Hearing impairment is a common chronic health condition of older age that has been linked to adverse changes in social and emotional well-being. Steeply sloping, high-frequency hearing loss (HFHL) is the most common sensorineural hearing loss profile for middle-aged and older adults with a past history of noise exposure. A common complaint of subjects with HFHL relates to a difficulty understanding speech in fluctuant background noise. Individual thresholds for recognizing speech in noise vary widely, are not well predicted from the audiogram, and are not reliably improved by amplification. The underlying motivation for this project is to identify physiological and perceptual biomarkers that more accurately predict speech in noise recognition in subject with HFHL, as compared to age-matched subjects with normal hearing (NH). Our underlying hypothesis is that impaired speech in noise processing for subjects with HFHL can be predicted from abnormal neural coding of low-frequency signals, where thresholds are normal.
In Aim 1, we employ a series of physiological and psychophysical tests to identify the stage of neural processing (from auditory nerve to cortex) and mode of neural processing (from the auditory nerve compound action potential to subcortical encoding of stimulus fine structure) that most directly map onto speech in noise outcomes in HFHL and NH subjects. To further probe the linkage between neural processing of low-frequency signals and speech in noise recognition, we will employ a new approach to enhance speech in noise processing through an immersive, closed-loop audiomotor software training interface. Our preliminary data suggest that speech in noise recognition can be significantly improved in subjects with sensorineural hearing loss that were randomly assigned to closed-loop audiomotor training, as compared to subjects assigned to a placebo auditory training interface. However, it is not known which physiological and perceptual predictors of speech processing are also modified to support a change in speech recognition thresholds.
Aim 2 will address this point through a randomized, double-blind placebo-controlled study design that will compare the neural and physiological predictors of speech processing before training, after training and at a follow-up test after training has been discontinued. By identifying the biomarkers of neural processing that not only predict speech outcomes in a baseline condition, but also track dynamic shifts in speech processing over the course of an intervention, these studies may identify the most robust neural predictors of speech in noise processing as well as possible targets for future therapies.
Understanding speech in background noise is vital for communication but is among the first aspects of hearing to fail with advanced age or excess environmental noise exposure. This project will identify neural and perceptual biomarkers that most accurately predict difficulties understanding speech in noise in older individuals with normal hearing and high-frequency hearing loss. In a longitudinal study, we will train subjects with immersive action `audio game' software and determine how temporary boosts in speech intelligibility are supported by shifts in underlying neural and perceptual sound processing.
|Wu, P Z; Liberman, L D; Bennett, K et al. (2018) Primary Neural Degeneration in the Human Cochlea: Evidence for Hidden Hearing Loss in the Aging Ear. Neuroscience :|
|Whitton, Jonathon P; Hancock, Kenneth E; Shannon, Jeffrey M et al. (2017) Audiomotor Perceptual Training Enhances Speech Intelligibility in Background Noise. Curr Biol 27:3237-3247.e6|