The Magnitude of Contralateral Suppression of Otoacoustic Emissions Is Ear- and Age-Dependent

The maturation of the uncrossed medial olivocochlear (UMOC) efferent remains poorly documented to date. The UMOC efferent system allows listeners to not only detect but also to process, recognize, and discriminate auditory stimuli. Its fibers can be explored non-invasively by recording the effect of contralateral acoustic stimulation (CAS), resulting in a decrease in the amplitude of transient evoked otoacoustic emissions (TEOAE). The objective of the present cross-sectional study was to investigate how the effectiveness of this system varies with age in healthy subjects aged 8 years to adulthood. For this purpose, 120 right-handed native French-speaking subjects (57 females and 63 males) were divided into five age groups of 24 subjects each: 8y–10y, 10y–11y6m, 11y6m–13y, 13y–17y, and ≥18y. TEOAE amplitudes with and without CAS were recorded. The equivalent attenuation (EA) was calculated, corresponding to the change in TEOAE amplitude equivalent to the effect generated by CAS. General linear models were performed to control for the effect of ear, sex, and age on EA. No sex effect was found. A stronger EA was consistently found regardless of age group in the right ear compared to the left. In contrast to the right ear, for which, on average, EA remained constant across age groups, an increasingly weaker TEOAE suppression effect with age was found in the left ear, reinforcing the asymmetrical functioning of the UMOC efferent system in favor of the right ear in adulthood. Further studies are needed to investigate the lateralization of the UMOC efferent system and its changes over time in cases of atypical or reversed cortical asymmetries, especially in subjects with specific learning disorders.


Introduction
The central auditory nervous system (CANS) consists of ascending and descending auditory pathways that are closely interdependent and act in an integrated manner. From the peripheral auditory receptor, afferent fibers carry information to the auditory cortex (AC) through the brainstem via, successively, the cochlear nuclei, the superior olivary complex (SOC), the lateral lemniscus, the inferior colliculus, and the medial geniculate body [1]. In turn, the AC is the source of multiple efferent projections [2,3] to each of the auditory nuclei [4]. A contingent of efferent fibers originates from the medial SOC and forms the "Uncrossed Medial OlivoCochlear (UMOC)" efferent system that projects to the ipsilateral cochlea reaching directly to the outer hair cells (OHCs) [5] (for a recent review, see [6]). In doing so, the SOC serves as a relay station for the corticofugal pathway that originates from pyramidal neurons in layers V/VI of the primary AC [7]. Therefore, the AC is able to control and modulate OHCs activity in a top-down manner [8].
The CANS undergoes maturational changes that allow children not only to detect but also to process, recognize, and discriminate auditory stimuli. Auditory evoked potentials (AEPs) have been extensively used to study the developmental time course of the auditory afferent pathway. It is well known that AEP peaks shorten and become sharper with age, suggesting an increase in neural efficiency [9][10][11][12]. However, having long assumed that the latency of early AEPs reaches adult values within the first two years of life [13], more recent research suggests a longer developmental time course, particularly for the auditory brainstem, which continues to mature during childhood [14] and even adolescence [15], leading to the notion of continuous maturational plasticity. Similarly, late AEPs have shown that changes in AC extend into late adolescence [16]. Maturation of the auditory thalamocortical connections occurs between two and five years of age, with the final stage of structural maturation of AC beginning later (at the earliest between 6 and 12 years) (see [17] for a detailed review). Such maturational changes underpin the development of increasingly complex auditory processing, allowing for a strong improvement in speech understanding, especially in noise, in late childhood and adolescence [18]. This is probably made possible not only by stronger cortical interaction within the same hemisphere but also by increasing communication between the two hemispheres. As cortical regions continue to mature in adolescence, functional changes may still occur in early adulthood [19].
In contrast, the maturation of the auditory corticofugal pathway remains much less documented. This is even more surprising as the UMOC efferent system can be explored non-invasively in clinical routine. One of the major advances in modern audiology is credited to Kemp [20,21], who first recorded, in the auditory canal, sounds produced by the contraction of OHCs in response to brief acoustic clicks and identified them as transient evoked otoacoustic emissions (TEOAEs). Due to the UMOC fibers being inhibitory, the addition of a contralateral acoustic stimulation (CAS) leads to a TEOAE amplitude decrease [22,23] called the "contralateral suppressive effect" (CSE). While recent studies in animals [24][25][26][27] provide strong anatomical and functional evidence for the existence of central control of OHCs by multiple corticofugal loops, in humans, such demonstrations have long remained indirect. For instance, in right-handers, the CSE has been reported to be stronger in the right ear, reflecting a rightward asymmetry in UMOC efferent system function [28,29], this asymmetry being absent in right-handed schizophrenic patients [30]. Surgical resection of the primary AC has been shown to suppress the CSE [31], whereas intracortical electrical stimulation of Heschl's gyrus increases it [32]. A recent functional MRI study in healthy adult subjects has confirmed the existence of direct connections between auditory structures, including Heschl's gyrus, the planum temporale, and the SOC [33].
From a developmental perspective, the CSE is present at birth [34] but is reduced in preterm neonates [35] with a greater suppressive effect on the right ear [36,37], which may reflect an early lateralization process. It has been reported that the magnitude of CSE slowly decreases during infancy until the age of 3 years [38]. A recent study found no effect of age on the amplitude of CSE in preschoolers aged 3-6 years [39]. Several studies have been conducted on older children, but to our knowledge, never on children over 14 years of age, and often on groups covering broad age ranges up to 4 years [40][41][42][43]. Data in the literature have reported differences in CSE amplitude between the right and left ear in children aged 9-10 years [44], 5-10 years [45], or 7-12 years [46]. Despite this, the fine-grained age-related changes in UMOC efferent system function during a highly plastic but vulnerable developmental period from childhood to early adulthood are still poorly understood.
As there are only a few studies on the effect of age on UMOC, and they are not conclusive, the aim of the present cross-sectional study was to examine how CSE varies with age in healthy subjects aged 8 to 35. Given that the cortical auditory areas are not fully mature until late adolescence and similar to the gradual maturation of the auditory afferent pathway, we expected to observe differences in CSE amplitude with age, but also depending on the test ear.

