Identifying predictors of speech-in-noise perception in older adults with hearing impairment: the key role of working memory

Background: Speech-in-noise (SiN) perception is a critical challenge for older adults, particularly with age-related hearing loss (ARHL). Despite the well-established association between hearing loss and SiN difficulties in older adults, a comprehensive understanding of the multiple, interrelated factors influencing SiN performance remains limited. The present study investigates these relationships and aims to identify key predictors of SiN in older adults with and without hearing impairment, with a focus on the contributions of peripheral hearing, cognitive function, quality of life and subjective hearing experiences. Methods: A total of 114 older adults (64–80 years) were categorized into the hearing loss group, the hearing aid users’ group, and the control group, with 38 participants in each group. Pure-tone audiometry findings, performance on the SiN perception task, standardized assessments of working memory, processing speed, fluid and crystallized intelligence, inhibitory control and verbal fluency, and self-ratings of quality of life and subjective hearing experiences were obtained from all participants. Group differences were examined, and Random Forest regression was applied to identify key predictors of SiN performance. Results: SiN perception in ARHL was influenced by a combination of peripheral hearing, cognitive, and self-reports on quality of life and subjective hearing. Cognitive measures, particularly working memory, inhibitory control and processing speed, were found to be the key predictors of SiN perception in the hearing loss group. While in the hearing aid users’ and control group, quality of life measures primarily contributed to SiN perception. Conclusions: Overall, this study emphasizes the importance of adopting a multidimensional approach to understanding and addressing age-related SiN perception challenges, highlighting the need for consideration of sensory, cognitive, and psychosocial factors in both research and clinical practice. Supplementary information: The online version contains supplementary material available at 10.1186/s12877-025-06841-5.