be very sensitive (see results). The contralateral

stimulus was administered manually from 0.551.0 s prior to, till 0.5- 1 .O s after the ipsilaterally administered 50 tone bursts (frequency: 0.25. 0.5.

1. 1.5. 2. 4 and 8 kHz and intensity 65. 115 dB SPL). Again, reproducibility was taken within 5%‘.

Next to the estimation of the latency by CM recordings the latency of tensor tympani muscle reflex was investigated by EMG recordings. using

a silver wire surface electrode on the belly of the muscle. For data-acquisition and signal processing

the Nicolet CA 2000 was used. The bandwidth

during registration of the EMG was 300 Hz-10 kHz (low and high pass filters 12 dB/octave).

Contraction was induced by single ipsilateral stimuli with a duration of 15.-200 ms, frequency

of 0.5. 1.0, 1.5. 2 and 4 kHz and intensity varying from 27 to 107 dB SPL. The latency was de- termined from the time between beginning of the stimulus and the beginning of the muscular action potentials.

Results

Fig. 1 (trace a) shows a typical electrocochleo- gram induced by ipsilateral 1.5 kHz sinusoidal stimulation at 67 dB SPL. In the first two millisec-

onds the response shows a fast increase in ampli- tude, which was also seen at higher stimulus inten- sities. In this time lapse the response predomi- nantly reflects contributions of the nerve action potential and the CM. After a few milliseconds

212

I

0 7.5 1s m

Fig. 1. Examples of cochlear microphonics in the rat. (a) Averaged CM recording of a stimulus of 67 dB SPL 1.5 kHz; stimulus onset at t = 0 (stimulus duration: 50 ms; repetition frequency 3.1 Hz; number of stimuli: 50). (b) The same record- ing as in (a), during total muscular relaxation by Pavulot?-

Vetalar” anesthesia.

1OC

1C

2

the response amplitude decreases slightly and stabilizes. Then the sinusoidal changes of the

potentials most likely reflects the CM. It was not to be expected that the decrease in CM amplitude

at t = 2-3 ms was induced by the middle ear

muscles. Nevertheless we checked this by measur- ing the ECoch during total muscle relaxation by

anesthesia with Pavulon8. Regardless the stimulus intensity, the initial response pattern is the same

with and without total muscle relaxation (see trace b in Fig. 1).

In Fig. 2a the CM amplitude and the standard deviation (SD) (IV = 4) of a 1.5 kHz stimulus is plotted as a function of the stimulus intensity.

There is a linear relations~p upto 74 dB SPL, whereas above 74 dB SPL a saturation, and above

405080708090 loo 110 1M 30 40 50 60 70 80 90 loo 110 120

Sttmulus Intensity WEPL.) B

Stkn&us Freciuemcy Wizf

Stimulus lntensty UEXSFU

Fig. 2. Relation between stimulus intensity and CM amplitude. (a) Mean CM amplitude of a stimulus of 1.5 kHz as a function of the stimulus intensity. Vertical bars: +/- 1 SD (b) The same form as the previous for different frequencies (D----- 0: 0.5 kHz: A-A: 1 kHz: O- o: 1.5 kHz; 0 -0: 2 kHz; I- l : 4 kHz). (c) Sensitivity curve of the CM amplitude. The

curve represents the mean intensity required to induce a 15 nV CM amplitude as a function of stimutus frequency (a-c: N = 4)

87 dB SPL a diminuation occurs. There are only slight inter-individual differences in amplitudes

between the four rats. The relationship between the CM response amplitude and stimulus intensity depends on frequency as is shown in Fig. 2b. Fig.

2c shows the sensitivity curve of the CM ampli-

tude. The intensity required to induce an CM amplitude of 15 PV is plotted as a function of the

stimulus frequency. It can be seen that the maxi-

mal sensitivity is found at 1.5 kHz. At stimulus intensities above 77 dB SPL the response pattern appears to be less sinusoidal for most frequencies.