Participants
One hundred and twenty native French-speaking right-handed subjects, with a balanced number of females (n = 57) and males (n = 63) and a mean age of 13 ± 0.3 years (range , participated in the study. The experiment was carried out with the agreement of the Regional Ethics Committee (CPP Sud-Est IV Lyon, approval No. 09/086 for adults and 04/008 for children). Informed consent was obtained from all participants (all children with verbal assent). Handedness degree was assessed using the Edinburgh Handedness Inventory [47]. None of the participants had a history of communication, cognitive, neurological, attention, or autism spectrum disorders. The participating children performed average or above average academically, never had to repeat a grade, and had no specific learning disorders. None of them had undergone speech therapy, and all of them had a normal reading age, as measured by the "L'alouette" test [48]. Participants were divided into five age groups of 24 subjects each, as follows: 8y-10y (group 1), 10y-11y6m (group 2), 11y6m-13y (group 3), 13y-17 (group 4), and ≥18y (group 5).
The demographic characteristics by age group and sex and the laterality by age group are shown in Table 1. All participants had normal otoscopy, air conduction hearing thresholds less than 15 dB HL at octave intervals from 0.25 to 8 kHz, and normal type A tympanograms (static acoustic admittance between 0.35 and 1.75 mmho and peak pressure between +50 to −100 daPa) in both ears. They also had ipsi-and contralateral acoustic reflex thresholds (ARTs) greater than 75 dB SPL in response to white noise bilaterally. TEOAEs were bilaterally present for all participants with an overall response amplitude of emissions exceeding the noise amplitude; that is, a signal-to-noise ratio (SNR) greater than 4 dB when measured with a non-linear click method at an intrameatal intensity of 81.5 ± 0.05 peak-equivalent (pe) SPL. The whole reproducibility level was greater than 50%, with stimulus stability exceeding 90%.

Apparatus
All measurements were performed in a soundproof booth. TEOAEs were recorded using the ILO92 system (Otodynamics Ltd., London, UK). The Otodynamics 5.6 clinical OAE software was used, and responses were recorded by a standard transducer (Otodynamics Model UGS TEOAE Probe). The probe, comprising a transmitter and a microphone, was fitted to the external auditory meatus with a foam rubber tip. CAS was delivered with a headphone (TDH49) using an AC40 clinical audiometer (Interacoustics, Middlefart, Denmark).

