Association of Occupational Noise Exposure and Extended High-Frequency Hearing Loss in Young Workers With Normal Hearing

Abstract

Objectives: Studies on the relationship between occupational noise exposure and extended high-frequency (EHF) hearing loss are limited. This study investigated the relationship between occupational noise exposure and EHF hearing loss in workers exposed to noise as measured by sound pressure level, exposure duration, and kurtosis to help provide a basis for early detection and prevention of hearing loss in noise-exposed workers.

Design: A cross-sectional survey was conducted among 602 participants with 472 noise-exposed workers and 130 non-noise-exposed controls. General demographic characteristics, noise exposure data, and hearing thresholds at conventional frequencies (0.25 to 8 kHz) and EHF (9 to 16 kHz) were collected and analyzed. Linear mixed-effects model analyses between hearing thresholds of EHF and noise exposure indicators including the 8-h equivalent continuous A-weighted sound pressure level ( LAeq,8h ), cumulative noise exposure (CNE), and kurtosis-adjusted CNE (CNE-K) were conducted.

Results: Among the 602 participants included in the analysis, 472 individuals (78.4%) were occupationally exposed to noise exposures ≥75 dBA. Significant differences ( p < 0.05) were observed in sex, exposure duration, LAeq,8h , CNE, and CNE-K between the noise-exposed group and the nonexposed group. The mean hearing thresholds for all tested extended high frequencies ranging from 9 to 16 kHz were significantly higher in the noise-exposed group than in the nonexposed group ( p < 0.05). The mean hearing thresholds of subjects in different groups of LAeq,8h exposures were generally stable with little variance in the conventional frequencies (0.25 to 8 kHz) but differed in the EHF range. Moreover, EHF hearing loss appeared to be most prominent in the subjects exposed to noise with 80 dBA < LAeq,8h ≤ 85 dBA. After the combination of the sound pressure level, exposure duration, and kurtosis by using the noise exposure indicators CNE and CNE-K, the subjects at the different noise exposures showed significant differences in hearing thresholds at EHF ( p < 0.05). Linear mixed-effected model analyses showed that the CNE-K was the best to indicate noise-induced hearing loss among the three noise exposure indicators at EHF.

Conclusions: The results indicate that the EHF hearing threshold testing is more sensitive to identifying early occupational noise-induced hearing loss than conventional audiometry. The CNE-K, an indicator combining noise energy, exposure duration, and kurtosis, is a more comprehensive and effective method for assessing the risk of EHF hearing loss due to occupational noise exposure.

Keywords: Complex noise; Cumulative noise exposure; Extended high-frequency audiometry; Kurtosis; Noise-induced hearing loss; Occupational noise exposure.

Fig. 1.

Mean hearing thresholds of noise-exposed subjects and nonexposed subjects (A), by sex (B–C), different age groups (D–G). The error bars refer to the SE of the mean. * indicates FDR-adjusted p value <0.05. For detailed results, see Tables S1 and S2 in Supplemental Digital Content, http://links.lww.com/EANDH/B545. FDR, false discovery rate.

Fig. 2.

Mean hearing thresholds of the conventional and extended high frequencies at different exposure levels of noise exposure indicators L_Aeq,8h (A), CNE (B), and CNE-K (C). The nonexposed group was defined as workers exposed to noise with L_Aeq,8h <75 dBA. The error bars refer to the SE of the mean. * indicates FDR-adjusted p value <0.05. For detailed results, see Tables S3 through S8 in the supplementary file http://links.lww.com/EANDH/B545. CNE indicates cumulative noise exposure; CNE-K, kurtosis-adjusted cumulative noise exposure; FDR, false discovery rate.

Fig. 3.

Mean hearing thresholds of the conventional and extended high frequencies at different kurtosis levels when 75 dBA ≤ L_Aeq,8h ≤ 85 dBA (A), 85 dBA < L_Aeq,8h ≤ 90 dBA (B), and L_Aeq,8h > 90 dBA (C). The nonexposed group was defined as workers exposed to noise with L_Aeq,8h < 75 dBA. The error bars refer to the SE of the mean. * indicates FDR-adjusted p value <0.05. For detailed results, see Tables S9–S12 in Supplemental Digital Content, http://links.lww.com/EANDH/B545. FDR, false discovery rate.