927 MENDELIAN RANDOMIZATION ANALYSIS SHOWS A CAUSAL EFFECT OF NOISE EXPOSURE ON ALZHEIMER’S DISEASE RISK

Abstract

Abstract Introduction Epidemiological studies suggested that exposure to transportation noise and long-term community noise was associated with a higher incidence of dementia, especially Alzheimer’s disease (ad). However, since subjects with low education and low income were also more likely to live or work in noisy environments, the association observed between noise exposure and ad may be distorted by education level, income level, and hearing impairment. Thus, whether noise exposure was causally associated with ad risk remained to be further clarified. Method A bidirectional Mendelian randomization (BMR) analysis was performed leveraging genome-wide association studies (GWAS) summary statistics of noise exposure (N = 456,380, UK Biobank) and ad (N = 54,162, IGAP) from the OpenGWAS database. Those single nucleotide polymorphisms (SNPs) with a P value less than 1E-05 were selected as instrumental variables (IVs). The TwoSampleMR package (version 0.5.6) was used to do BMR analysis using the inverse variance weighted (IVW) method with default parameters. Results The IVW results howed that exposure to noise at different times of day was causally associated with increased risk of ad (daytime: odds ratio [OR] = 2.2113, P = 0.0326; evening: OR = 2.2127, P = 0.0324; nighttime: OR = 2.2110, P = 0.0326; 16-hour: OR = 2.2114, P = 0.0325; 24-hour: OR = 2.2128, P = 0.0324). No obvious pleiotropic effects and heterogeneity were found in the BMR analysis, suggesting good reliability of the causal effect of noise exposure on ad. Moreover, the BMR results remained significant despite removing IVs associated with an education degree, average income, and hearing impairment. There was no reverse causal effect of ad on noise exposure in BMR analysis, which has been deposited in FigShare (https://figshare.com/s/09e30a1a0ccff450246b). Conclusion The results BMR analysis demonstrate that noise exposure is causally associated with an elevated risk of ad, independently of education degree, average income, and hearing impairment. Further studies are needed to confirm the underlying mechanism of this effect.