Furthermore, at these higher stimulus intensities (Fig. 3, trace a) the amplitude of the response

appears to decrease after about 15-30 ms depend-

ing on the stimulus intensity. The decrease itself takes about 5-15 ms. This response pattern is the

same at low (0.3 Hz) and at higher (3.1 Hz) stimulus repetition rates. This suggests that at

these repetition frequency no adaptation occurs. The time lapse between the beginning of the stimulus and the beginning of the decrease of the CM amplitude represents the latency of the mid- dle ear muscle reflex. Fig. 3 trace a, depicts the CM with both muscles intact. Trace b shows the CM recorded following tenotomy of the tensor tympani muscle. In both situations the amplitude

of the CM starts to decrease at about t = 15 ms

(onset stimulus at t = 0). However, the amplitude of the CM decreases more with both muscles

intact. Furthermore with the tensor tympani muscle tenotomized the decrease is followed by a

I 0 20 40

mm

Fig. 3. (a) Averaged CM recording of a 107 dE SPL. 1 kHz

stimulus (stimulus duration: 50 ms; r-petition frequency 3.1

Hz; number of stimuli: 50). (b) Tine same form as the previous.

however. tensor tympa?si muscle tenotomized (see text).

----__-.--__-._I’“-.~ , _?I-‘\*,., <“-^u--.H-

\/

,\ d

._-.__~l__-___.___Xr.- _A,_\. _n/--_-,-- __._.__-- e

0 I A’ 15 30 ms

Fig. 4. EMG recordings of the reflex response of the tensor

tympani muscle. (a) Example of a EMG response following a

ipsilateral stimulus of 107 dB SPL, 2 kHz (stimulus duration:

200 ms). (h-e) EMG responses at a shorter time scale at

decreasing stimulus intensity following a tone burst of 2 kHz.

(b) 107 dB SPL. (c) 97 dB 5PL. (d) 87 dB SPL. (e) 77 dB SPL.

Arrow: beginning of the stimulus (see text).

slight increase of the CM amplitude at about t = 25 ms.

In order to estimate the beginning of the musc- ular potentials of the tensor tympani muscle EMG

recordings were made. Fig. 4 shows the recordings of the EMG of the tensor tympani muscle at two

different time scales (traces a and b-d, respec- tively) following ipsilateral stimulation. The muscle

potentials could be induced by both ipsi- and

contralateral auditory stimulation. We recorded these potentials while the measuring electrode was placed on the belly of the tensor tympani. No such potentials could be detected when this electrode

was placed on the promontory. This confirms that these potentials are generated in the tensor tympani muscle and not in the muscles surround- ing the bulla. The rms value of the EMG activity increases with increasing stimulus intensity. Fig.

4b-d shows that the latency of the muscle re- sponses increases only slightly with decreasing stimulus intensity, in contrast to the stronger de-

crease of the latency when estimated from the CM recordings. This is depicted in Fig. 5. which shows the latency of the reflex as a function of stimulus intensity determined by CM recordings (curve a) and the latency of the tensor tympani muscle determined by EMG recordings (curve b). The curves show that the change in CM amplitude-

214

i 1:: *\ g lo- 2

8 *-----------------*_________________ * _J 5-

0’ 80 85 90 95 100 105 110

Stimulus ~Intenslty MBWL)

Fig. 5. Latency of the midole ear muscle reflex. Curve a: mean

latency determined from CM recordings as a function of

stimulus intensity (N = 5). Curve b: mean latency of the EMG

‘response of the tensor tympam muscle (N = 2) (see text).

latency is larger than the EMG-latency at every stimulus intensity. The latency decreases with in- creasing stimulus intensity.

Fig. 6 shows the mean time lapse between the beginning and the end of the decrease of the CM

amplitude (rise time of the middle ear muscle reflex) as a function of stimulus intensity with

both middle ear muscles intact. The rise time varies from 7 to 12 ms and decreases at higher

stimulus intensities. In Figs. 7 and 8 the effect of the contralaterally

induced middle ear muscle reflex on the ipsilateral CM amplitude of tones of various frequencies is shown. Examples of CM recordings with and without contralateral stimulation are shown in

80 85 90 95 100 105 110

Stmhs lntenslty (dBsPL)

Fig. 6. Mean rise time of the middle ear muscle reflex as a

function of stimulus intensity (N = 5) (see text).

