abnoramlities in efferent activities mutism

5/19/2018 Abnoramlities in Efferent Activities Mutism

1/9

E-Mail [email protected]

Original Paper

Audiol Neurotol 2013;18:353361

DOI: 10.1159/000354160

Abnormalities in Auditory EfferentActivities in Children with Selective Mutism

Chava Muchnika, d Daphne Ari-Even Rotha, d Minka Hildesheimera, d Miri Arieb

Yair Bar-Haimb, c Yael Henkina, d

aDepartment of Communication Disorders, Sackler Faculty of Medicine, bSchool of Psychological Sciences andcSagol School of Neuroscience, Tel Aviv University, Tel Aviv, and dHearing, Speech, and Language Center,

Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel

olds and decay functions and the suppression of transient-

evoked otoacoustic emissions, respectively. Auditory affer-

ent function was tested by means of auditory brainstem re-

sponses (ABR). Results indicated a significantly higher

proportion of children with abnormal MEAR and MOCB func-

tion in the SM group (58.6 and 38%, respectively) compared

to controls (9.7 and 8%, respectively). The prevalence of ab-

normal MEAR and/or MOCB function was significantly high-

er in the SM group (71%) compared to controls (16%). Intact

afferent function manifested in normal absolute and inter-

peak latencies of ABR components in all children. The finding

of aberrant efferent auditory function in a large proportion

of children with SM provides further support for the notion

that MEAR and MOCB may play a significant role in the pro-

cess of self-vocalization. 2013 S. Karger AG, Basel

Introduction

Verbal communication entails a continuous interac-tion between speech and hearing mechanisms [Curio etal., 2000; Borg et al., 2009; Ventura et al., 2009]. While wespeak, we continuously monitor the quality of our own

voice and speech in order to cope with incoming mes-

Key Words

Auditory efferent function Selective mutism Middle-ear

acoustic reflex Medial olivocochlear bundle reflex

Audio-vocal interaction

Abstract

Two efferent feedback pathways to the auditory periphery

may play a role in monitoring self-vocalization: the middle-

ear acoustic reflex (MEAR) and the medial olivocochlear bun-

dle (MOCB) reflex. Since most studies regarding the role of

auditory efferent activity during self-vocalization were con-

ducted in animals, human data are scarce. The working

premise of the current study was that selective mutism (SM),

a rare psychiatric disorder characterized by consistent failure

to speak in specific social situations despite the ability to

speak normally in other situations, may serve as a humanmodel for studying the potential involvement of auditory ef-

ferent activity during self-vocalization. For this purpose, au-

ditory efferent function was assessed in a group of 31 chil-

dren with SM and compared to that of a group of 31 nor-

mally developing control children (mean age 8.9 and 8.8

years, respectively). All children exhibited normal hearing

thresholds and type A tympanograms. MEAR and MOCB

functions were evaluated by means of acoustic reflex thresh-

Received: December 10, 2012

Accepted after revision: July 3, 2013

Published online: October 9, 2013

NeurotologyAudiology

Yael Henkin, PhD

Department of Communication DisordersSackler Faculty of MedicineTel Aviv University, Tel Aviv (Israel)E-Mail henkin @ post.tau.ac.il

2013 S. Karger AG, Basel

14203030/13/01860353$38.00/0

www.karger.com/aud


2/9

Muchnik/Ari-Even Roth/Hildesheimer/

Arie/Bar-Haim/Henkin


DOI: 10.1159/000354160

354

sages and to maintain production accuracy [Hood, 1998].An example of such an interaction is the Lombard reflexin which speakers raise their voice when an external noiseis introduced, as the auditory feedback of their own voiceis masked [Lamprecht, 1988].

Evidence for the existence of audio-vocal interaction

mechanisms at different levels of the auditory pathwayswere reported both in animals [Metzner, 1989, 1993; Eli-ades and Wang, 2003; Hage et al., 2006] and humans [Cu-rio et al., 2000; Houde et al., 2002; Ventura et al., 2009].Two efferent feedback pathways to the auditory periph-ery may play a role in monitoring and regulating self-

vocalization: the middle-ear acoustic reflex (MEAR) andthe medial olivocochlear bundle (MOCB) reflex. Thesetwo efferent reflexes have different neural pathways anddifferent peripheral targets. The MEAR consists of twomuscles, the tensor tympani and the stapedius. The path-way begins with the excitation of the auditory nerve

which, in turn, excites neurons within the ipsilateral co-chlear nucleus. Output from the cochlear nucleus thenexcites the motor neurons in the brainstem around theipsilateral facial motor nucleus. In addition, neuronsfrom the cochlear nucleus pass through the trapezoidbody and reach the superior olivary complex (SOC) onboth sides, thereafter exciting facial nerve nuclei both ip-silaterally and contralaterally. The neural circuit of theMEAR controls the contraction of the stapedius uponpresentation of loud sounds. Contraction of the stapediusstiffens the motion of the ossicular chain in the middle ear

and thus attenuates the transmission of the sound to theinner ear. MEAR-induced attenuations are largest forlow-frequency stimuli. Since the stapedius muscle canalso be activated in response to nonacoustic stimuli (e.g.during, and in anticipation of, vocalization) it is assumedthat there should also be additional higher descendingcontrol pathways from somewhere in the central nervoussystem [Liberman and Guinan, 1998; Gelfand, 2009].

