Associations of renal function with urinary excretion of metals: Evidence from NHANES 2003–2012

Abstract

BackgroundUrinary metals are considered measures of long-term exposures of metals, such as cadmium (Cd). Some studies indicate reduced renal function may affect the urinary excretion of several metals in general population making assessments difficult. ObjectivesTo examine whether reduced renal function is associated with reduced urinary excretion of 12 metals or their metabolites and, in turn, an underestimated measure of Cd in general population. MethodsWe conducted analyses using data from the National Health and Nutrition Examination Survey (NHANES) 2003–2012. Multiple linear regression models were used to examine the associations between urinary metal levels and estimated glomerular filtration rate (eGFR). Restricted cubic spline regression models were used to evaluate the nonlinearity. ResultsUrinary metal levels significantly increased (p < 0.001) with increasing eGFR, except for antimony (p = 0.172). Urinary levels of arsenic, dimethylarsonic acid, cobalt, molybdenum and tungsten increased linearly with eGFR, while Cd, lead, mercury, barium, cesium and thallium increased nonlinearly (p < 0.001) with eGFR. Based on a restricted cubic spline regression model, we found, corresponding to a fixed blood Cd adverse cutpoint of 5 μg/L, predicted urinary Cd cutpoints substantially varied from 0.78–1.21 μg/g for urinary Cd between those aged <40 years and who had chronic kidney disease and those aged 60 years or over with normal renal function, respectively. ConclusionReduced renal function is associated with reduced urinary metals; and associations are also observed across the eGFR range not just in the reduced range. Urinary abnormal cutpoints of metals are likely dependent on eGFR and age. The associations between urinary exposure of metals and disease risk are likely underestimated without considering the modifying effect of renal function.