Uncovering genetic comorbidities of speech-in-noise deficits using polygenic risk score analysis from two independent cohorts

Objective: Understanding speech in noisy environments is a crucial aspect of daily communication, and difficulties in this ability, known as speech-in-noise (SIN) deficits, can occur even in individuals with clinically normal hearing sensitivity. This study aimed to investigate the genetic comorbidities of SIN deficits by employing a polygenic risk score (PRS) analysis across two independent cohorts. PRS summarizes a complex genetic inheritance of health conditions or traits into a single score. The primary objective was to identify PRS predictors across the health spectrum associated with SIN deficits.Method: The study utilized two independent cohorts - the UK Biobank (N=343,104) served as the discovery sample, where SIN deficits were defined based on self-reported hearing difficulty in noise but self-reported normal hearing in quiet. The AudioGEN cohort (N=228), comprising healthy young adults with self-reported normal hearing, served as the replication sample. SIN deficits were measured using the Speech, Spatial, and Qualities of Hearing scale (SSQ12), QuickSIN, and 3-digit dichotic digit test (DDT). PRS were calculated for 2627 common health conditions and disease phenotypes from the polygenic risk score (PGS) catalog. Statistical analysis involved logistic regression in the UK Biobank to identify PRS predictors of SIN deficits, and linear mixed models (LMM) in the AudioGEN cohort to evaluate the association between PRS predictors and SIN measures, adjusting for relevant covariates. Enrichment analysis was performed to determine if PRS predictors associated with SIN deficits were over-represented in specific trait categories.