Disentangling peripheral hearing loss from central and cognitive effects on speech intelligibility in older adults

A research team from Wakayama University has published a new framework in arXiv that challenges conventional wisdom about age-related hearing loss. Using their proprietary Wakayama University Hearing Impairment Simulator (WHIS) and an objective intelligibility metric called the Gammachirp Envelope Similarity Index (GESI), the researchers created a method to disentangle physical hearing decline from cognitive and central processing factors. This approach allows for systematic comparison between young normal-hearing listeners and older adults with varying hearing levels.

The study's most surprising finding emerged when comparing predicted versus actual speech intelligibility scores. While young listeners' scores aligned with AI predictions based on simulated hearing loss, many older adults significantly outperformed these benchmarks. This contradicts previous research suggesting uniform decline in speech perception with age and points to potential cognitive compensation strategies developed over time. The GESI metric successfully accounted for peripheral hearing loss, leaving residual differences that likely reflect these central cognitive factors.

This research provides a crucial methodological advancement for both audiology and AI-assisted hearing technology development. By isolating cognitive contributions, researchers can now design more personalized hearing aids and cognitive training programs. The framework also offers a standardized way to assess individual differences in how the brain processes degraded auditory signals, moving beyond simple audiogram-based assessments.