Procedure
The participants were instructed to sit quietly in a comfortable position and to keep silent. To avoid falling asleep or to keep them calm, a silent movie was played during this passive CSE measurement (cartoon for children and film for adults). Measurements were successively conducted on the right ear and left ear in random order.
TEOAEs were recorded in a linear mode using 80-µs clicks at a rate of 50/s, successively in the absence and presence of a 35 dBSL continuous noise (0.125-12 kHz with a flat spectrum) in the contralateral ear (minimal threshold = 0 dBHL-maximal threshold = 15 dBHL). The stimulus level was randomly adjusted to obtain 5 peak pressures between 60 and 72 peSPL (±1 dB) that were presented in 3 dB steps. CAS was manually switched on prior to the onset of TEOAE recordings. For each of the 5 ipsilateral stimulation levels, the first measurement was always carried out without CAS. The noise rejection level was set at 47.3 dB peSPL. Three hundred responses were averaged in a 3.2 to 20 ms-time window. Signal parameters were analyzed as global values (for the whole signal). This protocol provides an input/output function for the global TEOAE amplitude with and without CAS and calculates the equivalent attenuation (EA) according to the so-called Lyon procedure (see for more details [44,46,49]). As such, EA corresponds to the change (in dB) of the ipsilateral click-evoked OAEs equivalent to the effect generated by CAS. The more negative the EA, the stronger the TEOAE suppressive effect. By subtracting the EA measured on each ear (EA right -EA left ), an asymmetry index (AI) can be obtained: the more negative it is, the stronger the CSE on the right ear.

Data Analysis
Two general linear models (GLMs), using JAMOVI version 1.6 (https://www.jamovi. org/, accessed on 20 June 2023), were run on three dependent variables (EA right , EA left , and AI) fitting the UMOC efferent system function. The first model used only one factor at a time: five age groups or sex. To ensure that the effect of age did not depend on our choice of age groups, we also investigated the relationships between age, as a continuous variable, and our three variables of interest (EA right , EA left , and AI) using the Spearman correlation test. A second model accounted for two factors: five age groups and sex with an ANOVA design. To account for multiple comparisons in the GLM analysis, all p-values were thresholded at a Bonferroni-corrected value of p = 0.5. Violin plots were drawn for data description, including a marker for the median value and a box for the interquartile range.
Finally, a two-way age group X sex factor test was performed to determine the influence of these two variables on CSE. Figure 4 shows the CSE distribution by age group and sex. A significant age group effect was found for EA left and AI (F(4, 119) = 9.85, p < 0.001, η 2 = 0.26) from 10 years to 13 years (t(114) = 2.05, p = 0.04) and from 8 years to adulthood (t(114) = 3.32, p = 0.001). In contrast, there was no significant age group effect for EA right (F(4, 119) = 1.64 p = 0.17, η 2 = 0.06).
No significant sex effect was found for EA left (F(1, 119) = 0.27, p = 0.60, η 2 = 0.14) and EA right (F(1, 119) Table 3 for male and female patients (all ages combined). Average EA and AI values for each ear are provided in Table 4, for male and female patients and each age group.  Finally, a two-way age group X sex factor test was performed to determine the influence of these two variables on CSE. Figure 4 shows the CSE distribution by age group and sex. A significant age group effect was found for EAleft and AI (F(4, 119) = 9.85, p < 0.001, η 2 = 0.26) from 10 years to 13 years (t(114) = 2.05, p = 0.04) and from 8 years to adulthood (t(114) = 3.32, p = 0.001). In contrast, there was no significant age group effect for EAright Average EA and AI values for each ear are provided in Table 4, for male and female patients and each age group.

Discussion
In this study of typically developing right-handed subjects, we identified ear-dependent differences in the evolution of UMOC function between childhood and adulthood. As maturation of the auditory afferent pathway is gradual, we expected to observe differences in CSE amplitude with age, but also as a function of the ear. Unlike the right

