Cong Peng, Guoqing Gong, Zhetian Chen, Jingwei Wu, Dan Liu, Bin Chen
Speech-in-noise hearing impairment was associated with a higher risk of incident dementia in the UK Biobank, and this association was broadly observed across inherited-risk strata. These findings support hearing impairment as a potentially modifiable marker relevant to dementia-risk assessment across polygenic, APOE ε4, and familial risk backgrounds.
OBJECTIVE: Hearing impairment is considered an important modifiable risk factor for dementia, but whether its association with incident dementia is consistent across inherited susceptibility profiles remains unclear.
METHODS: This prospective cohort study included 80,287 UK Biobank participants aged 40 to 69 years at baseline with valid Digit Triplet Test measurements, genetic data, linked follow-up data, and complete prespecified covariates. Hearing function was quantified using the better-ear speech reception threshold (SRT). Incident dementia was ascertained from linked hospital admission and death registry records. Cox proportional hazards models were used to estimate associations between baseline hearing status and incident dementia. The fully adjusted model included demographic, socioeconomic, lifestyle, cardiometabolic, visual, psychosocial, hearing-aid-use, and genetic ancestry covariates. Effect modification by inherited susceptibility, including dementia polygenic risk score (PRS), APOE ε4 carrier status, and parental history of dementia, was assessed using stratified analyses and multiplicative interaction tests.
RESULTS: During ∼988,462 person-years of follow-up, 1086 incident dementia cases were identified. Compared with normal hearing, mild and severe hearing impairment were associated with higher dementia risk in the fully adjusted model, with HRs of 1.33 (95% CI: 1.13-1.56) and 1.53 (95% CI: 1.15-2.02), respectively. Restricted cubic spline analyses supported a nonlinear dose-response association between higher SRT and dementia risk. The hearing impairment-dementia association was observed across PRS, APOE ε4 carrier status, and parental family-history strata, with no statistically significant multiplicative interaction by APOE ε4 status or parental family history and only suggestive evidence by PRS category.
CONCLUSIONS: Speech-in-noise hearing impairment was associated with a higher risk of incident dementia in the UK Biobank, and this association was broadly observed across inherited-risk strata. These findings support hearing impairment as a potentially modifiable marker relevant to dementia-risk assessment across polygenic, APOE ε4, and familial risk backgrounds.