Genome-wide association study identifies 44 independent genomic loci for self-reported adult hearing difficulty in the UK Biobank cohort

Abstract

# Joint senior authors *Corresponding authors 2 Age-related hearing impairment (ARHI) is the most common sensory impairment in the aging population; a third of individuals are affected by disabling hearing loss by the age of 65 1 . ARHI is a multifactorial condition caused by both genetic and environmental factors, with estimates of heritability between 35% and 55% [2][3][4] . The genetic risk factors and underlying biological pathology of ARHI are largely unknown, meaning that targets for new therapies remain elusive.We performed genome-wide association studies (GWAS) for two self-reported hearing phenotypes, hearing difficulty (HDiff) and hearing aid use (HAid), using over 250,000 UK Biobank 5 volunteers aged between 40-69 years. We identified 44 independent genome-wide significant loci (P<5E-08), 33 of which have not previously been associated with any form of hearing loss. Gene sets from these loci are enriched in auditory processes such as synaptic activities, nervous system processes, inner ear morphology and cognition. Immunohistochemistry for protein localisation in adult mouse cochlea indicate metabolic, sensory and neuronal functions for NID2, CLRN2 and ARHGEF28 identified in the GWAS. These results provide new insight into the genetic landscape underlying susceptibility to ARHI.ARHI is characterised by a non-syndromic bilateral, sensorineural hearing loss that progresses with increasing age and is an established risk factor for depression 6-8 and dementia 9-12 . Hearing loss was ranked fourth in the latest study into the Global Burden of Diseases 13 , yet hearing amplification devices are the only treatment option currently available for ARHI. ARHI is expected to be a highly genetically heterogeneous trait given that over 150 genetic loci have been identified in non-syndromic hereditary hearing loss alone (https://hereditaryhearingloss.org/). Previous GWAS of ARHI have identified a small number of promising candidate genes, though there has been poor replication of findings to date, possibly reflecting varied phenotyping approaches and limited sample sizes 14-24 .We conducted two GWAS using the self-reported hearing difficulty and hearing aid use of UK Biobank (UKBB) participants and refined our results using a combination of conditional analysis, replication analysis, in silico annotation and in vivo expression analysis (see Figure 1 for study design). Our aim was to identify the genetic components of adult hearing impairment in the UK population and provide insight into the pathology of ARHI.3 UKBB participants were categorised using a case-control design based on responses to questions regarding hearing difficulty (HDiff, n=498,281) and hearing aid use (HAid, n=316,629) ( Supplementary Figure 1). A linear mixed-effects model was used to test for association between 9,740,198 SNPs and the two traits, using BOLT-LMM v.2 25 , which corrects for population stratification and within sample relatedness. Following additional quality control filters and selection of white British participants (described ...

GWAS of Chronic Dizziness in the Elderly Identifies Novel Loci ImplicatingMLLT10, BPTF, LINC01225, andROS1

“…Consistent with other studies, this GWAS reveals genetic correlation with other cochlear disorders such as hearing loss and tinnitus [24, 25]. Further, despite differences in phenotype, and despite differences in phenotype definitions, genetic correlation with Ice/Fin/UKB/US is still reasonably strong and highly significant (rg = 0.67 se 0.07, p-value 5.34 × 10 −20 ), indicating that our “chronic dizziness” definition shares common genetic influences with other acute cochlear and vestibular syndromes.…”

“…Regional association plots were generated using LocusZoom with 400-kilobase windows around the index variant and linkage disequilibrium patterns calculated based on 1000 Genomes Project ancestries appropriate for each lead SNP. 27 After GWAS on individual ancestries, summary statistics were combined in metaanalysis using METAL. 28 Results of each study were weighted proportional to the square-root of each study's sample size.…”

“…The proportion of heritability explained by our GWAS was performed with linkage disequilibrium scores (LDSC) [23]. Pairwise genetic correlations (rg) between results on chronic imbalance, hearing [24] and tinnitus [25] were calculated using bivariate LDSC regression, with publicly available GWAS summary statistics. LDSC regressions were computed with the Ice/Fin/UKB/US replication cohort for comparison.…”

Genetics of age‐related hearing loss

“…TRIOBP was the only previously identified locus that was identified, with 34 of the other loci have no previous links to any form of HL phenotypes. Having assessed all loci previously highlighted in ARHL GWAS prior to the CHARGE study, the only SNPs that were replicated in this sample were variants located in TRIOBP and close to ISG20 (Wells et al, 2019a). Within the study, various bioinformatic approaches were used including pathway, correlation, and expression analysis.…”