The olivocochlear efferent system originates in theSOC and contains two fundamental subsystems: themedial and lateral olivocochlear bundles (MOCB andLOCB). The MOCB consists of myelinated fibers that

originate in the medial portion of the SOC on both sidesof the brain and project through the vestibular nerve, di-rectly onto the outer hair cells. The LOCB consists of un-myelinated fibers that originate predominantly on the ip-silateral side of the brain. Their axons also pass throughthe vestibular nerve and innervate the auditory nerve fi-bers under the inner hair cells [Guinan, 2006]. While theMOCB reflex is better understood than the LOCB, itscourse is complex and composed of the ipsilateral and

contralateral reflexes [see review by Guinan, 2006]. Acti-vation of the MOCB synapses on outer hair cells alterstheir properties and as a result reduces their contributionto the amplification of basilar membrane motion. Similarto the MEAR, the MOCB reflex can be evoked by ipsilat-eral, contralateral or binaural acoustic stimulation. In

contrast to the MEAR, MOCB-induced attenuations arelargest for mid- to high-frequency sounds [Liberman andGuinan, 1998]. Furthermore, the MOCB reflex is moreefficient when otoacoustic emissions are recorded at lowintensities. This leads to the assumption that the MOCBsystem (at least the uncrossed fibers) plays a more impor-tant role at low intensity. Thus, complementarity betweenMEAR and MOCB is plausible.

The involvement of the MEAR in the process of self-vocalization was already reported a few decades ago. TheStapedius muscle is active during vocalization both in an-imals [Carmel and Starr, 1963; Henson, 1965] and humans

[Klockhoff, 1961; Borg and Zakrisson, 1975]. When theMEAR is activated by self-vocalization, it is assumed tomoderate overloading of the sensory receptors of the co-chlea, and thereby maintains a fairly constant level of sen-sitivity. It is likely that aberration or inconsistency in mid-dle-ear muscle activity could occasionally cause auditoryadaptation or temporary threshold shift in the speech fre-quency range during self-vocalization [Curio et al., 2000].Another role of the MEAR that may be related to self-vo-calization is the prevention of interference associated withdecreasing the masking produced by the speakers own

voice (i.e. antimasking effect) even at normal vocal efforts.This may result in improved intelligibility of simultaneousincoming external speech [Borg and Zakrisson, 1975].

Most of the research regarding the MOCB functionduring vocalization has been conducted in animals. Au-ditory-vocal interaction was demonstrated by single-unitrecordings from the auditory cortex, medial geniculatebody, inferior colliculus and superior olivocochlear com-plex [Metzner, 1989, 1993; Eliades and Wang, 2003; Hageet al., 2006]. Data from animal models such as the singingcricket [Poulet and Hedwig, 2002] and mustached bat[Goldberg and Henson, 1998] provide compelling evi-

dence pointing to activation of an inhibitory action of theMOCB during self-vocalization. For example, intercellu-lar recordings from the singing cricket showed that pre-synaptic inhibition of auditory afferent and postsynapticinhibition of an interneuron occur in phase with the songpattern. Inhibition reduced the auditory interneuronsresponse to self-generated sounds and thus protected thecrickets auditory pathways from self-induced desensiti-zation [Poulet and Hedwig, 2002].


3/9

Auditory Efferent Function in Childhood

Selective Mutism


DOI: 10.1159/000354160

355

In humans, MOCB function can be assessed by thesuppression of transient-evoked otoacoustic emissions(TEOAE) during presentation of contralateral noise [Col-let et al., 1990; Berlin et al., 1993]. The functional role ofthe MOCB during vocalization, however, is still not fullyclear and is under continuous investigation [Robertson,

2009]. Some studies conducted in normal-hearing partic-ipants provide support for the view that under certainconditions the MOCB may play an antimasking role dur-ing speech perception in background noise [Micheyl andCollet, 1996; Giraud et al., 1997; Kumar and Vanaja, 2004].

Better understanding of the involvement of efferentactivity in audio-vocal interactions in humans may po-tentially be gained by studying clinical populations exhib-iting auditory efferent dysfunction. For example, reducedMEAR function has been observed in children with Wil-liams syndrome [Gothelf et al., 2006], and altered MOCBfunction was documented in children with dyslexia

[Veuillet et al., 2007], infantile autism [Khalfa et al., 2001]and auditory processing disorders [Muchnik et al., 2004;Sanches and Carvallo, 2006]. Nonetheless, these reduc-tions in MEAR or MOCB function were not reported tomanifest in aberrations in audio-vocal interactions.

The working premise of the current study was that se-lective mutism (SM), a rare psychiatric disorder charac-terized by consistent failure to speak in specific social sit-uations (e.g. school) despite the ability to speak normallyin other situations (e.g. home) [DSM-IV-TR; AmericanPsychiatric Association, 2000], may serve as a human

model to study the potential involvement of auditory ef-ferent activity during self-vocalization. Comorbid char-acteristics of SM include anxiety, shyness, timidity andsocial withdrawal [Sharkey and McNicholas, 2008]. SMhas also been associated with school failure, rejection bypeers and aggravated intrafamilial relationships. Preva-lence rates of SM range between 0.47 and 0.76% [Viana etal., 2009], and its etiology is still unknown [Sharkey andMcNicholas, 2008].

