Unveiling the molecular bridge between hearing loss and dementia via plasma proteomic mapping.

Abstract

BackgroundHearing loss (HL) is a potentially modifiable risk factor for dementia, but the circulating molecular features shared by hearing impairment and dementia remain unclear.ObjectiveTo derive a plasma proteomic signature of hearing impairment and examine its association with incident all-cause dementia and its statistical contribution to the hearing-dementia association.MethodsWe analyzed 13,941 UK Biobank participants with plasma proteomics and speech-in-noise hearing assessments. A proteomic signature of hearing impairment was derived using elastic-net regression with 10-fold cross-validation. A proteomic signature was constructed and tested for associations with incident all-cause dementia using Cox models. Mediation, functional enrichment, and exploratory drug-signature enrichment analyses were also performed.ResultsElastic-net regression selected 64 proteins, and the composite signature correlated with hearing impairment (point-biserial r = 0.232). Compared with normal hearing, insufficient (HR = 1.338, 95% CI: 1.015-1.763) and poor hearing (HR = 2.021, 95% CI: 1.244-3.283) were associated with higher dementia risk after multivariable adjustment. The proteomic signature was associated with incident dementia (HR = 1.791, 95% CI: 1.312-2.445). Estimated proportions mediated were 22.99% (95% bootstrap CI: 4.87%-129.05%) and 14.61% (95% bootstrap CI: 2.21%-64.21%), respectively. Drug-signature enrichment yielded exploratory compound-protein signals.ConclusionsIn this prospective cohort, hearing impairment and a derived plasma proteomic signature were associated with incident all-cause dementia. These findings provide hypothesis-generating evidence for circulating molecular features shared by hearing impairment and dementia but require external validation and mechanistic investigation before causal or clinical interpretation.