Abstract

OBJECTIVES: Four previously conducted epidemiological studies in more than 1200 grain workers were used to compare exposure-response relations between exposure to grain dust and respiratory health. METHODS: The studies included Dutch workers from an animal feed mill and a transfer grain elevator and Canadian workers from a terminal grain elevator and the docks. Relations between forced expiratory volume in one second (FEV1) and exposure were analysed with multiple regression analysis corrected for smoking, age, and height. Exposure variables examined included cumulative and current dust exposure and the numbers of years a subject was employed in the industry. Sampling efficiencies of the Dutch and Canadian measurement techniques were compared in a pilot study. Results of this study were used to correct slopes of exposure-response relations for differences in dust fractions sampled by Dutch and Canadian personal dust samplers. RESULTS: Negative exposure-response relations were shown for regressions of FEV1 on cumulative and current exposure and years employed. Slopes of the exposure-response relations differed by a factor of three to five between industries, apart from results for cumulative exposure. Here the variation in slopes differed by a factor of 100, from -1 to -0.009 ml/mg.y/m3. The variation in slopes between industries reduced to between twofold to fivefold when the Dutch transfer elevator workers were not considered. There was evidence that the small exposure-response slope found for this group is caused by misclassification of exposure and a strong healthy worker effect. Alternative, but less likely explanations for the variation in slopes were differences in exposure concentrations, composition of grain dust, exposure characteristics, and measurement techniques. CONCLUSION: In conclusion, this study showed moderately similar negative exposure-response relations for four different populations from different countries, despite differences in methods of exposure assessment and exposure estimation.