Relation between dietary cadmium intake and biomarkers of cadmium exposure in premenopausal women accounting for body iron stores

BACKGROUND: Cadmium is a widespread environmental pollutant with adverse effects on kidneys and bone, but with insufficiently elucidated public health consequences such as risk of end-stage renal diseases, fractures and cancer. Urinary cadmium is considered a valid biomarker of lifetime kidney accumulation from overall cadmium exposure and thus used in the assessment of cadmium-induced health effects. We aimed to assess the relationship between dietary cadmium intake assessed by analyses of duplicate food portions and cadmium concentrations in urine and blood, taking the toxicokinetics of cadmium into consideration. METHODS: In a sample of 57 non-smoking Swedish women aged 20-50 years, we assessed Pearson's correlation coefficients between: 1) Dietary intake of cadmium assessed by analyses of cadmium in duplicate food portions collected during four consecutive days and cadmium concentrations in urine, 2) Partial correlations between the duplicate food portions and urinary and blood cadmium concentrations, respectively, and 3) Model-predicted urinary cadmium concentration predicted from the dietary intake using a one-compartment toxicokinetic model (with individual data on age, weight and gastrointestinal cadmium absorption) and urinary cadmium concentration. RESULTS: The mean concentration of cadmium in urine was 0.18 (+/- s.d.0.12) μg/g creatinine and the model-predicted urinary cadmium concentration was 0.19 (+/- s.d.0.15) μg/g creatinine. The partial Pearson correlations between analyzed dietary cadmium intake and urinary cadmium or blood concentrations were r = 0.43 and 0.42, respectively. The correlation between diet and urinary cadmium increased to r = 0.54 when using a one-compartment model with individual gastrointestinal cadmium absorption coefficients based on the women's iron status. CONCLUSIONS: Our results indicate that measured dietary cadmium intake can reasonably well predict biomarkers of both long-term kidney accumulation (urine) and short-term exposure (blood). The predictions are improved when taking data on the iron status into account.