Binaural speech processing in individuals with auditory neuropathy

Abstract

Auditory neuropathy disrupts the neural representation of sound and may therefore impair processes contingent upon inter-aural integration. The aims of this study were to investigate binaural auditory processing in individuals with axonal (Friedreich ataxia) and demyelinating (Charcot–Marie–Tooth disease type 1A) auditory neuropathy and to evaluate the relationship between the degree of auditory deficit and overall clinical severity in patients with neuropathic disorders.

Twenty-three subjects with genetically confirmed Friedreich ataxia and 12 subjects with Charcot–Marie–Tooth disease type 1A underwent psychophysical evaluation of basic auditory processing (intensity discrimination/temporal resolution) and binaural speech perception assessment using the Listening in Spatialized Noise test. Age, gender and hearing-level-matched controls were also tested.

Speech perception in noise for individuals with auditory neuropathy was abnormal for each listening condition, but was particularly affected in circumstances where binaural processing might have improved perception through spatial segregation. Ability to use spatial cues was correlated with temporal resolution suggesting that the binaural-processing deficit was the result of disordered representation of timing cues in the left and right auditory nerves. Spatial processing was also related to overall disease severity (as measured by the Friedreich Ataxia Rating Scale and Charcot–Marie–Tooth Neuropathy Score) suggesting that the degree of neural dysfunction in the auditory system accurately reflects generalized neuropathic changes.

Measures of binaural speech processing show promise for application in the neurology clinic. In individuals with auditory neuropathy due to both axonal and demyelinating mechanisms the assessment provides a measure of functional hearing ability, a biomarker capable of tracking the natural history of progressive disease and a potential means of evaluating the effectiveness of interventions.

Introduction

The auditory system is exquisitely sensitive. Accurate representation of complex and continually varying acoustic signals (such as speech), requires a high degree of precision. Subtle changes affecting any point in the peripheral or central auditory mechanisms can have a significant impact on perception. The consequences of cochlear damage on sound detection and the processing of (particularly) frequency information have been explored in detail over the past 40 years (Moore, 1995). More recently, the effects of auditory neuropathies on the transmission of electrical signals through the auditory neural pathways (and the consequent effects on functional hearing) have also been considered (Starr et al., 1991, Starr et al., 1996, Rance et al., 2004, Zeng et al., 2005).

The term “auditory neuropathy” (AN) describes a group of hearing disorders in which cochlear (outer hair cell) function is normal, but afferent conduction through the auditory nerve and central pathways is disrupted (Starr et al., 1996). This disruption may be the result of reduced neuronal population (axonopathy) or impaired neural function due to demyelination or abnormal synaptic transmission (Starr et al., 1996). Auditory neuropathy may present in infancy, when it is typically the result of perinatal insult such as oxygen deprivation or hyperbilirubinaemia (Rance et al., 1999), or at a later stage when associated with generalized neurodegenerative processes (Starr et al., 1996, Starr et al., 2000, Rapin and Gravel, 2003).

Friedreich ataxia (FRDA) is one such neurodegenerative disease which affects both motor and sensory systems through mutations in the FXN gene (Campuzano et al., 1996). Physical manifestations of the disease include limb and trunk ataxia, scoliosis, hypertrophic cardiomyopathy and an increased rate of diabetes mellitus. Sensory deficits have been observed across multiple modalities including the visual, vestibular systems and auditory systems (Fahey et al., 2008, Rance et al., 2008, Fielding et al., 2010). Auditory neuropathy occurs in a high proportion of cases (>90% in the latter disease stages [Rance et al., 2008, Rance et al., 2010a]) as a result of axonopathy in the cochlear nerve (Spoendlin, 1974).