In a previous study we hypothesized that aberrant au-ditory efferent function may underlie deficient auditoryprocessing during self-vocalization and thus impair the

ability of children with SM to simultaneously speak andprocess incoming auditory signals [Bar-Haim et al.,2004]. We further speculated that faced with the negativeconsequences of vocalization on the capacity to processexternal sounds, children with SM may adapt by whisper-ing, restricted vocalization and speech avoidance. Thishypothesis was in line with reports in the literature citingchildren with SM describing distortions in the perceptionof their own voice (e.g. my voice sounds funny and I

dont want others to hear it [Black and Uhde, 1992] ormy brain wont let me speak because my voice soundsstrange [Boon, 1994]). Indeed, preliminary results froma group of 16 children with SM indicated that abnormalauditory efferent activity was evident in two thirds of thesample [Bar-Haim et al., 2004]. Compared with normally

speaking control children, children with SM displayedsignificant aberrations in MEAR thresholds and decayfunctions and a diminished suppression effect of TEOAE,indicating reduced activity of efferents from the MOCB.

The goal of the current study was, therefore, to expandand substantiate our preliminary finding of aberrant ef-ferent function in children with SM [Bar-Haim et al.,2004]. By studying an enlarged sample of children withSM and normally developing controls, and by applyingmore stringent criteria for abnormal MEAR and MOCB,we sought to elucidate the involvement of auditory effer-ent activity in SM.

Materials and Methods

Participants

A total of 62 children participated in the study, 31 diagnosedwith SM (mean age = 8.9 years, range = 5.216.8 years, SD = 3.1,22 females) and 31 healthy, normally developing and freely speak-ing controls (mean age = 8.8 years, range = 4.915.5 years, SD =2.6, 18 females). Participants with SM were recruited through ad-vertisement and referrals from affiliated clinics informed of thestudy. Control children were recruited from public schools in the

same district areas as the children in the SM group.All children were screened for psychopathology using a struc-tured research diagnostic interview: Schedule for Affective Disor-ders and Schizophrenia for School Age Children present andlifetime version (K-SADS-PL) [Kaufman et al., 1997] conductedwith the parents by an experienced clinical psychologist. Diagnosisof SM was further substantiated using DSM-IV-TR criteria[American Psychiatric Association, 2000] by means of a semistruc-tured interview with the parents. Furthermore, normal speechproduction at home was verified through homemade audio- orvideotapes of the children fluently conversing with members oftheir nuclear family.

Prevalence of comorbid psychiatric diagnoses included socialphobia (53%), simple phobia (16%) and enuresis (17.7%). All Chil-

dren from the SM and control groups were medication-free basedon parental reports. Children in the control group had to be freeof any current or lifetime diagnosis. Children with SM with comor-bid developmental delays or signs of pervasive developmental dis-order were excluded from the study.

For all participants, audiological inclusion criteria were the fol-lowing: (1) hearing thresholds for pure tones within the normalrange (air conduction thresholds at 0.5, 1, 2 and 4 kHz 15 dB HL)in both ears, (2) intelligibility of phonetically balanced monosyl-labic words within the normal range and (3) type A tympano-grams.


4/9




DOI: 10.1159/000354160

356

The study was approved by the Institutional Review Board ofthe Sheba Medical Center. Informed consent was obtained fromparents and assent was obtained from children.

Methods

Middle-Ear Acoustic ReflexA Grason-Stadler GSI-33 middle-ear analyzer (V.2) was usedfor the MEAR measurements at 226 Hz probe tone: (1) the acous-tic reflex thresholds (ART) and (2) the reflex decay test (RDT).Both tests were conducted in the ipsi- and contralateral mode ofstimulation, for each ear. Order of ear and mode of stimulationwere counterbalanced between participants.

ARTs were obtained at 0.5, 1 and 2 kHz. The threshold was de-fined as the lowest intensity level of the tone needed to elicit a 0.2-ml decrease in middle-ear admittance on at least two of three trials.The initial presentation level was 70 dB HL and levels were elevat-ed by 5-dB steps until threshold was detected.

RDTs were obtained at 0.5, 1 and 2 kHz. During each test run,the stimulus was presented 10 dB above the ART at the tested fre-

quency for a period of 10 s. For both ART and RDT measurementsthe order of tested frequency was randomized.The criterion for absent MEAR in the present study was de-

fined as no response at maximum output (ipsilateral mode: 110,105, 100 dB HL; contralateral mode: 120, 120, 120 dB HL for 0.5,1 and 2 kHz, respectively) of the middle-ear analyzer in at least twoout of six conditions in the same ear: three tested frequencies (0.5,1 and 2 kHz) at either mode of stimulation (ipsi- or contralateral).

The criterion for abnormal MEAR decay was defined as a de-cline of 50% or more in MEAR amplitudes within the 10 s of test-ing in at least two out of four conditions in the same ear: two test-ed frequencies (0.5 and 1 kHz) at either mode of stimulation (ipsi-or contralateral).

TEOAE RecordingsThe TEOAE were recorded using an ILO92 Echoport OAE an-alyzer V.42 with a SDG-type probe in both ears in a counterbal-anced fashion. The quick screen mode was used (recording win-dow of 2.512.5 ms). Click stimuli were produced by 80-s rectan-gular electric pulses presented at 80/s in the nonlinear mode ofstimulation. The stimulus level as measured in the sealed ear canalwas adjusted to a peak pressure of 80 3 dB SPL. The noise rejec-tion level was set at 54.9-dB peak-equivalent SPL. Emissions wereaveraged in response to 260 sweeps.

