Confounding variables in the environmental toxicology of arsenic

Abstract

Arsenic is one of the most important global environmental toxicants. For example, in regions of West Bengal and Inner Mongolia, more than 100 000 persons are chronically exposed to well water often strongly contaminated with As. Unfortunately, a toxicologically safe risk assessment and standard setting, especially for long-term and low-dose exposures to arsenic, is not possible. One reason is that the key mechanism of arsenic’s tumorigenicity still is not elucidated. Experimental data indicate that either DNA repair inhibition or DNA methylation status alteration may be causal explanations. Moreover, when comparing epidemiological data, it cannot be ruled out that the susceptibility to arsenic’s carcinogenicity may be different between Mexican and Taiwanese people. Some other studies indicate that some Andean populations do not develop skin cancer after long-term exposure to As. It is not known yet how this resistance could be mediated. Finally, the situation is even more complicated when taking into consideration that there are several compounds suspected to modulate the chronic environmental toxicity of arsenic, variables that may either enhance or suppress the in vivo genotoxicity and carcinogenicity of the metalloid. Among them are nutritional factors like selenium and zinc as well as drinking water co-contaminants like antimony. Further, yet unidentified factors influencing the body burden and/or the excretion of arsenic are possibly prevailing: preliminary data from own human biomonitoring studies showed a peaking of As in urine samples of non-exposed people which was not caused by elevated exposure to As through seafood consumption. The relevance of these putative confounding variables cannot be finally evaluated yet. Further experimental as well as epidemiological studies are needed to answer these questions. This would help to conduct a toxicologically improved risk assessment, especially for low-dose and long-term exposures to arsenic.