Effects of hearing difficulty on stroke: retrospective cohort and mendelian randomization studies

Despite observational studies associating hearing difficulty (HD) with an increased risk of stroke, biases, such as confounding and reverse causation, remain. This study aimed to elucidate the causal relationship between HD and stroke more definitively using Mendelian randomization (MR) and retrospective cohort studies. We employed a multifaceted, two-sample Mendelian randomization approach using large-scale GWAS data from the UK Biobank, FinnGen, and MEGASTROKE. This approach was complemented by retrospective cohort analyses of the Korean National Health Insurance Service Health Screening (NHIS-HEALS) and UK Biobank databases, applying propensity score matching and time-dependent Cox proportional hazards models to evaluate stroke risk associated with HD. The inverse variance weighted analysis demonstrated a negative causal association between HD and stroke (FinnGen: β = − 1.71; 95% confidence interval [CI] = − 2.43 to − 1.00; p = 2.38 × 10–6; MEGASTROKE: β = − 0.92; 95% CI = − 1.53 to − 0.31; p = 3.29 × 10–3). The weight median method similarly supported a negative causal association between HD and stroke (FinnGen: β = − 1.28; 95% CI = − 2.31 to − 0.24; p = 1.56 × 10–2; MEGASTROKE: β = − 0.81; 95% CI = − 1.59 to − 0.04; p = 0.039). In contrast, observational studies revealed that HD significantly increased the risk of stroke in both the Korean NHIS-HEALS (adjusted hazard ratio [aHR] = 1.83; 95% CI, 1.69 to 1.98; p < 0.001) and UK Biobank (aHR = 1.83; 95% CI, 1.53 to 1.19; p < 0.001) databases. This study provided robust evidence for a causal association between genetic predisposition to HD and stroke, which was a conflicting result compared with observational retrospective cohort studies. Future studies should focus on elucidating the biological pathways involved in HD that may influence stroke, which can potentially lead to novel preventive strategies and therapeutic interventions for stroke.