The presence of a reliable response was determined by a wholereproducibility level 50% and a signal-to-noise ratio 3 dB inthree of the four frequency bands (1.6, 2.4, 3.2 and 4 kHz).

MOCB Function: Suppression of TEOAEThe TEOAE suppression effect was tested via six successiveTEOAE measurements alternately without and with contralater-al acoustic stimulation (CAS). Measurements were always con-ducted first without noise, in order to ascertain the presence ofTEOAEs at the intensity of 71-dB peak-equivalent SPL, which waschosen for testing the suppression effect. This intensity was foundto be more effective for obtaining higher suppression values inprevious studies conducted in our laboratory. The CAS was com-posed of white noise produced by a B200C Beltone audiometer anddelivered through an SM-N insert earphone. The CAS was pre-

sented at 40 dB sensation level above the behavioral threshold towhite noise (which was obtained prior to the TEOAE measure-ments for each participant). The suppression effect was calculatedby subtracting the mean TEOAE amplitude value with CAS fromthe mean TEAOE amplitude value without CAS.

The criterion for abnormal TEOAE suppression was defined as0.5 dB SPL or less in at least one ear, the lower cut-off level havingbeen introduced by Prasher et al. [1994].

Recording of Auditory Brainstem ResponsesIn order to assess auditory afferent functionality, auditory

brainstem responses (ABRs) were recorded in both ears in a coun-terbalanced fashion using the Bio-logic Navigator Pro evoked po-tential system. Electrodes were placed at Cz and at the earlobe ip-silateral to the stimulated ear. A ground electrode was placed at thecontralateral earlobe. Impedance was kept below 5 k. Responseswere amplified with a gain of 100,000 and digitally filtered with abandwidth of 0.13 kHz. Each ear was stimulated by alternating85-dB nHL clicks with presentation rate of 21/s. Clicks were deliv-ered using ER-3 insert earphones. Responses in each conditionwere averaged over 2,000 individual sweeps, with a sweep time of16 ms. Peak absolute latencies of ABR waves I, III and V were reli-ably obtained from all tested participants. Interpeak latencies IIII,IV and IIIV were also calculated.

ProcedureChildren from both groups who met the studys inclusion cri-

teria were invited for a day of audiological assessment. Efferentactivity was assessed by means of MEAR thresholds, MEAR decayfunction, and contralateral suppression of TEOAE. In addition,afferent auditory function at the acoustic nerve and brainstem lev-els was assessed by means of ABR. All tests were carried out in asound-treated room.

Statistical AnalysisOne-way (group) analysis of variance (ANOVA) with repeat-

ed measures (ear) was conducted to test the differences betweenthe groups (SM, controls), ears (right, left), and group-by-ear in-teraction for all ABR absolute and interpeak latencies. Two-wayANOVA was conducted to test the effect of group, ear and group-by-ear interaction on TEOAE amplitude and suppression values.Fishers exact or 2tests were used to assess the difference in prev-alence of abnormal between-group findings and ears for the fol-lowing parameters: (1) absent MEAR, (2) abnormal MEAR decay,(3) abnormal suppression of TEOAE and (4) combination of ab-normal findings in the MEAR and/or MOCB. Differences wereconsidered statistically significant when p < 0.05.

Results

Auditory Brainstem ResponsesABR measurements were obtained from 26 children

with SM and from all children (31) in the control group;5 children from the SM group did not comply with thetesting procedure that required lying still on a bed witheyes closed for 1520 min. For all tested children from theSM and control groups, absolute latencies of waves I, III


5/9


Selective Mutism


DOI: 10.1159/000354160

357

and V of the ABR as well as interpeak latencies IIII, IVand IIIV were within the normal range. One-wayANOVA with repeated measures indicated nonsignifi-cant main effects of group, ear and group-by-ear interac-tion for all absolute and interpeak latencies (table 1).

Middle-Ear Acoustic ReflexPrevalence of Absent MEARThe prevalence of cases exhibiting absent MEAR was

significantly higher in the SM group (11/31 subjects,35.5%) compared to the control group (1/31, 3.2%; Fish-ers exact test, p = 0.003).

The number of right versus left ears that showed ab-sent MEAR was similar in the SM group (9/31 and 7/31ears, respectively; 2= 0.34, p = 0.56). In the control grouponly one (left) ear showed absent MEAR.

MEAR Decay

The MEAR decay test was performed in 29 childrenwith SM due to absent MEAR or elevated thresholds thatprevented recording of MEAR decay (output limitationsof the middle-ear analyzer system). The prevalence of cas-es exhibiting abnormal MEAR decay was significantlyhigher in the SM group (13/29, 45%) compared to thecontrol group (3/31, 9.7%; Fishers exact test, p = 0.003).

The number of right versus left ears that showed ab-normal MEAR decay was similar in both study groups

(SM group: 9/29 and 8/29, right and left, respectively;2= 0.08, p = 0.77; control group: 2/31 and 2/31, right andleft, respectively).

Prevalence of Absent MEAR and/or AbnormalMEAR DecayTable 2 summarizes the prevalence of cases in which

only absent MEAR, or only abnormal MEAR decay, or acombination of both was evident in the SM and controlgroups. The total prevalence of abnormal findings inMEAR tests was significantly higher in the SM group(17/29, 58.6%) compared to the control group (3/31,9.7%; Fishers exact test, p = 0.0001).

