LAUSR

Background: Noise pollution is common and can affect health, but mechanisms underlying this relationship remain incompletely characterized. We therefore examined the novel association between aircraft noise and DNA methylation (DNAm), an environmentally modifiable epigenetic phenomenon that affects gene expression and is associated with cardiovascular and neurological disease. Methods: We conducted a methylome-wide association study in race/ethnicity-stratified subpopulations of 4,535 post-menopausal Women’s Health Initiative participants (mean age: 64.4 years; 59.6% white; 24.7% African American; 15.7% Hispanic/Latina). We estimated DNAm in whole blood as a proportion of methylated cytosines at each Cytosine-phosphate-Guanine (CpG) site using the Illumina 450k BeadChip. In cooperation with the Federal Aviation Administration, we generated day-night-average sound level (DNL) contours around 90 major U.S. airports using the Aviation Environmental Design Tool. The DNL applies a penalty for nighttime exposure and is the primary metric for informing policy in the U.S. We estimated geocoded participant address-specific, annual, DNL aircraft noise exposures in decibels (dB) from the contours on the day of blood draw. We categorized aircraft noise exposures as non-exposed or exposed (DNL < 45 or ≥ 45dB) and used linear mixed-effects models to estimate CpG site-specific DNAm-noise associations adjusted for sociodemographic, behavioral, and technical covariates. We combined site-specific associations across race-ethnicity strata in fixed-effects, inverse variance-weighted meta-analyses and then identified the significant ( P < 1.0x10 -7 ), non-heterogeneous ( P Cochran’s Q > 0.10) associations among them. Results: Overall, DNAm was inversely associated with aircraft noise at cg21525369 (Chromosome 15; P = 6.40x10 -8 ). Cg21525369 is 56 kb upstream from ACAN , a gene encoding an extracellular matrix protein (aggrecan). Aggrecan is expressed in response to ultrasound; forms perineural nets around cochlear hair cells and auditory brainstem neurons; affects afferent extension, synaptic transmission, and neuronal vulnerability; deposits in tympanic membranes of patients with sensorineural hearing loss, and redistributes in auditory brainstem after deafferentation. Conclusions: The biologically plausible DNAm-noise association among postmenopausal U.S. women suggests aircraft noise pollution may affect methylation of DNA near a gene previously associated with hearing, although its cardiovascular, otologic, and neurological implications remain unclear pending sensitivity analysis, external replication, mediation analysis, and functional characterization.