Urine is better for rare earth elements bimonitoring in long-term exposed population: An exposure-response relationship study

With the soaring use of rare earth elements (REEs) worldwidely in high-technology and clean energy industries, there were growing concerns for adverse health effect from the REEs exposure. However, there is a lack of biomonitoring research concerning both urine and blood in population with definite exposure. We performed a biomonitoring study that involved 103 REEs exposed males and 110 males as non-REEs exposed controls. We measured the levels of REEs in environment and urine and blood samples from participants, and explored the exposure-response relationship between REEs in environment and body fluids. The effects of exposure duration and smoking status on the internal exposure level of REEs were also investigated. The results showed environmental REEs level of exposure group was significantly higher than that of control group (range of geometric mean of exposure vs. control: 1.08-4.07 x 10(4) ng/m(3) vs. <LOD-2.16 x 10(2) ng/m(3)). Six elements with detection rates higher than 60% in blood or urine samples were lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), samarium (Sm), gadolinium (Gd). We found the REEs concentrations both in urine and blood of exposure population were significantly higher than controls, median range of the above 6 elements of urine and blood was 0.02-1.06 mu mol/mol vs. <LOD-0.01 mu mol/mol creatinine and 0.01-0.79 mu g/L vs. <LOD-0.38 mu g/L. The correlations between both blood and urine level of REEs and environment level showed significant. The correlation coefficients with urine levels are higher than with blood. Biomonitoring results showed good exposure-response relationship in urine REEs, while no positive response in blood samples. Smoking status, drinking status and years of exposure showed little effect on the level of REEs. Our results suggested that both blood and urine can be used to monitor REEs exposure, while urinary REEs is promising for risk assessment in population.