Table 1.Click-ABR absolute peak and interpeak latencies for the left and right ears in the SM and control groups

I III V IIII IIIV IV

SMRight (n = 26) 1.470.11 3.610.14 5.410.17 2.140.13 1.790.13 3.940.16Left (n = 25) 1.470.11 3.670.19 5.450.20 2.20.20 1.780.12 3.980.21

ControlRight (n = 31) 1.490.10 3.630.15 5.450.17 2.140.16 1.820.12 3.960.20Left (n = 31) 1.50.11 3.670.18 5.460.18 2.160.18 1.790.15 3.960.16

Group effectd.f. (n = 1, d = 55)F 1.37 0.03 0.35 0.28 0 0.46p value 0.24 0.86 0.56 0.6 0.96 0.5

Ear effectd.f. (n = 1, d = 55)F 0.0 1.94 1.71 1.19 1.2 0.03p value 0.97 0.17 0.2 0.28 0.28 0.87

Group ear interactiond.f. (n = 1, d = 55)F 0.06 0.23 0.86 0.32 0.88 0.24p value 0.8 0.64 0.36 0.57 0.35 0.62

Latency values are expressed as means SD, in milliseconds.

Table 2.The prevalence of cases exhibiting absent MEAR, abnor-mal MEAR decay, or both in the SM and control groups

AbsentMEAR

AbnormalMEAR decay

Absent MEAR/abnormal MEAR decay1

Total

SM 4/29 8/29 5/29 17/292

Control 0/31 2/31 1/31 3/31

1 Nonoverlapping conditions (e.g. different mode of stimula-tion/different ear).

2 The number of SM children in which both MEAR thresholdsand decay could be performed.


6/9




DOI: 10.1159/000354160

358

MOCB FunctionTEOAE amplitude and suppression measurementswere obtained from 21 children with SM and from 24control children in at least one ear. This was due to chil-drens inability to comply with the TEOAE suppressiontesting procedure that required sitting still for approxi-mately 30 min. Table 3 provides mean TEOAE amplitudeand suppression values of the study groups in the rightand left ears. Two-way ANOVA revealed that the maineffects of group, ear and the group-by-ear interaction onTEOAE amplitudes were not significant (group, p = 0.56;ear, p = 0.76; group-by-ear interaction, p = 0.32). Group

differences in TEOAE amplitudes were further analyzedin the four frequency bands (1.6, 2.4, 3.2 and 4 kHz). Re-sults indicated that for all participants from both groupsemissions were present for the tested frequency bands,excluding 3 cases from the SM group and 1 case from thecontrol group at 1.6 kHz.

Analysis of the suppression data indicated a significantmain effect of group [F(1, 75) = 5.89; p = 0.02] that man-ifested in lower TEOAE suppression values in the SM

group compared to the control group. The main effect ofear and the group-by-ear interaction were not significant(p = 0.88 and 0.42, respectively).

The prevalence of cases with abnormal TEOAE sup-pression was significantly higher in the SM group (8/21,38%) compared to the control group (2/24, 8%; Fishersexact test, p = 0.03). The number of right versus left earsthat showed abnormal TEOAE suppression did not differsignificantly in both groups [SM group: 4/17 (23.5%) vs.6/19 (31.8%), respectively; Fishers exact test, p = 0.7; con-trol group: 1/23 (4.3%) vs. 1/20 (5%), respectively].

Prevalence of Abnormal MEAR and MOCB FindingsWe further evaluated the prevalence of cases exhibit-

ing abnormal efferent function as demonstrated by find-ings of abnormal MEAR function (i.e. absent MEAR and/or abnormal MEAR decay), MOCB function (i.e. sup-pression of TEOAE) and the combination of both MEARand MOCB. We analyzed data of 21 children from the SMgroup and 24 children from the control group who hadfull data sets. Figure 1 shows that 71% of the children with

Table 3.TEOAE amplitude and suppression values by ear in the SM and control groups

Group SM Control Group effect(d.f. = 1, 75)

F, p value

Ear effect(d.f. = 1, 75)

F, p value

Group ear interaction(d.f. = 1, 75)

F, p valuen mean SD n mean SD

TEOAE amplitude 0.32, 0.57 0.15, 0.7 0.98, 0.32Right ear 17 13.554.57 23 13.173.95Left ear 19 12.293.78 20 13.723.96

TEOAE suppression 5.82, 0.02 0.01, 0.93 0.65, 0.42Right ear 17 1.030.75 23 1.710.79Left ear 19 1.181.05 20 1.521.09

Values are expressed in decibels.

8%

8%

84%

Control

29% 33%

19%19%

SM

Only MEAR

Only MOCB

MEAR and MOCB

Normal

Fig. 1. The prevalence of abnormal find-ings in MEAR, MOCB, and the combina-tion of abnormal MEAR and MOCB func-tion in the SM and control groups.


7/9


Selective Mutism


DOI: 10.1159/000354160

359

SM (15/21) showed abnormal findings in MEAR and/orMOCB, whereas only 16% of the control children (4/24)showed abnormal findings (Fishers exact test, p = 0.0003).