Discussion
In this study of typically developing right-handed subjects, we identified ear-dependent differences in the evolution of UMOC function between childhood and adulthood. As maturation of the auditory afferent pathway is gradual, we expected to observe differences in CSE amplitude with age, but also as a function of the ear. Unlike the right ear, for which, on average, EA remained constant across age groups, an increasingly weaker TEOAE suppression effect with age was found in the left ear. These results confirm those of previous studies observing that in adult right-handers, CSE is significantly stronger in the right ear [49]. This RE advantage in TEOAE suppression reinforces the concept of laterality of the UMOC function, which favors the action of the right olivocochlear tract over the left [50].
Here, no significant age group effect was found for EA on the right ear and AI. On the other hand, a significant age group effect was found for EA from age 8 to 13 and over the entire developmental trajectory from age 8 to adulthood for the left ear. Previous studies on groups with a wider age range revealed no significant asymmetry in a group of children aged 6 to 13 [51], while a mean laterality score significantly lower than 0, indicating rightear predominance, was found in typically developing children aged 8 to 14 [42]. Notably, the present study is the only one to date to compare smaller age groups over a longer period (from childhood to adulthood), with a large and identical number of participants (N = 24) in each group. It underlines the fact that asymmetry of UMOC function may continue to develop well into childhood and adolescence.
Our results indicate that changes in the function of the UMOC efferent system with age are underpinned by a reduction in the CSE on the left ear, with stability on the right. This raises the question of what factors might explain this reduction in contralateral suppression, limited to the left ear. Firstly, one factor that may play a role is the presence or absence of TEOAEs of comparable amplitude between the two ears. Given that the CSE is based on the measurement of a change in TEOAE amplitude under the effect of acoustic stimulation, the functional integrity of the OHCs is paramount, and the presence of TEOAEs reflects this [21]. This was the case for all the participants in the present study since they all presented a global TEOAE response value of at least 4 dBpeSPL with more than 50% reproducibility in both ears. Strict selection criteria for tympanometry and audiometry were also used to exclude any potential peripheral effects on TEOAE amplitude. We observed a significant decrease in TEOAE amplitude with age, confirming previous studies [52,53]. This decrease was identical in both ears, with no difference in TEOAE amplitude between ears in each age group. Thus, if the TEOAE amplitude factor alone explained the decrease in contralateral suppression observed in the left ear, the latter should also have been observed in the right ear. Secondly, there is evidence that handedness may play a role in TEOAE amplitude and suppression [28]. However, this is unlikely to have been the case here since all participants were right-handed.
Rather, it may be asked whether the relative constancy of CSE as age increases can be explained by the completion of maturation processes. In a post-mortem histological and immunohistochemical study, Moore and Guan traced cortical maturation through the fetal period, infancy, childhood, and young adulthood [54]. Maturation of layers 4, 5, and 6 and the medial geniculate to the cortex was found to be complete by age 6. Later in childhood, maturation would take place in superficial layers IIIb, IIIa, and II of the cortex. By the age of 12, the density of mature axons has become equivalent to that of an adult [54]. The longer duration of axonal maturation could have implications for the emergence of auditory cortical functions [17,54]. Other central auditory processes undergo maturation during adolescence that continues into adulthood, including those associated with the predominance of right-ear listening in right-handed individuals [55,56]. It is possible that the asymmetrical changes in CSE observed here with age are linked to the maturation of more central processes, reinforcing the hypothesis of the existence of central control of OHCs function.
Finally, one may wonder why the UMOC efferent system function in the left ear does not reach stability as rapidly as in the right. This could be because the subcortical or cortical structures that control it develop later. However, descending loops are anatomically complex, and their function is still poorly understood. Our results suggest that, in righthanded children, UMOC fibers are involved in different maturation processes between the right and left ears. In the left ear (receiving uncrossed efferent fibers from the left superior olivary complex explored by a CAS passing through the right cochlear nucleus), this system would continue to mature by "calming down". In the right ear (receiving uncrossed efferent fibers from the right superior olivary complex explored by a CAS passing through the left cochlear nucleus), the system could mature earlier as a reflection of the predominance of the left auditory cortex in right-handed subjects.
Here, we observed no statistically significant gender differences in TEOAE suppression and AI. Gender disparities in TEOAEs exist in both neonates and adults. Mac Faden et al. suggested that the degree of androgen exposure during prenatal development may be a contributing factor to these differences [57]. While women are known to have stronger click-induced OAEs than men [57], a study investigating contralateral TEOAE suppression as a function of gender showed that men had greater contralateral suppression, the exact reason for this difference being unclear [58]. Other studies have found no sex differences in the contralateral suppression of OAEs [59,60]. Among them, Jedrzejczak et al. measured TEOAE with and without CAS in 126 normal-hearing children aged 3-6 years. They found no significant effect of sex on TEOAE suppression [39,61].
Some questions remain: what is the exact starting point of this asymmetrical development, which in this study only manifests from the age of 11y6m? Previous research has shown protracted subcortical auditory maturation during adolescence, which may represent a "transitional point" between the enhanced response in childhood and the mature, albeit smaller, response in adulthood [15].
It is still possible that the samples studied here were too small; moreover, we observed a large inter-individual variability in adults but also in children in terms of CSE. In terms of learning acquisitions, all participants were described as having a "normal" reading age. Yet, each age group in the study covered a wide range of reading ages, which may have increased within-group variability.
Further studies are needed to study the maturation of UMOC function when cortical asymmetries are reversed or absent, as is the case in dyslexic subjects. If feasible, longitudinal studies would neutralize the inter-individual variability observed in the variables fitting the UMOC efferent system function. Funding: This study was funded by "Direction Générale de l'Offre des soins-Programme de Recherche: PHRC Interrégional" (ASTRAUCOLE Project ID: RCB 2013-A00732-43). The funding was acquired and received by Hung Thai-Van. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Regional Ethics Committee (CPP Sud-Est IV, approval No. 09/086 for adults and 04/008 for children).

Informed Consent Statement:
Informed consent was obtained from all participants or parents/legal guardians involved in the study. Written informed consent was also obtained prior to publishing this paper.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to ethical, legal, and privacy issues.