作者: Ana Navas-Acien , Ellen K. Silbergeld , Roberto Pastor-Barriuso , Eliseo Guallar
DOI: 10.1097/EDE.0B013E3181AFEF88
关键词:
摘要: The role of inorganic arsenic exposure in chronic diseases, including type 2 diabetes, is a major public health research question. This has been underscored by recent epidemiologic1-4 and experimental5-8 evidence supporting increased risks at low levels. In this context, it critical to understand the biology technical limitations biomarkers exposure, usually measured urine.9-11 Total urine integrates from multiple sources (arsenite, arsenate) organic (mainly arsenobetaine, arsenosugars arsenolipids) compounds their metabolites (Figure 1). population-based studies, speciation important differentiate because arsenicals, mostly found seafood, have little toxicity relative its metabolites. Despite analytic advances measurement arsenosugars, arsenolipids metabolites,9 determination remains technically challenging epidemiologic studies. For example, those were not 2003-2004 National Health Nutrition Examination Survey (NHANES). Moreover, arsenite, arsenate methylarsonate—species that directly reflect metabolism 1)—were NHANES with high limits detection,9,12-14 only total arsenic, dimethylarsinate, arsenobetaine arsenocholine (a minor seafood arsenical) available for analyses endpoints. Figure 1 Inorganic exposure: relevant general population, source urine. 2003-2004, arsenate, methylarsonate (MA) below limit detection 96%, 94%, 65% 98% ... To evaluate association prevalence diabetes 2003-2004,2 we reported two main strategies remove contribution arsenicals marine origin arsenic. First, conducted between adjusted sociodemographic factors, risk markers intake (urine blood mercury). More importantly, factors but restricted participants levels (ie unlikely intake, whom would be derived mainly arsenic). magnitude subgroup was similar analysis adjusting mercury, demonstrating identified whole sample driven dependent on statistical method used control origin. In reanalysis published issue EPIDEMIOLOGY, Steinmaus et al.15 minus (in 13 individuals detected) as an “estimate exposure,” attempt arsenocholine, however, inadequate estimate does other unmeasured seafood. Indeed, correlation remained moderately strong (r=0.51, Figure 2). Discussion, authors indicated after subtracting 24% variance remaining still explained “probably due some dimethylarsinate are present same seafoods contain arsenobetaine”.15 misclassification induced when estimating can further appreciated comparing distribution (20th, 50th 80th, 90th 99th percentiles 2.7, 6.0, 11.9, 18.4 73.6 μg/L, respectively) among undetectable (<0.4 μg/L) 1.9, 3.9, 7.7, 11.8 30.4 respectively). As consequence, any uses needs adjust such itself. Futhermore, using index aggravated categorical analyses, e.g. quintiles. Many highest quintile (≥80th percentile) exposed substantial amounts categorization situation could reduce efficacy adjustment multivariable models. Figure 2 Relationship (AB) (AC) 2003-2004. Arsenocholine subtracted detectable Lines represent dose-response relationships ... Given longstanding experience al. investigating effects difficult why exposure” without controlling conclusions. findings consistent out 2008 paper.2 (Table 3 manuscript).