Metabolic profiling of DNA methylation age acceleration, associations with hearing loss and interaction with genetic susceptibility

Limited studies have examined the association between DNA methylation age acceleration (DNAm AA) and hearing loss (HL). We aimed to assess whether plasma metabolite alterations are associated with the DNAm AA-HL link and explore the potential modifying effect of polygenic risk score (PRS). A total of 2068 individuals aged 65.7 years from the Dongfeng-Tongji cohort in 2013 were included. Hearing threshold was assessed using pure tone average (PTA), and HL was defined as PTA >25 dB. DNAm AA was calculated using 6 epigenetic clocks (HorvathAge, HannumAge, PhenoAge, GrimAge, DunedinPACE and DNAm mortality risk score). Two hundred and ten plasma metabolites were profiled by targeted metabolomics approaches. We applied generalized linear model coupled with enrichment analysis to characterize metabolomic profiles of DNAm AA, and elastic-net regression to construct metabolomic risk score (MRS). A weighted PRS for HL was constructed using 37 single nucleotide polymorphisms. Among 2068 participants, 1075 (52.0%) had HL and exhibited significantly higher DNAm age. HorvathAgeAccel, GrimAgeAccel, and DunedinPACE were associated with poorer PTA thresholds (β range: 0.59-1.56) and greater HL risks (odds ratios range: 1.11-1.21). GrimAgeAccel was linked to 28 metabolites and DunedinPACE to 65, involving 11 enrichment pathways and 6 metabolite sets. The MRSs were positively associated with PTA level or speech- and high-frequency HL (SFHL and HFHL) risk, and explained 33.4% to 48.2% of the associations of GrimAgeAccel and DunedinPACE with PTA or SFHL. Additionally, GrimAgeAccel and DunedinPACE exhibited positive associations with PTA or HL only in the high-PRS subgroup, and significant interaction was found between MRS for DunedinPACE and PRS on the risk of HFHL. GrimAgeAccel and DunedinPACE were associated with poorer PTA thresholds and higher HL risks, with their linked MRS partially explaining these associations. PRS modified the MRS-HFHL risk association, with a potentially amplified effect in high genetic susceptibility subgroups.