Discovery of an arsenic and mercury co-elevation in the Midwest United States using reference laboratory data

Abstract

This study aimed to determine if there is a co-elevation of human blood arsenic and mercury levels in the Midwestern population of the United States (U.S.) and to determine any geographical patterns and variation of arsenic and mercury that may exist in Michigan. 58,800 blood specimens along with associated demographic/geographic data from the contiguous United States were reviewed. Univariate and multivariable logistic regression were used to analyze demographic/geographic variables associated with elevated arsenic concentrations. Furthermore, blood data from patients in Michigan were aggregated to the ZIP code tabulation area (ZCTA) in order to assess geographic variation using spatial regression models. SaTScan software was also used to analyze potential clustering of arsenic and mercury across Michigan ZCTAs. Within the contiguous United States, elevated mercury blood concentrations, older age, female sex, and coastal status were all associated with elevated arsenic blood concentrations (elevated mercury odds ratio (OR) 3.18 (3.04–3.33); female sex OR 1.06 (1.02–1.11); +10 yr age OR 1.12 (1.11–1.14); coastal state OR 1.33 (1.27–1.40). Within the state of Michigan, as with the continuous U.S., elevated mercury blood concentrations and older age were associated with elevated arsenic blood concentrations (elevated mercury OR 2.75 (2.38–3.18); female sex OR 1.06 (0.95–1.19); +10 yr age OR 1.10 (1.06–1.13). Using spatial regression, it was determined that within Michigan, economic inequality (measured via the Gini coefficient) was also associated with elevated concentrations of mercury in the blood. Clinical reference laboratory data, in conjunction with spatial analysis methods, may enhance our understanding of how elemental exposure affects human health and should be considered for studying how environmental contaminants, socioeconomics and geography affect the health of populations.