Abstract:
Objectives: The study investigated predictors of speech-in-noise perception among younger (18-24 years) and older (55-70 years) adults with normal hearing sensitivity, examining the relative contributions of hearing sensitivity, psychoacoustic abilities, cognitive function, and lexical access.
Design: Fifty-five participants (30 young, 25 older adults) with normal hearing were assessed with a comprehensive test battery. Measurements included speech perception of Kannada sentences in speech-shaped noise and speech babble, hearing sensitivity (conventional, extended high frequency), Psychoacoustic tasks (amplitude modulation, frequency modulation, binaural temporal fine structure), cognitive assessments (working memory, lexical access), and subjective questionnaires (SCAP-A, Hearing handicap questionnaire). Stepwise-regression and mediation analyses were used to identify key predictors and examine age-related effects.
Results: In older adults, lexical access emerged as the strongest predictor of speech-in-noise performance, explaining 35.8% of variance for speech-shaped noise and 34.7% for babble conditions. Binaural temporal fine structure processing contributed additional significant variance in both conditions. For speech-shaped noise only, the Screening Checklist for Auditory Processing (SCAP-A) provided further predictive power. In younger adults, the models explained substantially less variance (~15%), with listening effort predicting performance in speech-shaped noise and extended high-frequency thresholds predicting performance in babble. Mediation analysis revealed that lexical access mediated 33-40% of age-related effects on speech perception.
Conclusions: Speech-in-noise perception among older adults with normal hearing was primarily predicted by lexical access abilities and binaural temporal processing, rather than general cognitive measures like working memory. The strong mediating role of lexical access suggests it may be a key pathway through which aging affects speech perception. Further research in this direction is indicated, both in terms of using lexical access as an evaluation tool as well as in aural rehabilitation programs. Different predictive patterns between age groups and noise types indicate that clinical assessments should incorporate both energetic and informational masking conditions to better capture real-world listening challenges.
