Objectives: Chronic exposure to high levels of noise is believed to be associated with increased risk of cardiovascular disease but results of many studies to date have been equivocal. This inconsistency may in part be due to limitations in exposure assessment resulting in misclassification of exposure and attenuation of the exposure-disease relation. We undertook a quantitative retrospective exposure assessment using predictive statistical modeling to estimate historical exposures to noise among a cohort of 27,499 sawmill workers as part of an investigation of acute myocardial infarction mortality.
Methods: Noise exposure data was gathered from research, industry and regulatory sources. An exposure data matrix was defined and exposure level estimated for job-title/mill/time-period combinations utilizing regression analysis to model determinants of noise exposure. Cumulative exposure, and duration of exposure metrics were calculated for each subject. These were merged with work history data, and exposure-response associations were tested in subsequent epidemiologic studies, reported elsewhere.
Results: Over 14,000 noise measurements were obtained. A subset, comprising 1,901 measurements from was used in producing a predictive model (R2 = 0.51). The model was then used to estimate noise exposures for 3,809 "cells" of an exposure data matrix representing 81 jobs at 14 mills over a several decade period. Mean cumulative exposure was 101 dBA*yr. Mean duration of employment in jobs with exposure above thresholds of 85, 90 and 95 dBA, were 9.9, 7.0 and 3.2 years, respectively.
Conclusions: The utility of predictive statistical modeling for occupational noise exposure was demonstrated. The model required input data that was relatively easily obtained even retrospectively.
Statistics from Altmetric.com
Web only appendices 66;6:388-94
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.