Levels and predictors of airborne and internal exposure to chromium and nickel among welders--results of the WELDOX study.

The objective of this analysis was to investigate levels and determinants of exposure to airborne and urinary chromium (Cr, CrU) and nickel (Ni, NiU) among 241 welders. Respirable and inhalable welding fume was collected during a shift, and the metal content was determined using inductively coupled plasma mass spectrometry. In post-shift urine, CrU and NiU were measured by means of graphite furnace atom absorption spectrometry, with resulting concentrations varying across a wide range. Due to a large fraction below the limits of quantitation we applied multiple imputations to the log-transformed exposure variables for the analysis of the data. Respirable Cr and Ni were about half of the concentrations of inhalable Cr and Ni, respectively. CrU and NiU were determined with medians of 1.2 μg/L (interquartile range <1.00; 3.61) and 2.9 μg/L (interquartile range <1.50; 5.97). Furthermore, Cr and Ni correlated in respirable welding fume (r=0.79, 95% CI 0.74-0.85) and urine (r=0.55, 95% CI 0.44-0.65). Regression models identified exposure-modulating variables in form of multiplicative factors and revealed slightly better model fits for Cr (R(2) respirable Cr 48%, CrU 55%) than for Ni (R(2) respirable Ni 42%, NiU 38%). The air concentrations were mainly predicted by the metal content in electrodes or base material in addition to the welding technique. Respirable Cr and Ni were good predictors for CrU and NiU, respectively. Exposure was higher when welding was performed in confined spaces or with inefficient ventilation, and lower in urine when respirators were used. In conclusion, statistical modelling allowed the evaluation of determinants of internal and external exposure to Cr and Ni in welders. Welding parameters were stronger predictors than workplace conditions. Airborne exposure was lowest inside respirators with supply of purified air.