1326-P: Diabetes Is Associated with a Higher Hearing Loss Risk in an Observational and Mendelian Randomization Study

Abstract

Observational studies on diabetes and hearing loss have been controversial, and their potential causal relationship has not been investigated. We conducted Mendelian randomization (MR) analyses to determine whether type 1 diabetes (T1D), type 2 diabetes (T2D), fasting insulin (FI), 2-h glucose (2hG), fasting glucose (FG), and glycated hemoglobin (HbA1c) were associated with hearing loss. Among 497,998 individuals in the UK biobank, T1D and T2D were identified by medical history in questionnaires and clinical records. Hearing loss was primarily defined by speech-in-noise tests in the better ear (SiN hearing loss, SiNHL) and secondarily identified by International Classification of Diseases code (ICD-defined hearing loss, ICDHL) and self-reported data (hearing difficulty, HDiff; hearing difficulty with noise, HDiffn; and hearing aid use, HAid). In observational analyses, odds ratios (ORs) in people with T1D than those without diabetes were 1.27 (95% CI 1.06-1.53) for SiNHL, 1.36 (1.14-1.62) for HAid, and 1.78 (1.49-2.12) for ICDHL. The corresponding ORs in people with T2D were 1.17 (1.09-1.24), 1.16 (1.08-1.23), and 1.25 (1.16-1.34), respectively. In MR analyses, genetically predicted T1D and T2D were associated with a higher risk of SiNHL (OR 1.02 per 1-unit higher log-odds of T1D 1.02 [1.00-1.03]; T2D 1.03 [0.99-1.06]). Further analysis of glycemic traits supported an inverse association of genetically elevated FI on SiNHL (per 1-log unit increase in pmol/L FI 0.71 [0.54-0.94]), and a positive association for genetically elevated 2hG (per 1-log unit increase in mmol/L 2hG 1.18 [1.06-1.31]). Results were largely robust to sensitivity analyses accounting for horizontal pleiotropy, sample overlap, and definitions of hearing loss. Our study suggested a potential causal relationship between diabetes and hearing loss risk. The finding warrants confirmation by larger population-based studies and experimental studies on potential biological pathways. Disclosure T.Li: None. M.Sun: None. M.Wang: None. G.Zong: None. W.Gan: Employee; Novo Nordisk A/S. C.Feng: None. Y.Liu: None. Y.Cao: None. H.Han: None. D.Fu: None. H.Xing: None. H.Qin: None. Funding National Science Fund for Excellent Young Scholars (81922060)