Associations between blood arsenic and urinary arsenic species concentrations as an exposure characterization tool

Blood arsenic has various toxicities including carcinogenicity, but urinary concentrations are often substituted to determine the exposure level. Since there is little information on the relation of urinary arsenic species to blood arsenic, the aim was to investigate relationships between blood total arsenic (T-As) and the urinary species adjusted by creatinine and specific gravity (SG). Blood and spot urine samples were collected from 109 Japanese subjects aged 18-66 years without occupational exposure. Positive correlations of blood T-As (median, 3.49 μg/L) with urinary creatinine-adjusted and SG-adjusted T-As and arsenobetaine were statistically significant and greater than those with the unadjusted ones. The magnitude of associations of blood T-As with creatinine-adjusted arsenic species was significantly larger than those with unadjusted or SG-adjusted ones. Most of the correlation coefficients among urinary arsenic species concentrations were significant in three adjustment methods, but there was not a significant correlation between monomethylarsonic acid and arsenobetaine after urinary creatinine and SG corrections. Given multiple regression analysis, plasma T-As concentrations showed significant relations to creatinine-adjusted T-As, dimethylarsinic acid, and arsenobetaine concentrations, but erythrocyte T-As did hardly reflect the variation of urinary arsenic species. In conclusion, creatinine-adjusted arsenic concentrations are suggested to be the most appropriate predictor of blood T-As; by contrast, use of the urinary unadjusted arsenic concentration may result in a misleading interpretation of inorganic arsenic toxicity because the associations between inorganic and organic arsenic species based on the unadjusted concentration were mutually close. Plasma T-As appeared to be the best indicator of low-level exposure in blood samples.