Discussion

In the current study human auditory efferent func-tionality was studied in children with SM, a rare psychi-atric disorder characterized by consistent failure to speakin specific social situations despite the ability to speaknormally in other situations. Our goal was to expand andsubstantiate our previous preliminary findings [Bar-Haim et al., 2004] in a larger cohort of children with SMusing more stringent criteria for abnormal MEAR andMOCB function. The rationale for studying this clinicalgroup was based on evidence suggesting involvement ofthe middle-ear efferent system during self-vocalization

[Klockhoff, 1961; Borg and Zakrisson, 1975]. The find-ings indicate that the majority of children with SM (71%)demonstrated auditory efferent abnormalities manifestedin aberrant MEAR and/or MOCB function, thus support-ing and extending our previous assertion that abnormalefferent activity may underlie reduced self-vocalization atleast in some children with SM. As previously reported[Bar-Haim et al., 2004; Arie et al., 2007; Henkin et al.,2010] auditory afferent function was found to be intact inchildren with SM as demonstrated by normal hearingthresholds, normal speech discrimination and normal

ABR absolute and interpeak latencies. These findingspresumably rule out afferent dysfunction in the presenceof efferent aberrations.

The prevalence of children with SM who exhibited ab-sent MEAR and/or abnormal MEAR decay was 59%, sig-nificantly higher than that found in the control group(10%). Applying a more stringent criterion in an enlargedgroup in the current study resulted in a similar prevalencecompared to that previously reported in a smaller cohortof children with SM (63%) [Bar-Haim et al., 2004]. Fur-thermore, the use of more stringent criteria resulted in adecrease in the prevalence of abnormal MEAR findings

in the healthy control group (10%) compared to that pre-viously reported (37%) [Bar-Haim et al., 2004]. These re-sults suggest that applying more stringent criteria forMEAR abnormality does not change the detection ratesin children with SM but reduces detection rates in healthycontrols that were marginally aberrant in our previousreport.

The MEAR is known to be activated during vocaliza-tion in humans and is thought to play an important role

in reducing distortions, nonlinearities and upward spreadof masking [Borg and Zakrisson, 1975]. It is assumed thatthe MEAR activity during vocalization has the effect ofdecreasing the masking influence of the speakers own

voice. This results in an improved capacity of the speakerto hear other external sounds while vocalizing [Borg and

Counter, 1989]. The finding of abnormal MEAR functionin the current cohort of children with SM provides fur-ther support to the idea that reduced auditory efferentactivity during self-vocalization may restrict the ability ofSM children to simultaneously process incoming audi-tory signals [Bar-Haim et al., 2004]. In fact, in a previousbehavioral study we showed that children with SM whopresented aberrations in auditory efferent activity, com-pared to children with SM without auditory efferent de-ficiency and healthy controls, exhibited impaired audi-tory processing during vocalization [Arie et al., 2007].

Absent MEAR has also been reported as a major char-

acteristic of children with Williams syndrome [Gothelf etal., 2006], a multiple congenital syndrome affecting the

vascular, connective tissue and central nervous system.Interestingly, two of the core features of Williams syn-drome are hyperacusis and phonophobia. Gothelf et al.[2006] postulated that MEAR dysfunction may underlieoversensitivity to sound due to excessive exposure to loudsounds, leading these children to avoid noisy environ-ments. In other words, for individuals with MEAR dys-function, certain environmental sounds could be per-ceived as disproportionately noisy or distorted. It may be

the case that aberrant MEAR function may underlie, atleast to some extent, averseness to loud environmental(external) sounds in children with Williams syndrome,and to ones own voice in children with SM. Further sup-port for this notion is the finding of a higher proportionof introverted, socially withdrawn children and adults ex-hibiting abnormal MEAR compared to their extravertedpeers, and coinciding with their increased auditory sensi-tivity and preference for more quiet environments [Bar-Haim, 2002].

An additional efferent reflex that is assumed to play arole in monitoring self-vocalization is the MOCB reflex.

The prevalence of reduced MOCB function in childrenwith SM was 38 versus 8% in the control group. Applyingthe lower cut-off criterion of Prasher et al. [1994] for ab-normal suppression (i.e. 0.5 dB) in the current study re-sulted in a lower prevalence compared to that found inour previously reported smaller cohort (58%) [Bar-Haimet al., 2004]. Prevalence of abnormal findings in the healthycontrol group, however, was also significantly lower (8 vs.33%, current and previous study, respectively).


8/9




DOI: 10.1159/000354160

360

The relevance of the MOCB reflex function to the pro-cess of self-vocalization has been studied predominantlyin animal models and may be associated with the un-masking phenomena. For example, in anesthetized cats,single auditory nerve responses to tone bursts in noisewere measured in two conditions, with and without con-

tralateral noise (noise and quiet conditions). In thequiet listening condition MOCB activation inhibited theresponse, while in the noise condition MOCB enhancedthe auditory nerve response by reducing the response tothe noise [Kawase et al., 1993]. In humans, reducedMOCB activity manifested in smaller suppression valuesof TEOAE in children with learning disabilities [Veuilletand Collet, 1999] and in children with auditory process-ing disorders [Muchnik et al., 2004; Sanches and Carval-lo, 2006; Yalcinkaya et al., 2010] known to exhibit diffi-culties understanding speech in background noise.

