|
|
|
|
|
|
|
The authors ask you:
Background: Workers in the flavor industry are at risk for bronchiolitis obliterans (BO) from exposure to butter flavoring chemicals including diacetyl. (1) We conducted longitudinal medical surveillance and workplace diacetyl sampling in flavor companies with the goals of controlling hazardous exposures and assessing workers at risk for BO. This study is designed to assess longitudinal respiratory symptom and spirometric changes in participating flavor workers.
Methods: Medical surveillance included spirometry and a respiratory symptom/exposure questionnaire. We measured workplace diacetyl samples during “worst case” production processes. Workers were classified as having low, medium or high cumulative diacetyl exposure based on years of exposure, frequency of exposure and mean diacetyl concentration by company. For population characterization of longitudinal spirometry, we used a mixed regression model. We are comparing individual spirometries to predicted thresholds for abnormal rates of decline based on the Longitudinal Limit of normal Decline (LLD) (2), the ACOEM threshold (3), and the cross-sectional LLN. A non-linear mixed model will be used to characterize changes in respiratory symptoms over time, with participants stratified by cumulative diacetyl exposure, job category and smoking.
Initial Results: 194 workers from 7 flavor manufacturing companies participated at least twice in our consultation program between 2004 and 2007. The rate of decline in FEV1 among production workers was not significantly different from non-production workers. After stratifying for smoking, however, non-smoker production workers had a greater rate of decline in FEV1 compared to current-smoker production workers. Among workers in the highest exposure category, both never- and former-smokers had significantly greater rates of decline in FEV1 compared to current smokers.
Cross-sectional studies of flavor (4) and diacetyl production and microwave popcorn manufacturing workers (1) have shown that exposure to butter flavoring chemicals containing diacetyl is associated with high rates of respiratory symptoms and fixed airflow limitation.
Design: This is a prospective clustered cohort study of flavor industry workers.
Study setting: Seven flavor manufacturing companies participated at least twice in the National Jewish Medical and Research Center Respiratory Health Consultation Program between 2004-2007. Participants were included from 2 categories: Production & Non-Production (Lab/QC, Warehouse, & Administration).
Medical surveillance: Medical surveillance included a symptom/exposure questionnaire and spirometry. Questionnaires were completed by individual workers and reviewed with a physician. Smoking status was assessed as never-smoker, former-smoker & current-smoker.
Spirometry: Spirometry was obtained by NIOSH-certified technicians following ATS guidelines, with quality assurance review. Frequency and duration of repeat testing varied among companies.
Exposure sampling: Mean and peak diacetyl samples were obtained during “worst case exposure” production processes. Personal and area monitoring was performed using NIOSH methods and analytical laboratories. Individual cumulative exposure was calculated by multiplying mean diacetyl level by company (ppm) by frequency of diacetyl use by years worked in current job, Participants were assigned to low (less than 0.0002 ppm), medium (0.0002 – <0.03), and high (0.03 or greater) exposure groups.
Analysis plan: 194 workers underwent 2 or more tests. We used the general linear mixed model to evaluate longitudinal change in FEV1. (5) We stratified the overall longitudinal change among workers by smoking status, and then by cumulative diacetyl exposure, or by job task (production versus non-production). After assessing longitudinal change for the population, we will compare individual changes with the predicted longitudinal limit of normal decline (LLD) (2). Additionally we will compare the individual’s change to thresholds of abnormal rate of decline predicted by ACOEM (3), and to the cross-sectional lower limit of normal (LLN) using SPIROLA software developed by NIOSH. After assessing individual changes, we will calculate the sensitivity and specificity of the LLD, ACOEM and cross-sectional LLN methods compared to definitions of “true” excessive decline in FEV1 of >60 ml/year versus > 90 ml/year versus > 330 ml/year. We will then use the nonlinear mixed model to characterize change in respiratory symptoms over time, stratifying the overall longitudinal change by cumulative diacetyl exposure, job task and smoking. All analyses will be performed using SAS© software.
Preliminary analysis shows that the data fit a linear pattern of FEV1 versus age. The estimated mean slope was -37.94 ml/yr (-0.03794 L) with standard error 0.004575 L. While there were no overall differences between production and non-production workers, non-smoker production workers had greater rates of decline in FEV1 compared to current-smoker production workers. See Table 1.
There were no significant differences among high, medium or low cumulative exposure groups. Among workers in the high-exposure group, however, both never- and former-smokers had significantly greater rates of decline in FEV1 compared to current smokers. See Figure 1.
While analysis is incomplete, initial review reveals that the general linear mixed model is an appropriate model for analysis of our data. This preliminary analysis suggests that non-smoking flavor workers may experience greater rates of decline in FEV1 compared to currently smoking flavor workers. This is consistent with cross-sectional evaluations of popcorn and flavor worker cohorts. Verification of this observation and in-depth assessment of longitudinal changes in spirometry in flavor workers at risk for bronchiolitis obliterans awaits further data analysis.
The findings and conclusions in this poster are those of the author(s) and do not necessarily represent the views of the National Institute for Occupational Safety and Health. Citations to Web sites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these Web sites.
View Comments on this Poster