On the bio-accessibility of 14 elements in welding fumes.

The bio-accessibility of 14 elements in welding fume particulate matter was investigated in 325 personal air samples collected during welding in two shipyards and one factory producing heavy machinery. The apparent solubility in a synthetic lung lining fluid (Hatch's solution) was used as proxy for the bio-accessibility. The Hatch solubility of the different elements was highly variable with a median < 1% for Al, Fe, Pb, Ti, between 4 and 6% for Co, Cr, Ni, V, W, between 13 and 27% for Cd, Cu, Mn, Zn, and 41% for Mo. For many elements, the solubility varied over a wide range of several tens of percent. The welding techniques used influenced the solubility of Co, Cr, Cu, Mn and V significantly. The plants investigated (i.e., the welded materials and used electrodes) had a significant influence on the solubility of Co, Cr, Cu, Mn, Mo, V and W. According to principal component analysis (PCA), the variation in solubility can be described by four components, which explain 69% of the variance. The first principal component mostly comprises elements that can predominantly occur as divalent cations, the second principal component elements often forming oxyanions. The principal components are independent of the absolute value of the Hatch solubility. The results of PCA indicate that the co-variation of Hatch solubility is mainly controlled by the most soluble compounds in contrast to the absolute value of apparent solubility, which is strongly influenced by the distribution of the elements between compounds with different equilibrium solubilities. The observed large variability and the significant differences between welding techniques and plants clearly show that the bio-accessibility cannot be obtained from the literature but has to be studied experimentally at each location of interest.

Air exposure assessment and biological monitoring of manganese and other major welding fume components in welders.

In a cross-sectional study, 96 welders were compared with 96 control subjects. Also 27 former welders, all diagnosed as having manganism, were examined. Exposure to welding fumes was determined in the 96 welders, while the concentration of elements in whole blood and urine was determined in all subjects. The geometric mean (GM) concentrations of manganese (Mn) and iron in the workroom air were 97 microg m(-3) (range 3-4620 microg m(-3); n=188) and 894 microg m(-3) (range 106-20 300 microg m(-3); n=188), respectively. Thus the Mn concentration in the workroom air was on average 10.6% (GM) of that of the Fe concentration. No substantial difference was observed in the air Mn concentrations when welding mild steel as compared to welding stainless steel. The arithmetic mean (AM) concentration of Mn in whole blood (B-Mn) was about 25% higher in the welders compared to the controls (8.6 vs. 6.9 microg l(-1); p < 0.001), while the difference in the urinary Mn concentrations did not attain statistical significance. A Pearson's correlation coefficient of 0.31 (p < 0.01) was calculated between B-Mn and Mn in the workroom air that was collected the day before blood sampling. Although the exposure to welding fumes in the patients had ceased on average 5.8 years prior to the study (range 4 years-7 years), their AM B-Mn concentration was still higher than in referents of similar age (8.7 microg l(-1) vs. 7.0 microg l(-1)). However, their urinary concentrations of cobolt, iron and Mn were all statistically significantly lower.