Noise-induced hearing loss (NIHL) is a major occupational health risk in industrialized countries worldwide that affects people of all ages, sex, and races. About 22-30 million US workers are exposed to hazardous noise levels at work and an estimated US $242 million is spent annually on compensation for disability associated with hearing loss^[1-2]. Data from the Centers for Disease Control and Prevention (China) showed that NIHL is the third most serious occupational disease in China^[3]. NIHL not only affects workers' health, but also causes social isolation, impaired communication, and decreased productivity.

NIHL is a complex form of hearing loss induced by interactions between genetic and environmental factors. Noise is the most studied environmental factor associated with hearing loss; it is harmful over 85 dB and causes both mechanical and metabolic damage. However, not all workers develop NIHL after exposure to identical noise levels. Thus, genetic factors might also influence the susceptibility to NIHL.

PCDH15 is a member of the cadherin superfamily of calcium-dependent cell-cell adhesion molecules^[4], which is localized in the inner ear hair cell stereocilia and retinal photoreceptors. Mutations in PCDH15 have been associated with both non-syndromic (DFNB23) and syndromic hearing loss (Usher syndrome type1F, USH1F)^[5-6]. In 2009, research in Swedish and Polish populations first reported that single nucleotide polymorphisms (SNPs) in PCDH15 were associated with NIHL risk^[7]. Even though the reported variation is relatively common in certain European populations, racial differences need to be considered. Zhang et al. were the first to conduct research on this topic in China, reporting that the rs11004085 genetic variation in PCDH15 was associated with NIHL^[8]. However, the findings in this study were restricted to men living in South China; thus, more studies in other independent samples are necessary to confirm these findings. Therefore, this study investigated whether susceptibility to NIHL was associated with PCDH15 genetic variations in a northern Chinese population.

This study included 6, 309 workers exposed to continuous and steady occupational noise in a steel factory in Henan province, China. A detailed description of the inclusion and exclusion criteria can be found elsewhere^[9-10]. The case group was defined as those individuals with an average high-frequency (3, 4, and 6 kHz) binaural hearing level (HL)≥40 dB. The individually matched control group was defined as binaural HL of any frequency (0.5, 1, 2, 3, 4, and 6 kHz) < 25 dB. Finally, 344 matched pairs of participants were recruited from the steel factory cohort from September to December in 2013 and September to November in 2014. The study was approved by the Ethics Committee of the Henan Institute of Occupational Medicine.

All the subjects answered a structured questionnaire and underwent physical examinations by trained physicians. Noise exposure levels were assessed during their working time, which was evaluated with equivalent continuous dB (A)-weighted sound pressure levels (L_{Aeq, 8h}). Cumulative noise exposure (CNE) was calculated to determine the actual noise exposure for each subject^[9-10].

A total of 12 candidate SNPs were selected based on the inclusion criteria^[9-10], including rs10825112, rs10825113, rs1900443, rs12258253, rs2135720, rs11004085, rs11004142, rs996320, rs7081730, rs978842, rs11004439, and rs7922254. The genotypes were determined using the commercial SNPscan^TM multiplex SNP genotyping kit (Genesky Biopharm Technology Co., Ltd, Shanghai, China).

Hardy-Weinberg equilibrium (HWE) was checked for each SNP in the control subjects using χ²-tests. Paired samples t-tests were used to compare demographic information for continuous variables, while χ²-tests were used for categorical variables. Adjusted ORs with 95% CIs were computed by conditional logistic regression analysis to test for associations between NIHL risk and the genotypes. Bonferroni correction was performed to control for multiple testing, which resulted in a corrected significance level of 0.004 (P=0.05/12=0.004). Haploview was used to estimate the haplotypes and investigate the linkage disequilibrium (LD) between the SNPs. All statistical analyses were two-sided and performed using IBM SPSS Statistics for Windows, Version 20.0, with a significance level of 0.05.

The basic characteristics of the subjects are shown in Table S1 (www.besjournal.com). The average binaural HL of high frequency noise exposure in the case group was significantly higher than that of the control subjects (P=0.024). The percentage of smokers was also higher in the case group (P < 0.001). The two groups appeared to be well matched by age, gender, body mass index (BMI), hypertension, drinking status, exposure time, the use of earplug, noise exposure level, and CNE (P > 0.05).