Fig. 7. The effect of a contralaterally induced middle ear

muscle reflex on the CM amplitude of an ipsilaterally adminis-

tered stimulus, without and with total muscular relaxation. (a)

CM response of a 67 dB SPL 1.5 kHz tone (averaged response

of 50 tone bursts administered at a repetition rate of 3.1 Hz,

stimulus rise and fall time: 1 ms; single stimulus duration: 15

ms), (b) CM response of the same stimulus like in (a), during

contralateral stimulation with a 97 dB SPL and 1.5 kHz tone.

(c-d) The same stimulus situation as in a-b respectively,

however, during Pavulon@-Vetalar” anesthesia. Note the de-

crease in CM amplitude during contralateral stimulation (b

versus a) This effect disappears during total muscular relaxa-

tion (c-d).

Fig. 7 (traces a and b). These show that con- tralateral stimulation with a stimulus of 97 dB

SPL at 1.5 kHz causes a decrease of the CM amplitude of the ipsilaterally administered stimu- lus (67 dB SPL, 1.5 kHz). The lower two traces (c and d) show the same stimulus situation during total muscle relaxation by means of Pavulon@-

Vetalar @ anesthesia. Contralateral stimulation (trace d) did not affect the CM amplitude any- more. Inspection of the tensor tympani muscle

with the operation microscope did not reveal any sign of contraction in this situation.

In Fig. 8a the mean change in CM amplitude of ipsilateral stimuli of 67 dB SPL at various fre-

quencies (+/ - 1 SD) during 97 dB SPL con- tralateral stimulation (1.5 kHz) is shown ( N = 6). As can be seen in the figure, for most intensities, the maximal change was found at 1 kHz, whereas at lower and higher frequencies the decrease was less. The SD is relatively large. Fig. 8b shows the same curves for the various intensities of the con- tralateral stimulus. At all contralateral intensities 1 kHz is influenced most. This is shown in another way in Fig. 8e: the contralateral stimulus intensity (1.5 kHz) required to induce a fixed decrease in

215

Measulng Frequency (kt-+zJ

65 70 75 80 85 90 95 100 105 110

C Sttmulus lntenslty (dElSPL)

65 ,

I . . ..I

02 1

B Measwng Freauency CkHz)

g -7 -..-..*_. f *;L --‘--------o._.._~~

---______ -..-ir- ..-.. ----O -+ - .__ *

5- ~~~_~...,~_.%_

4

‘---__&, “--::.;_*_

-..:;y_ ---._ * -_p._$

c --.. --_*

a 5 ‘0

--.

65 70 7s 80 85 90 95 loo 105 110

D Stimulus lntenslty (dBSPL)

02 1 IO

E Measuring Frequency (ktiz)

Fig. X. The effect of contralaterally induced contraction of the middle ear muscles on CM amplitudes of ipsiiaterally administered

tones. (a) The effect of contralateral stimulation (97 dB SPL, 1.5 kHz) on the mean CM amplitudes (+ / - 1 SD) of ipsilaterally

administered 67 dB SPL tones with various frequencies, (b) The same form as the previous; with increasing contralateral stimulus

mtensity (77-107 d3 SPL) (O- o: 77 dB SPL: A-A: 87 dB SPL; l - 0: 97 dB SPL; + ~ + : 107 dB SPL). (c),

The same stimulus situationas in (a-b); decrease in CM amplitude of 1 kHz tones plotted as a function of contralateral stimulus

intensity (+ / - 1 SD). (d) The same form as (c), for different frequencies (0.2558 kHz) (+ - +: 0.25 kHz; .- A*: 0.5

kHz; O- l : 1 kHz: A -A: 1.5 kHz; o- o: 2 kHz; 8----- I: 4 kHz; Cl- 0: 8 kHz. (e), iso-decrease in CM

amplitude curves; frequency (ipsilateral) as a function of stimulus intensity (contralateral). o -0: 1.5 dB: CI- D: 2 dB;

0-W 2.5 dB: A- A: 3 dB; .- .: 4.5 dB; n -a: 6 dB. (a-e: N = 6).

