Objective: Hearing loss (HL) has been associated with cognitive impairment in high-income countries. However, no study has investigated this association in low-and middle-income countries. Therefore, our aim was to investigate the association between cognitive function and HL in the Brazilian Longitudinal Study of Adult Health.Design: Cross-sectional analysis of Longitudinal Study of Adult Health (ELSA-Brasil) with 802 individuals (35-74 years old). Hearing was measured using pure-tone audiometry. A pure-tone average (s) of thresholds at 500, 1000, 2000, and 4000 Hz was calculated. HL was defined as a PTA above 25 dB in the better ear or either ear, as a categorical variable. Cognitive performance was measured using the Consortium to Establish a Registry for Alzheimer's Disease word list memory test, the semantic and phonemic verbal fluency (VF) tests, and the Trail Making test version B. To investigate the association between cognitive performance and HL, we used linear regression models adjusted for sociodemographic and clinical variables.Results: Of the total of participants, 7.6% had HL. After adjustment for sociodemographic and health confounding variables, only VF was associated with HL; a 10 dB increase in the PTA in the better ear was associated with worse performance in the phonemic VF test (β = −0.115 [95% CI, −0.203 to −0.027], p = 0.01). We found a significant interaction between HL and age in the VF domain (p = 0.01). HL was related to poor VF performance among older adults only.
Conclusion:In a community-dwelling sample of most middle-aged adults, objectively measured HL was associated with lower VF. These results should be evaluated with caution, given the likelihood of residual confounding and the fact that only VF showed an association with HL.