Evaluating Metal Mixture Effects across Multiple Media

Abstract

Background/Aim: We developed multi-media biomarkers (MMBs) that integrate information across chemical exposure estimates when multiple biomarker media are used. Since each chemical has unique toxicokinetics and each medium (hair, urine, etc.) reflects differences in kinetics, no single medium is an ideal biomarker of exposure for all chemicals in a mixture. MMBs integrate exposure measures into a common metric that can be used in mixtures analyses. We propose that this will reduce exposure misclassification. Methods: Levels of lead (Pb) and manganese (Mn) were quantified in 5 media- blood, hair, nails, urine, and saliva. MMBs were constructed for both metals using weighted quantile sum (WQS) regression and the 5 media. We estimated the joint effects of both metals on Wechsler Intelligence Scale for Children (WISC) IQ scores in the Public Health Impact of Metals Exposure (PHIME) project, a study of Italian adolescents aged 10-14 years (N~720). Models were adjusted for age, sex, and SES. We compare the mixture effect of Pb and Mn, estimated using WQS regression, when Pb and Mn levels are quantified using MMBs and the original biomarkers. Results: Joint exposure to Pb and Mn was associated with decreased Verbal IQ when Pb and Mn exposure levels were estimated using hair (-0.7, 95% CI [-1.3, -0.1]), saliva (-0.8, 95% CI [-1.3, -0.2]), and MMBs (-1.1, 95% CI [-1.8, -0.5]). Total IQ was associated with Pb and Mn when exposure was assessed using MMBs (-0.71, 95% CI [-1.3, -0.1]) and not with individual biomarkers. Weights indicate that blood and hair contributed most to the MMB for Pb, whereas hair and saliva contributed most to the MMB for Mn. Conclusions: Mixture analyses are improved by using MMBs to estimate exposure. Combining information across multiple biomarkers through MMBs enables greater accuracy in measuring mixture effects and provides increased statistical power over single biomarkers.