1 IO

Sttmulus Frequency (kHz) B Sttmulus Frequency (kt-iz)

0 St8mulus Intensity k33SPi_)

02 1 10

E Stimulus Frequency (kHr)

Fig. 9. The influence of the frequency and the intensity of contrrdateral acoustic stimuli in eliciting the middle ear muscle reflex. (a)

Mean change in CM amplitude (+ / - 1 SD) of a 67 dB SPL 1.5 kHz tone (averaged response of 50 tone bursts administered at a

repetition rate of 3.1 Hz, stimulus rise time: 1 ms; single stimulus duration: 15 ms) as a function of the contralateral stimuhrs

frequency at 97 dB SPL (eliciting a middle ear muscle reflex). (bj The same form as the previous, at 77-107 dB SPL. (O-O:

77-84 dB SPL; A- A: 87-94 dB SPL; +----- l : 97-104 dB SPL: + ~ f: 107-114 dB SPL). (c) The change in CM amphtude plotted as a function of (contralateral) stimulus intensity at 1 kHZ ( +/ - 1 SD). (d) The same form as the previous, at

0.25-8 kHz. ( + - + : 0.25 kHz; .----- A: 0.5 kHz; 0 -0: 1 kHz; A- A: 1.5 kHz; Q- 0:2kHz;W------ m: 4 kHz. (e) Ho-change in CM amplitude curves as a function of the contralateral stimulus frequency, o -0: 3 dB; 0-U: 6

dB. (a-e: N = 6).

217

CM amplitude depends on the (ipsilateral) mea- suring frequency. The decrease is maximal at 1 kHz. In Fig. 8c and 8d the change in CM ampli-

tude of various frequencies is plotted versus the intensity of the contralateral stimulus. For 1 kHz (Fig. 8c), the CM amplitude decreases with in-

creasing stimulus intensity. The mean maximal

decrease in CM amplitude at this frequency is 9-10 dB. Fig. 8d shows the intensity dependence

at other frequencies. At 1.0. 1.5, 2.0 and 4.0 kHz we also measured the effect of the middle ear muscle reflex on the CM amplitude of a 57 dB

SPL (instead of the standard 67 dB SPL) sub- threshold ipsilateral stimulus. The signal to noise

ratio was reduced but the relative changes of the CM amplitude remained the same within 5510%.

Apparently no substantial subthreshold facilita- tion occurs (see Simmons. 1965; Blood and

Greenburg, 1981). Fig. 9 shows the influence of the frequency and

the intensity of the contralateral (middle ear

muscle reflex eliciting) stimulus on the CM ampli- tude of a 67 dB SPL, 1.5 kHz ipsilateral stimulus. Fig. 9a, b illustrate that the frequencies above 2

kHz are the most effective stimulus frequencies in

inducing a middle ear muscle reflex. This trend of lower frequencies being less effective then higher frequencies is found for all stimulus intensities.

Furthermore, the CM amplitude decreases some- what more with increasing stimulus intensity. This

is more clearly shown in Fig. 9c, d, where attenua- tion is plotted against stimulus intensity. The ef- fectiveness of the contralateral stimulus is shown in another way in Fig. 9e. The contralateral stimulus intensity required to induce a fixed de- crease in CM amplitude (1.5 kHz) depends on the

(contralateral) stimulus frequency. 2.5 kHz is the most effective stimulus. A similar curve is ob-

tained when the ipsilateral frequency was changed to 4 kHz (data not shown).

Discussion

The aim of this study was to investigate the latency, the rise time and the effect on sound transmission of the middle ear muscle reflex in the rat.