Charcot–Marie–Tooth (CMT) disease is another neurodegenerative syndrome with auditory neuropathy as a phenotypic variant (Satya-Murti et al., 1979, Starr et al., 1996, Kovach et al., 2002). The disease presents with a range of symptoms reflecting peripheral nerve dysfunction including weakness, muscle atrophy, orthopaedic abnormality and sensory deficits. CMT is subdivided into demyelinating forms (dominantly inherited [CMT type 1] and recessively inherited [CMT type 4], axonal forms (generally dominantly inherited [CMT type 2]) and intermediate forms including X-linked varieties. Auditory neuropathy is relatively common in both axonal and demyelinating forms. Prevalence in affected adults is yet to be determined, but recent evidence in school-aged children has suggested that >85% of individuals with CMT2 and >60% with CMT disease type 1A (CMT1A) show abnormal auditory-evoked potentials and impaired perception (Rance et al., 2012b).

A cardinal feature of auditory neuropathy is impaired temporal resolution – the ability to perceive rapid signal changes over time (Starr et al., 1991, Starr et al., 1996, Rance et al., 2004, Zeng et al., 2005). In both demyelinating and axonal forms, this is thought due to neural desynchrony producing a time-smeared representation of the acoustic signal. In the case of demyelinating neuropathy, desynchrony occurs when damage to the neural insulator produces slow/inconsistent conduction and an impaired ability to transmit trains of pulses (Brown and Watson, 2002). In axonal disorders desynchrony may be the result of irregular firing patterns in partially damaged fibres, secondary demyelination or conduction block (Brown and Watson, 2002). Degree of temporal-processing deficit may be equally severe in both AN-forms (Rance et al., 2012b).

The functional consequence of temporal processing disorder is impaired speech perception (particularly in background noise). Monaural (individual ear) tests of speech understanding have shown that most listeners with auditory neuropathy struggle to understand speech even in low levels of noise. This “figure-ground” deficit is thought to be due to a combination of increased masking effects and an inability to use brief quiet periods (in a fluctuating noise) to access the speech signal (Zeng et al., 2005, Rance et al., 2007, Rance et al., 2010b).

Temporal-processing deficit may also affect a listener’s ability to combine signals from the two ears to maximize perception. When sound emanates from any point other than directly in front of the listener (0° azimuth), the signal travels an unequal distance to each ear. This gives rise to subtle (<1 ms) inter-aural time differences (ITDs). Furthermore, the head acts as a barrier when the sound source is not directly in front which results in inter-aural level differences (ILDs) of approximately 5–15 dB. These timing and intensity differences are initially processed in the lower brainstem where neural impulses originating in the left and right auditory nerves meet at the superior olivary complex (SOC) and are converted to location (and movement) information (Reidel and Kollmeier, 2002) before being relayed to higher centres in the auditory pathway (Spitzer and Semple, 1993, Riedel and Kollmeier, 2002, Warren et al., 2002, Kuwada et al., 2006). As a result, listeners with normal hearing may perceive a three-dimensional listening environment and be able to accurately judge the direction of sound sources. This capacity has obvious benefits alerting the listener to possible information sources and environmental dangers. In addition, sound localization cues may also be used to improve perception in background noise when a target signal (such as speech) and competing noise arise from different directions. That is, acoustic sequences from different locations may be parsed into separate “streams” and a target stream may be selectively attended as a separate entity within a complex signal (Micheyl et al., 2007, Cameron and Dillon, 2008).

Auditory neuropathy disrupts the neural representation of sound and may therefore impair processes contingent upon inter-aural comparison. Binaural auditory processing has not been explored in detail for listeners with AN, but preliminary evidence suggests that spatial perception is disrupted. Zeng et al. (2005) measured sound localization ability in three affected subjects using a range of binaural stimuli and found that while localization based upon inter-aural level differences was unimpaired, the AN listeners showed little or no ability to use inter-aural timing cues to judge sound direction.

This study had three main aims: (1) to investigate binaural auditory-processing deficits in individuals with axonal and demyelinating auditory neuropathy; (2) to evaluate the relationship between degree of auditory-processing deficit and overall clinical severity in patients with neurodegenerative disease; (3) to track binaural processing changes over time in affected individuals.