The Longitudinal Association of Subclinical Hearing Loss With Cognition in the Health, Aging and Body Composition Study

Objectives: To examine the longitudinal association between subclinical hearing loss (SCHL) and neurocognitive performance. Design: Longitudinal analyses were conducted among 2,110 subjects who underwent audiometric testing in a US multi-centered epidemiologic cohort study. The primary exposure was better ear hearing (pure tone average). SCHL was defined as hearing <= 25 dB. The primary outcome was neurocognitive performance, measured by Digit Symbol Substitution Test (DSST), Modified Mini Mental State Examination (3MS), and CLOX1. Linear mixed models were performed to assess the longitudinal association between hearing and cognitive performance, adjusting for covariates. Models were fit among all individuals and among individuals with SCHL only. Results: Among 2,110 participants, mean (SD) age was 73.5 (2.9) years; 52.3% were women. Mean (SD) better ear pure tone average was 30.0 (13.1) dB. Mean follow-up was 9.1 years (range 3-16). Among all participants, worse hearing was associated with significantly steeper cognitive decline measured by the DSST [0.054-point/year steeper decrease per 10 dB worse hearing, 95% confidence interval (CI): 0.026-0.082] and 3MS (0.044-point/year steeper decrease per 10 dB worse hearing, CI: 0.026-0.062), but not CLOX1. Among those with SCHL, worse hearing was associated with significantly steeper cognitive performance decline as measured by DSST (0.121-point/year steeper decrease per 10 dB worse hearing, CI: 0.013-0.228), but not CLOX1 or 3MS. Conclusion: Among those with SCHL, worse hearing was associated with steeper cognitive performance declines over time as measured by DSST. The relationship between hearing loss and cognition may begin at earlier levels of hearing loss than previously recognized.