The latency of the acoustic reflex can be mea- sured by various methods. However, the most

direct methods are EMG and the recording of the CM. From our recordings of the CM it was shown that middle ear muscle reflex latency in the rat is

very short. While it was shown that both muscles

take part in the acoustic reflex in the rat. no differentiation in latency of the two muscles could

be revealed. When the tendon of the tensor

tympani muscle was cut the latency did not change. This would indicate that the latency of the stapedius muscle reflex is at least the same, or

possibly shorter than that of the tensor tympani muscle reflex. In most mammals the latency of the

stapedius muscle is shorter than that of the tensor tympani muscle, while in man only the stapedius

muscle is believed to be active in the acoustic reflex (Borg et al.. 1984). The EMG’s of the tensor

tympani muscle confirm that the reflex latency of the muscle is shorter than the latency of the reflex measured at the level of the cochlea by means of

CM recordings. This difference can be attributed to the time lapse between activation of the muscles

(EMG) and, via the influence on the biomechani- cal system of the ossicular chain, the ultimately

resulting effect on the amplitude of the CM. Com-

pared to most other mammals the latency of the tensor tympani response in the rat measured by

EMG is relatively short (See Wersall. 1958: Fisch and Schulthess. 1963). Only the EMG latency of

tensor tympani responses in the mouse and cat were shown to be as short as about 4 and 7 ms.

respectively (Horner. 1986; Eliasson. 1955). One should be careful when comparing all these results

as the latency depends on various stimulus char- acteristics as rise time, frequency and intensity of

the activating stimulus (Bosatra et al., 1984). Fur- thermore. as was shown in this study, latency

depends on the method of measurement (e.g. EMG. ECoch or impedance). In our opinion, the

latency determined by the diminuation of the am- plitude of the CMs seems to be the most relevant one, as it reflects the influence of the muscles on sound transmission at the level of the cochlea.

The change in CM amplitude in time can be considered as the result of contraction of the middle ear muscles as, firstly. this phenomenon disappeared during total muscular relaxation, and, secondly, it was influenced by tenotomy of the tensor tympani muscle. The time lapse between the beginning and the end of the decrease of the

21x

CM amplitude reflects the rise time of the reflex and can be considered as a measure for the con- traction time of the muscles. The rise time of the

reflex together with its CM-latency determines the clinically important time lapse between the begin- ning of the acoustic stimulus and the effect of

middle ear muscle contraction at the level of the cochlea. As far as we know measurements of the rise time of the muscles by CM recordings have

not been done. Measurements of the contraction

time have been done by stimulating the motor nerves of the muscles (Werdll, 1958; Teig, 1972a).

Teig (1972a) showed that with supramaximal

stimulation of the motor nerves, the contraction time of the muscles in the cat and and rabbit was 20-30 ms. This confirmed findings of Wersgll

(1958) who found a contraction time of the stapedius muscle in the rabbit of about 20 ms. The rise time of about 7-12 ms for both muscles

together in our study is very short and supports the results of morphological studies which showed

that both muscles are composed of mainly fast twitch muscle fibres (Berge and Wirtz, 1989a; Berge and Wirtz, 1989b; Berge and Wal, 1990). In this study no exact differentiation could be made

between the rise time of the tensor tympani and stapedius muscles. However, it was shown that the rise time of the stapedius muscle alone was shorter

than that of the two muscles together. This can be explained by either a shorter latency of the

stapedius muscle or a longer rise time of the tensor tympani muscle. The relatively short rise time of the reflex with both muscles intact to- gether with the relatively short latency of the

reflex indicate that the acoustic reflex in the rat has a very fast reaction pattern. This fast reaction pattern is an important characteristic in the light of prevention of noise damage to the inner ear.

