The Search for a Diagnostic Indicator of Cochlear Deafferentation: Predicting Age from Auditory Evoked Potential Measures

Objectives: Cochlear synaptopathy, a type of cochlear deafferentation that occurs with aging, is expected to be common in humans and to have negative impacts on auditory perception. However, there is currently no means for diagnosing cochlear deafferentation in living humans. Auditory brainstem response (ABR) wave I amplitude and the envelope following response (EFR) are auditory evoked potentials that have been proposed as potential non-invasive indicators of cochlear deafferentation. However, these measures may be impacted by outer hair cell (OHC) dysfunction and changes in endocochlear potential (EP), making them difficult to interpret. One potential method for estimating the degree of deafferentation in individual patients is to combine evoked potential and distortion product otoacoustic emission (DPOAE) measurements with a computational model of the auditory periphery (CMAP). The goal of this study was to evaluate the ability of auditory evoked potentials, with and without the CMAP, to predict age, a risk factor for cochlear deafferentation. Design: In a population with up to a mild sensorineural hearing loss, a CMAP was used with Bayesian regression to predict the frequency-specific degree of deafferentation for individual human participants based on their ABR, EFR, and/or DPOAE measurements and determine whether changes in EP could influence these predictions. Linear regression models were then used to evaluate the ability of the estimated degree of deafferentation and various ABR wave I amplitude, EFR magnitude, and DPOAE measurements to predict age. Results: High frequency ABR wave I amplitude measurements and estimates of deafferentation generated from high frequency ABR wave I amplitudes performed the best at predicting participant age. Accounting for potential changes in OHC function (as indicated by DPOAEs) or the EP in the estimates of deafferentation or by including DPOAEs in the linear regression models had limited impact on the ability of ABR wave I amplitudes to predict age. Conclusions: High frequency ABR wave I amplitudes and estimates of cochlear deafferentation generated from high frequency ABR wave I amplitudes were able to predict participant age within approximately 6 years, with or without incorporating DPOAE measurements. Modeling results indicate that changes in the EP are not a compelling explanation for this outcome. This suggests that high frequency ABR wave I amplitude measurements are good candidates for non-invasive diagnosis of age-related cochlear deafferentation and the CMAP can be used with ABR measurements to predict deafferentation in individual patients.