Occupational noise exposure assessment using O*NET and its application to a study of hearing loss in the US general...

Although occupational noise is a well known risk factor for hearing loss, little epidemiological evidence has been reported on its association with hearing loss in the general population, in part, because of the difficulty in exposure assessment. This study introduced a quantitative occupational noise exposure assessment tool using the Occupational Information Network (O*NET) database and evaluated its applicability for epidemiological research using data from the National Health and Nutrition Examination Survey (NHANES) 1999-2004.

The O*NET noise exposure data were assessed by questionnaires across numerous occupations, asking the frequency of exposure to sounds and noise levels that are distracting and uncomfortable (with five possible responses from 'never' to 'every day'). Means of the O*NET noise scores were computed to correspond to NHANES occupational categories and assigned to 3828 adults aged 20-69 years, who participated in the 1999-2004 NHANES. Pure-tone averages (PTA) of hearing thresholds at 0.5, 1, 2 and 4 kHz were computed, and hearing loss was defined as a PTA >25 dB in either ear. Linear and logistic regression models with either continuous or quintiles of the O*NET noise scores were fitted on log-transformed PTA and binary hearing loss, respectively.

Noise scores ranged from 1.80 to 4.37 with mean±SE of 3.06±0.02. After controlling for potential confounders, the highest (vs lowest) noise score quintile had a 22.5% (95% CI 11.0% to 35.2%) increase in PTA, and there was a linear dose-dependent trend across the quintiles of noise scores (p trend<0.0001). The adjusted OR for hearing loss comparing the highest with the lowest noise score quintiles was 2.1 (95% CI 1.2 to 3.6).

This study suggests that the O*NET noise score is a useful tool for examining occupational noise-induced health effects in the general population in the absence of actual occupational noise exposure assessment data.