With regard to the quantitative effect of the middle ear muscle reflex on sound transmission it became clear that the mean decrease of CM am- plitude in the rat is limited to a maximum of about 10 dB at I kHz. From the slope of the CM amplitude versus stimulus intensity curve at 1 kHz (Fig. 2b), it can be calculated that this corre- sponds with an attenuation of sound transmission of about 15 dB SPL. This is less in comparison with findings in, for example, the cat, the guinea pig and the rabbit (Wiggers, 1937; Wever and

Bray, 1937. 1942; Wever and Vernon. 1955; Teig. 1972a; Teig, 1973) and in man (Borg, 1968; Borg and Zakrisson, 1974). However several aspects have to be taken into account.

Firstly, in this study measurements were per- formed leaving both middle ear musles and their innervation intact, while in a number of other

studies attenuation was measured after tenotomy or attenuation was calculated from indirect meth- ods like monaural masking in patients with stape-

dial paralysis (Borg and Zakrisson, 1974) and con-

traction force studies (Wever and Bray, 1942; Teig,

1972a). Under the experimental conditions of this

study, when the middle ear muscles are activated, attenuation occurs, resulting in a diminished sound level. As a result the contraction of the muscles

might be adjusted to this lower level (as was shown in Fig. 3 curve b), ultimately resulting in a

steady state. Secondly, measurements were performed only

up to 107 dB SPL for most frequencies. This was done because measurements took many hours and had appeared to result in inner ear damage at these high sound levels in pilot studies. Finaily.

our data were derived from contralateral activa- tion of the middle ear muscle reflex. Activation of the reflex might occur at lower intensities and can

result in higher attenuation when stimulated ipsi- laterally or bilaterally (Guinan and McCue, 1987).

As regards the stimulus intensity-attenuation curves it was shown that the largest effect oc-

curred between 65-77 dB SPL. Further increase of the stimulus intensity did not result in a accord-

ingly larger increase of attenuation. This could possibly be explained by the findings of Teig (1973) who showed in the cat that stronger con-

tractions gave less reduction of sound transmis- sion per gram tension.

The finding that tones of 1 kHz are attenuated most is difficult to explain. In most animals at- tenuation is maximal at lower frequencies (i.e. less than 1000 Hz) and decreases towards higher fre- quencies. However, it seems likely that the acous- tic reflex is adapted to the special needs for a certain species, as was shown, for example, in the bat (Pollak and Henson, 1973). Supposedly, in the rat, 1 kHz is a frequency that needs to be at- tenuated, possibly to avoid masking of higher tones during vocalisation. In addition it appeared

219

that the higher tones were the most effective eiici- tor of the middle ear muscle reflex. This might

indicate the importance of high tones for the rat.

Sensitivy studies of the CM in the rat have shown peak sensitivity at 3 kHz and 40 kHz (Crowley

and Hepp-Reymond, 1966). However, in the Lewis rat we found peak sensitivity at 1.5 kHz (Fig. 2~).

This seems to be paradoxal since attenuation is most at 1.0-1.5 kHz. In summary. the acoustic

reflex in the rat has been shown to be fast reactive both in latency and rise time. The attenuation of 0.25-8 kHz upto stimulus intensities of 115 dB SPL is (while both middle ear muscles and their

innervation are intact) limited to a mean maxi- mum of 15 dB SPL. The tones above 2 kHz

appear to be the best elicitor of the middle ear muscle reflex. However, relevant inter-indivudual

differences appear to exist. One kHz appears to be attenuated the most in the rat, the reason of which remains obscure. In addition to this phenomenon,

interesting aspects for further research into the function of the middle ear muscles would be;

measurements during ipsilateral activation of the muscles, the fatiguability of the reflex, as well as

the activity of the muscles following noise ex-

posure.

Aknowledgements

We thank V. Bongers and C. ter Riet for their contributions to the experiments and J. Huntjens for the technical assistance. This study was sup- ported by the Heinsius-Houbolt foundation.

References

Berge, H. van den, Wirtz, P. (1989a) Detailed morphology of

the tensor tympani muscle of the rat. An integrated light-

microscopical, mo~hometrical, enzyme histochemical, im-

munohistochemical and eiectronmicroscopical study in re-

lation to function. J. Anat. 164. 215-228.