In the present study no laterality effect on TEOAE

amplitude and suppression was found for both groups.One factor that may affect TEOAE laterality is handed-ness. There are, however, controversial reports, withsome studies showing nonsignificant effects [Driscoll etal., 2002] and others showing some effect of handednesson the suppression of TEOAE in males [Khalfa et al.,1998]. In the current study, information regarding hand-edness was not acquired; therefore, one cannot rule outthe possibility that TEOAE amplitude and suppressionmay have been influenced by dexterity. Thus, futurestudies could benefit from adding a dexterity measure to

their protocol.Auditory efferent abnormality as manifested by MEARand/or MOCB dysfunction was evident in 71% of chil-dren with SM. Despite the use of more stringent criteriain the present study, this finding is similar to that report-ed for our previous smaller cohort suggesting that ap-proximately 75% of children with SM exhibit abnormali-ties in auditory efferent activity; 19% of the children withSM exhibited abnormal findings in both MEAR andMOCB, whereas none of the healthy children showedsuch a result; 52% of children with SM, however, exhib-ited either MEAR or MOCB dysfunction. This distribu-

tion of results may be explained by the notion of func-tional complementarity of the MEAR and MOCB efferentreflexes. Specifically, MEAR activation is largest for low-frequency, high-intensity sounds, whereas MOCB activa-tion is largest in response to mid-to-high-frequency, low-intensity sounds [Liberman and Guinan, 1998]. Lastly,the finding that some children with SM did not demon-strate aberration in MEAR or MOCB function cannotrule out a possible impairment in higher descending ef-

ferent control pathways [Liberman and Guinan, 1998]and thus requires further investigation.

Based on the current findings we conclude that insome children with SM, MEAR and MOCB dysfunctionmay be associated with an auditory processing deficit.Consequently, a child with SM may have difficulty in si-

multaneously coping with incoming sounds and self-vo-calization and thus faces the dilemma of consciously orsubconsciously choosing between speaking and listening.The combination of social anxiety, typical of childrenwith SM, and listening difficulties due to an auditory pro-cessing deficit, may lead the child to resolve this dilemmaby avoiding vocalization. The assumption that some chil-dren with SM may suffer from an auditory processing def-icit is supported by our previous findings which showedthat 9 of 18 children with SM, who had abnormal audi-tory efferent activity, demonstrated impaired auditoryprocessing ability during a vocalization task, compared to

9 children with SM who had normal auditory efferent ac-tivity [Arie et al., 2007]. Nonetheless, further investiga-tion is required to substantiate this particular notion.

In conclusion, the current data solidifies and extendsour previous findings showing that a large proportion ofchildren with SM exhibit aberrant efferent auditory func-tion that may be involved in speech avoidance in selectedsituations. The suggested aberrant audio-vocal interac-tion in SM supports the premise that the efferent reflexes,MEAR and MOCB, may have an important role in theprocess of self-vocalization.

Disclosure Statement

All authors declare that they have no conflict of interest orfunding arrangement related to the current research.

References American Psychiatric Association: Diagnosticand Statistical Manual of Mental Disorders,ed 4, text revision (DSM-IV-TR). Washing-ton, American Psychiatric Association, 2000.

Arie M, Henkin Y, Lamy D, Tetin-Schneider S,

Apter A, Sadeh A, Bar-Haim Y: Reduced au-ditory processing capacity during vocaliza-tion in children with selective mutism. BiolPsychiatry 2007; 61: 419421.

Bar-Haim Y: Introversion and individual differ-ences in acoustic reflex function. Int J Psycho-physiol 2002; 46: 111.

Bar-Haim Y, Henkin Y, Ari-Even-Roth D, Tatin-Schneider S, Hildesheimer M, Muchnik C:Reduced auditory efferent activity in child-hood selective mutism. Biol Psychiatry 2004;55: 10611068.


9/9


Selective Mutism


DOI: 10.1159/000354160

361

Berlin CI, Hood IJ, Wen H, Szaho P, Cecola RP,Righy P, Jackson DF: Contralateral suppres-sion of non-linear click-evoked otoacousticemissions. Hear Res 1993; 7: 111.

Black B, Uhde TW: Elective mutism as a variantof social phobia. J Am Acad Child AdolescPsychiatr 1992; 31: 10901094.

Boon F: The selective mutism controversy (con-tinued). J Am Acad Child Adolesc Psychiatry1994; 33: 283.

Borg E, Bergkvist C, Gustafsson D: Self-masking:Listening during vocalization. Normal hear-ing. J Acoust Soc Am 2009; 125: 38713881.

Borg E, Counter SA: The middle-ear muscles. SciAm 1989; 260: 7480.

Borg E, Zakrisson JE: The activity of the stapediusmuscle in man during vocalization. Acta Oto-laryngol 1975; 79: 325333.

Carmel PW, Starr A: Acoustic and nonacoustic fac-tors modifying middle ear muscle activity inwaking cats. J Neurophysiol 1963; 26: 598616.

Collet L, Kemp DT, Veuillet E, Duclaux R, MoulinA, Morgan A: Effect of contralateral auditorystimuli on active cochlear micro-mechanical

properties in human subjects. Hear Res 1990;43: 251261.

Curio G, Neuloh G, Numminen J, Jousmaki V,Hari R: Speaking modifies voice-evoked ac-tivity in the human auditory cortex. HumBrain Mapp 2000; 9: 183191.

Driscoll C, Kei J, McPherson B: Handedness ef-fects on transient evoked otoacoustic emis-sion in schoolchildren. J Am Acad Audiol2002; 13: 403406.

Eliades SJ, Wang X: Sensory-motor interaction inthe primate auditory cortex during self-initi-ated vocalization. J Neurophysiol 2003; 89:21942207.

Gelfand SA: Essentials in Audiology, ed 3. NewYork, Thieme, 2009.

Giraud AL, Garnier S, Micheyl C, Lina G, ChaysA, Chery-Croze S: Auditory efferent involvedin speech-in-noise intelligibility. Neuroreport1997; 8: 17791783.

