Distortion product otoacoustic emission components as a potential marker of preclinical noise injury to the cochlea

Abstract

Background: Excessive noise exposure can lead to noise induced hearing loss (NIHL) and hearing conservation programmes are essential to prevent occupational hearing loss. These programmes rely on monitoring hearing using pure tone audiometry. However, audiometry is not very sensitive and only detects a permanent hearing loss after significant injury has occurred. Distortion product otoacoustic emissions (DPOAEs) are generated by cochlear non-linear physiological processes associated with the outer hairs cells that underpin detection of sound, thus DPOAEs have promise as a tool to detect and monitor cochlear hair cell damage in NIHL. The DPOAE has two components, one generated at the point of interaction between the primaries (generator), the second reflected from the characteristic place of the DPOAE frequency (reflected). Some of the variability in DPOAE is believed to result from the interaction of the components, termed the DPOAE fine structure, and extracting the components may reduce the variability and improve the sensitivity of the DPOAE to injury. This study investigated the feasibility and reliability of a new approach to separating the DPOAE components, and their relationship to hearing measures in people with and without exposure to noise. Methods: Logarithmically swept primaries with a least-squares-fit (LSF) analysis were used to separate the DPOAE components, and the approach was evaluated and optimised in normally hearing subjects. Two studies followed that investigated the components of the DPOAE and how they correlate with behavioural measures of hearing. Firstly, in subjects (n=16) with ‘normal’ hearing (<20 dB HL), both behavioural measures of the auditory filter (notched-noise maskers) and threshold were compared to the components of the DPOAE. In the second study subjects (108 men) with a mixture of thresholds, noise exposure and age were recruited and then divided into two equal sized and age matched groups (noise exposed and non-noise exposed) based on a detailed history of self-reported noise exposure. The DPOAE components were compared to audiometric threshold and behavioural measures of compressive nonlinearity using Schroeder-phase masked thresholds. Results: Overall, psychoacoustic measures of auditory filter and cochlear non-linearity were influenced by audiometric threshold, and exhibited no relationship with DPOAE level in subjects with ‘normal’ hearing (<20 dB HL). The generator and combined DPOAE showed very similar patterns for all associations and correlations throughout both studies, but the generator component was less variable. The reflected component of the DPOAE was lower in level and was mostly weakly correlated with other measures. Some trends were noted in the DPOAE level that holds some promise for future research. Consistently across the two studies, younger and or normally hearing (<20 dB HL) subjects that reported significant noise exposure exhibited a reduction in DPOAE level at lower frequencies (f2 between 1500-3000 Hz). These differences were most apparent in the generator component for high level primaries (>55 dB SPL). Furthermore the relationship between audiometric threshold and DPOAE level showed a different pattern between those with and without noise exposure. Audiometric thresholds for the noise exposed subjects tended to correlate more strongly to DPOAEs of the same frequency, in contrast the thresholds in the non-noise exposed subjects correlated with higher frequency (4000-6000Hz) DPOAEs. The differences in patterns were more apparent for low frequency audiometric thresholds (below 3000Hz). Conclusions: The swept primaries approach with the LSF analysis provides an efficient and detailed method to separate the level and phase of the DPOAE components. The generator-component and combined-DPOAE levels generated in 1500-3000 Hz f2 region were consistently lower in noise-exposed subjects even when thresholds were clinically ‘normal’, and warrants further investigation as an early marker of preclinical NIHL.