Berge, H. van den, Wirtz. P. (1989b). Detailed morphology of

the stapedius muscle of the rat. An integrated lightmicro-

scopical. morphometrical, enzyme histochemical, immuno-

histochemical and electronmicroscopical study in relation

to function. J. of Anat. 166, 157-166.

Berge, H. and Wal, J.C. (1990) The innervation of the middle

ear muscles of the rat. J. of Anat. 170, 99-109.

Blood I.M. and Greenburg H.J. (1981) Low level acoustic

reflex thresholds. Audiology 20, 244-250.

Borg, E. (1968) A yuantitave study of the effect of the acoustic stapedius reflex on sound transmission through the middle

ear of man. Acta Otolaryngol. (Stockh.) 66. 461-472.

Borg, E. and Zaknsson, J.-E. (1974) Stapedius reflex and

monaural masking. Acta Otolaryngol. (Stockh.) 7X. 155-

161.

Borg, E., Counter, S4.. Riisler, G. (1984) Theories about

middle ear muscle function. In: S. Silman (Ed.). The acoub-

tic reflex: Basic principles and clinical applications

Academic Press, Orlando, Florida.

Bosatra. A., Russolo. M.. and Silverman, C.A.. (1984) Acous-

tic-reflex latency: State of the art. In: S. Silman (Ed.). The

acoustic reflex; Basic principles and clinical applications,

Academic Press. Orlando. Florida.

Crowly. D.E and Hepp-Reymond. M.C. (1966) Development

of cochlear function in the ear of the infant rat. J. Comp.

and Physiol. Psychol. 62, 4277432.

Eliasson. S. and Gisselsson, L. (1955) Electromyographic stud-

ies of the middle ear muscles of the cat. EEG Clin. Neuro-

physiol. 7. 399-406.

Fisch. U. and Schulthess, v. G. (1963) Electromyographic

studies on the human stapedial muscle. Acta Otolaryngol.

(Stockh.) 56. 287.-297.

Guinan J.J. and McCue M.P. (1987) Asymmetries in the acous-

tic reflexes of the cat stapedms muscle. Hear. Res. 26. l-10.

Horner. KC. (1986) The tensor tympani muscle reflex m the

mouse. Hear. Res. 24. 117-123.

Metz. 0. (1946) The acoustic impedance measured on normal

and pathological ears. Acta Otolaryngol. (Stockh.) suppl.

63. l-254.

Metier. A.R. (1965) An expe~mental study of the acoustic

impedance of the middle ear and its transmissi~~n proper-

ttes. Acta Otolaryngol. (Stockh.) 60. 129-149.

Simmons F.B. (1965) Binaural summation of the acoustic

reflex. J. Acoust. Sot. Am. 37. 834-836.

Teig. E. (1972a) Force and contraction velocity of the middle

ear muscles in the cat and the rabbtt. Acta Physiol. Stand.

84. l-10.

Teig. E. (1972b) Tension and contraction time of motor units

of the middle ear muscles in the cat. Acta Physiol. Stand.

84, 11-21.

Teig, E. (1973) Differential effect of graded contraction of

middle ear muscles on the sound transmission of the ear.

Acta Physiol. Stand. 88, 3822391.

Wemall, R. (1958) The tympanic muscles and their reflexes.

Acta Otolaryngol. (Stockh.) suppl. 139, 1-l 12.

Wever. E.G and Bray C.W. (1937) The tensor tympani muscle

and its relation to sound conduction. Ann. of Otol. 46,

947-961.

Wever. E.G and Bray, C.W. (1942) The stapedius muscle in

relation to sound conduction. J. Exp. Psychol. 31. 3543.

Wever E.G. and Vernon J.A. (1955) The effect of the tympanic muscle reflexes upon sound transmission. Acta Otolaryn-

gol. 45. 433-439.

Zakrisson, J.-E. (1979) The effect of the stapedius muscle reflex

on attenuation and poststimulatoy fatigue at different fre-

quencies. Acta Otolaryngol. (Stockh.) suppl. 360. 118-121.