Abstract
Background
Self-reported hearing difficulty is a common hearing-related complaint and may reflect perceived auditory impairment, communication difficulties, or clinically relevant hearing loss. Observational studies have linked cardiometabolic traits to hearing-related outcomes, but these associations are difficult to interpret because of residual confounding and reverse causation.
Methods
We performed a two-sample Mendelian randomization (MR) study to evaluate the potential associations of five genetically predicted cardiometabolic traits—body mass index (BMI), C-reactive protein (CRP), systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), and liability to type 2 diabetes (T2D)—with self-reported hearing difficulty. The primary outcome was self-reported hearing difficulty in the UK Biobank (field 2247; OpenGWAS: ebi-a-GCST90013961). External replication was conducted using a large genome-wide association meta-analysis of hearing loss/difficulty in hearing (OpenGWAS: ebi-a-GCST90018857). The inverse-variance weighted (IVW) method was prespecified as the primary estimator, complemented by sensitivity analyses and multivariable MR (MVMR) to assess robustness and the independence of the observed association.
Results
Genetically predicted higher BMI was associated with higher odds of hearing difficulty in the primary analysis (IVW OR 1.076, 95% CI 1.030–1.124; P = 0.0011), and a directionally concordant association was observed in the external replication dataset (OR 1.092, 95% CI 1.012–1.178; P = 0.0226). By contrast, IVW estimates for CRP, SBP, LDL-C, and T2D liability were close to the null and showed no consistent evidence of association across the two outcome datasets. In MVMR including BMI and CRP, the BMI estimate attenuated toward the null, whereas CRP showed no independent association.
Conclusions
These findings provide suggestive genetic evidence that higher BMI is associated with increased liability to self-reported hearing difficulty, whereas the other cardiometabolic traits examined showed limited evidence of consistent associations. However, the BMI association should be interpreted cautiously given the heterogeneity, pleiotropy-related findings, and attenuation observed in MVMR.