Pharmacokinetics/pharmacodynamics (PK/PD) modeling of risks of kidney toxicity from exposure to cadmium: estimates of dietary risks in the U.S. population.

An analysis of epidemiological studies of associations between exposure to cadmium and kidney toxicity was conducted. Dose-response functions relating low-molecular-weight (LMW) proteinuria to various indices of cadmium dose (dietary cadmium intake, urinary cadmium excretion, or tissue cadmium burden) were obtained from 15 studies of diverse exposures (occupational, general environmental, environmental contamination). Estimates of the dose corresponding to probabilities of LMW proteinuria of 0.1, 0.15, or 0.2 were transformed from the reported dose units into corresponding estimates of target organ dose (microgram Cd/g renal cortex, RC) by simulation using a pharmacokinetics (PK) model. The median RC associated with a 0.1 probability (RC10M) of LMW proteinuria was predicted to be 153 micrograms Cd/g cortex (95% confidence interval [CI]: 84-263). The lower confidence limit on the RC10M (RC10L, 84 micrograms/g cortex) was predicted to be attained with a constant chronic intake of 1 microgram/kg/d in females or 2.2 micrograms/kg/d in males. The RC10L was 2.5-5 times higher than the median RCs predicted to result from dietary cadmium intake in U.S. nonsmokers (microgram Cd/g cortex: 33, females; 17, males) and 1.6-3 times higher than the corresponding 95th percentile RCs (53, females; 27, males). Additional exposure from smoking cigarettes (approximately 20 cigarettes/d, 3 micrograms Cd inhaled/d) was predicted to increase the median RC (microgram/g cortex) by approximately 45-70% (48, females; 29, males); however, predicted 95th percentile RCs for smokers (66, females; 38, males) were lower than the RC10L. These results indicate that, for most of the U.S. population, dietary-derived risks are likely to be negligible, in the absence of exposures from other sources.