Document resource
Introduction Observational studies suggest an association between hyperglycaemia and hearing loss, but causal evidence remains uncertain due to potential confounding.Objective To evaluate the causal influence of hyperglycaemia on hearing loss using a multi-ancestry Mendelian randomisation (MR) framework.Methods A two-sample MR study used individual-level data from the Taiwan Biobank and summary statistics from public databases. Instrumental variables for HbA1c, fasting glucose and type 2 diabetes mellitus (T2DM) were analysed in relation to non-specific hearing loss (primary outcome) and specific subtypes—sensorineural, age-related and conductive hearing loss—as secondary outcomes. Causal effects were estimated via inverse variance weighting and pooled using random-effects models. Multivariable Mendelian randomisation (MVMR) adjusted for smoking, blood pressure, lipids and body mass index while bidirectional MR assessed causality direction.Results A total of nine non-specific hearing loss phenotypes and five specific subtypes were analysed across Asian and European populations. Pooled MR estimates showed that elevated HbA1c, glucose and T2DM were significantly associated with an increased risk of non-specific hearing loss with maximum ORs of 1.54 (95% CI 1.34 to 1.78), 1.45 (95% CI 1.22 to 1.71) and 1.13 (95% CI 1.07 to 1.19), respectively. Sensitivity analyses (MR-Egger, MR-PRESSO (Pleiotropy Residual Sum and Outlier)) showed no horizontal pleiotropy. MVMR confirmed these effects were independent of metabolic confounders, and bidirectional MR established a unidirectional pathway from hyperglycaemia to auditory impairment. Significant causal associations were also observed across all specific subtypes.Conclusion This study provides robust genetic evidence supporting a causal link between hyperglycaemia and hearing impairment. Findings emphasise the clinical importance of glycaemic control in hearing preservation strategies.