LAUSR

Cadmium (Cd) exposure has become a growing public health issue and the level of urinary Cd is commonly used as the internal biomarker of overall Cd exposure. There has been raised a concern whether the level of Cd in a single spot urine actually reflects individual internal exposure over a long-term period. We aimed to examine the variability of urinary Cd levels over three years. Levels of urinary Cd were determined repeatedly in 2238 general adults during a follow-up of three-year from a community-based prospective study. We estimated the intra-class correlation coefficients (ICCs) of urinary Cd level over three years using the three-level random-effects mixed models to assess their variations. We found that the Pearson correlations for urinary Cd over three years were 0.521 for uncorrected Cd, 0.632 for creatinine (Cr)-corrected Cd, and 0.551 for specific gravity (SG)-corrected Cd, respectively (all P < 0.001). Moderate reproducibility was obtained for urinary Cd over three years, where ICCs of the three methods all exceeded 0.50. Of note, Cr-corrected urinary Cd levels achieved high reproducibility [0.773, 95%CI (0.750-0.794)]. Additionally, positive dose-response associations of smoking amount with Cr-corrected urinary Cd level were observed (P trend <0.05). Our findings suggest that Cr-corrected urinary Cd level in a single measurement was a credible biomarker for the relatively long-term levels of urinary Cd in the general population and cigarette smoking plays a part of urinary Cd exposure.