Hearing impairment and missing cognitive test scores in a population‐based study of older adults: The Atherosclerosis Risk in Communities neurocognitive study

INTRODUCTION Hearing impairment is associated with poor cognitive test performance in older adults. However, hearing's impact on cognitive test completion is poorly described, and missing cognitive data due to hearing impairment could misestimate the association. METHODS We investigated if hearing impairment is associated with missing neurocognitive scores in 3678 adults (72-94 years). Hearing impairment was defined by the better-ear pure tone average of speech-frequency thresholds (0.5-4 kHz) >25 decibels. RESULTS Hearing impairment was associated with greater missingness on all auditory-only tests, including Logical Memory (prevalence ratio [PR] comparing ≥ moderate impairment vs normal hearing:1.68, 95% confidence interval [CI] 1.26, 2.25) and Digits Backwards (PR 1.62; 95% CI 1.21, 2.17); and two non-auditory tests, Boston Naming (PR 1.61; 95% CI 1.21, 2.17) and Trail Making B (PR 1.55; 95% CI 1.29, 1.86). Models that imputed missing cognitive scores showed the strongest hearing-cognition associations. DISCUSSION Older adults with hearing impairment are less likely to complete cognitive testing, thereby underestimating the hearing impairment-cognition relationship.