The distributions of the PCDH15 genotypes and alleles in the case and control subjects are shown in Table 1. Of the 12 SNPs, only one significant association of genotype rs11004085 was observed between the two groups (P=0.039). As shown in Table S2 (www.besjournal.com), for rs11004085, after adjusting for BMI, smoking, drinking, and CNE, the frequencies of the CT compared with the TT genotype in the case group was significantly higher than that in the control group (adjusted OR=3.03; 95% CI: 1.26-7.33, P=0.014). Compared with subjects carrying the TT genotype, subjects with CT/CC genotypes were at increased risk of NIHL (adjusted OR=2.64; 95% CI: 1.14-6.11, P=0.024). Thus, this SNP genotype (CT/CC on rs11004085) was identified as a risk factor associated with NIHL. These findings were in agreement with two previously published studies of Polish and Swedish populations and a southern Chinese population^[7-8]. Mutations on rs11004085 might decrease calcium binding capacity, weaken its interaction with other genes such as CDH23, and affect its adhesion function, which increase the stereocilia bundle susceptibility to noise injury^[4-6]. However, no significant differences between two groups in terms of the distribution of genotypes or alleles of the other 11 SNPs were found in our study, as shown in Table S3 (www.besjournal.com). Similarly, no significant differences were detected when Zhang and colleagues compared allele and genotype frequencies for the SNPs on rs12258253^[8]. This SNP did not seem to play an important role in NIHL in the Chinese population, although more studies are needed to confirm these findings.

As noise is the most common cause of NIHL, stratified analysis by noise exposure level or CNE was conducted, the results of which are shown in Table 2. For rs11004085, compared with the TT genotype, CT/TT genotypes resulted in an increased risk of NIHL for noise exposure levels > 85 dB (A) (adjusted OR=4.61; 95% CI: 1.25-17.04, P=0.022); for CNE > 95 dB (A), the CT/TT genotypes were also at increased risk (P=0.011), with adjusted OR of 5.26 and 95% CI of 1.47-18.79. For rs978842, no overall significant associations with NIHL were observed before stratification. Compared to those with TT genotypes, noise exposure levels > 85 dB (A) were associated with an increased risk for NIHL among subjects carrying TC/CC genotypes (adjusted OR=1.58; 95% CI: 1.03-2.42, P=0.035). Therefore, this study iden fied significant SNP-environment interac ons between two SNPs, rs11004085 and rs978842, and noise exposure, a finding concordant with that of previous research^[8]. It is not surprising that NIHL is posi vely correlated with noise exposure, as workers are more susce ble to NIHL when exposed to greater noise exposure levels. Taken together, these data suggest that the interac ons between PCDH15 polymorphisms and noise exposure might play important roles in the incidence of NIHL.

In order to ide fy true associa ons that might be missed because of the incomplete informa on provided by the individual SNP, we primarily applied a haplotype-centric approach, taking into account different levels of noise exposure, to test for interac ons. The pairwise LD between the 12 SNPs is shown in Table S4 (www.besjournal.com). From Table 3, no significant P-values were obtained before stra fica on (P > 0.05). When noise exposure level > 85 dB (A) were compared with haplotype TAGCT, the frequencies of haplotype TAGCC was significantly higher in the case group (adjusted OR=1.84; 95% CI=1.01-3.33, P < 0.05); for CNE > 95 dB (A), subjects with haplotype TAGCC were more susceptible to NIHL (adjusted OR=1.79; 95% CI=1.08-2.95, P < 0.05), indicating it to be a risk haplotype. Our finding was concordant with that of a previous study^[8], suggesting that multiple genetic variations in PCDH15 modify NIHL risk and that higher noise exposure might increase risk.

However, after applying Bonferroni correction, the associations were no longer statistically significant.

These findings are of value for the prevention of NIHL. People with the CT/CC rs11004085 genotype or the TAGCC PCDH15 risk haplotype should take care to avoid high levels of noise exposure in their workplaces.

The limitations of this study should be acknowledged when interpreting the results. Firstly, although some workers might be exposed to community noise, such as sleeping with the television on and listening to music with headphones, these factors were too complicated to consider in the current study. Secondly, while this study identified TAGCT as a potential risk haplotype, the molecular mechanism are not clear and further functional studies are warranted.

In conclusion, our findings indicated that genetic variations in PCDH15 may play an important role in genetic susceptibility to NIHL. The effect of gene-environment interactions and multiple loci on the development of NIHL were detected. However, after Bonferroni correction, these differences were not found to be significant. Further studies involving a larger number of individuals and independent populations are required to assess these findings.