Goldberg RL, Henson OW Jr: Changes in cochle-ar mechanics during vocalization: evidencefor a phasic medial efferent effect. Hear Res1998; 122: 7181.

Gothelf D, Ferber N, Raveh E, Apter A, Attias J:Hyperacusis in Williams syndrome: charac-teristics and associated neuroaudiologic ab-normalities. Neurology 2006; 66: 390395.

Guinan JJ Jr: Olivocochlear efferents: anatomy,physiology, function, and the measurement ofefferent effects in humans. Ear Hear 2006; 27:589607.

Hage S, Jurgens U, Ehret G: Audio-vocal interac-tion in the pontine brainstem during self-ini-tiated vocalization in squirrel monkey. Eur JNeurosci 2006; 23: 32973308.

Henkin Y, Feinholz M, Arie M, Bar-Haim Y: P50suppression in children with selective mut-ism: a preliminary report. J Abnorm ChildPsychol 2010; 38: 4348.

Henson OW: The activity and function of themiddle ear muscle in echo-locating bats. JPhysiol 1965; 180: 871887.

Hood LJ: An overview of neural function andfeedback control in human communication. JCommun Disord 1998; 31: 461470.

Houde JF, Nagarajan SS, Sekihara K, MerzenichMM: Modulation of the auditory cortex dur-ing speech: an MEG study. J Cog Neurosci2002; 14: 11251138.

Kaufman J, Birmaher B, Brent D, Rao U: Schedulefor Affective Disorders and Schizophrenia for

School-Age Children Present and LifetimeVersion (K-SADS-PL): initial reliability andvalidity data. J Am Acad Child Adolesc Psy-chiatry 1997; 36: 980988.

Kawase T, Delgutte B, Liberman MC: Anti-mask-ing effects of the olivocochlear reflex. II. En-hancement of auditory nerve response tomasked tones. J Neurophysiol 1993; 70: 25332549.

Khalfa S, Bruneau N, Roge B,Georgieff N, VeuilletE, Adrien JC, Barthelemy C, Collet L: Periph-eral auditory asymmetry in infantile autism.Eur J Neurosci 2001; 13: 628632.

Khalfa S, Veuillet E, Collet L: Influence of hand-edness on peripheral auditory asymmetry.Eur J Neurosci 1998; 10: 27312737.

Klockhoff I: Middle ear muscle reflexes in man. Aclinical and experimental study with specialreference to diagnostic problems in hearingimpairment. Acta Otolaryngol Suppl 1961;164: 192.

Kumar UA, Vanaja CS: Functioning of olivoco-chlear bundle and speech perception in noise.Ear Hear 2004; 25: 142146.

Lamprecht A: Changes of the voice in the Lom-bard reflex. Laryngol Rhinol Otol 1988; 67:350354.

Liberman MC, Guinan JJ Jr: Feedback control ofthe auditory periphery: anti-masking effectsof middle ear muscles vs. olivocochlear effer-ents. J Commun Disord 1998; 31: 471482.

Metzner W: A possible neuronal basis for Dop-pler-shift compensation in echo-locatinghorseshoe bats. Nature 1989; 341: 529532.

Metzner W: An audio-vocal interface in echo-lo-cating horseshoe bats. J Neurosci 1993; 13:18991915.

Micheyl C, Collet L: Involvement of olivocochlearbundle in the detection of tones in noise. JAcoust Soc Am 1996; 99: 16041610.

Muchnik C, Ari-Even Roth D, Othman-Jebara R,Putter-Katz H, Shabtai E, Hildesheimer M:Reduced medial olivocochlear bundle systemfunction in children with auditory process-ing disorders. Audiol Neurotol 2004; 9: 107114.

Poulet J, Hedwig B: A corollary discharge main-tains auditory sensitivity during sound pro-duction. Nature 2002; 418: 872876.

Prasher D, Ryan S, Luxon L: Contralateral sup-pression of transiently evoked otoacousticemissions and neuro-otology. Br J Audiol1994; 28: 247254.

Robertson D: Centrifugal control in mammalianhearing. Clin Exp Pharmacol Physiol 2009; 36:

603611.Sanches SG, Carvallo RM: Contralateral suppres-

sion of transient evoked otoacoustic emis-sions in children with auditory processingdisorder. Audiol Neurotol 2006; 11: 366372.

Sharkey L, McNicholas F: More than 100 years ofsilence, elective mutism: a review of the lit-erature. Eur Child Adolesc Psychiatry 2008;17: 255263.

Ventura MI, Nagarajan SS, Houde JF: Speech tar-get modulates speaking-induced suppressionin auditory cortex. BMC Neurosci 2009; 10:5868.

Veuillet E, Collet L: Objective evidence of periph-eral auditory disorders in learning-impairedchildren. J Audiol Med 1999; 8: 1829.

Veuillet E, Magnan A, Ecalle J, Thai-Van H, Col-let L: Auditory processing disorder in chil-dren with reading disabilities: effect of audio-visual training. Brain 2007; 130: 29152928.

Viana AG, Beidel DC, Rabian B: Selective mut-ism: a review and integration of the last 15years. Clin Psychol Rev 2009; 29: 5767.

Yalcinkaya F, Yilmaz ST, Muluk NB: Transientevoked otoacoustic emissions and contralat-eral suppression in children with auditory lis-tening problems. Auris Nasus Larynx 2010;37: 4754.

abnoramlities in efferent activities mutism

Documents