Efferent unmasking of speech-in-noise encoding?

Abstract

Objective: The medial olivocochlear (MOC) reflex provides efferent feedback from the brainstem to cochlear outer hair cells. Physiologic studies have demonstrated that the MOC reflex is involved in "unmasking" of signals-in-noise at the level of the auditory nerve; however, its functional importance in human hearing remains unclear.

Design: This study examined relationships between pre-neural measurements of MOC reflex strength (click-evoked otoacoustic emission inhibition; CEOAE) and neural measurements of speech-in-noise encoding (speech frequency following response; sFFR) in four conditions (Quiet, Contralateral Noise, Ipsilateral Noise, and Ipsilateral + Contralateral Noise). Three measures of CEOAE inhibition (amplitude reduction, effective attenuation, and input-output slope inhibition) were used to quantify pre-neural MOC reflex strength. Correlations between pre-neural MOC reflex strength and sFFR "unmasking" (i.e. response recovery from masking effects with activation of the MOC reflex in time and frequency domains) were assessed.

Study sample: 18 young adults with normal hearing.

Results: sFFR unmasking effects were insignificant, and there were no correlations between pre-neural MOC reflex strength and sFFR unmasking in the time or frequency domain.

Conclusion: Our results do not support the hypothesis that the MOC reflex is involved in speech-in-noise neural encoding, at least for features that are represented in the sFFR at the SNR tested.

Keywords: Efferent; frequency following response; medial olivocochlear reflex; otoacoustic emissions.

Figure 1.

CEOAE amplitudes evoked by 60 and 70 dB ppeSPL clicks without (black dots) and with CN (grey dots) for one subject. The three methods for quantifying MOC reflex inhibition are shown: Amplitude reduction is indicated by the downward-pointing arrow; Effective attenuation at 65 dB is indicated by the horizontal arrow; I/O inhibition is the difference between the linear regression slopes for Q and CN conditions (I/O = CNslope – Qslope). Note that each condition was replicated three times. While most measurements are highly replicable, one CEOAE + CN measurement at 70 dB was larger than the other two in this subject, hence the appearance of two grey dots.

Figure 2.

CEOAEs were inhibited by CN at both 60 and 70 dB ppeSPL presentation levels (left). There was a marginally significant interaction between stimulus level and condition demonstrating that inhibition was larger for 60 than 70 dB ppeSPL clicks. Inhibition did not significantly differ between sexes (right), although CEOAEs were generally larger in females than males ( * = p < 0.025).

Figure 3.

A) Average sFFR waveforms for CN (red), IN (green), and IN+CN (blue) conditions are overlaid onto Q (black) waveforms for comparison. B) Envelope spectra for each condition overlaid onto Q. C) Average latencies of onset, transition, steady-state, and offset components are plotted as a function of delay re: Q. D) F0 amplitudes by condition.

Figure 4.

CEOAE inhibition was not correlated with sFFR latency unmasking for transition (top) or F0 for steady-state (bottom). Unmasking was quantified as the difference between IN and IN+CN conditions. Pearson’s r and p-values for each correlation are presented in the top right corner of each frame.