Use of background inorganic arsenic exposures to provide perspective on risk assessment results

Abstract

Background exposures provide perspective for interpreting calculated health risks associated with naturally occurring substances such as arsenic. Background inorganic arsenic intake from diet and water for children (ages 1–6 years) and all ages of the U.S. population was modeled stochastically using consumption data from USDA, published data on inorganic arsenic in foods, and EPA data on arsenic in drinking water. Mean and 90th percentile intakes for the U.S. population were 5.6 and 10.5 μg/day, assuming nationwide compliance with the 10 μg/L U.S. drinking water standard. Intakes for children were slightly lower (3.5 and 5.9 μg/day). Based on the current EPA cancer slope factor for arsenic, estimated lifetime risks associated with background diet and water at the mean and 90th percentile are 1 per 10,000 and 2 per 10,000, respectively. By comparison, reasonable maximum risks for arsenic in soil at 20 (higher typical background level) and 100 mg/kg are 4 per 100,000 and 2 per 10,000, using EPA default exposure assumptions. EPA reasonable maximum estimates of arsenic exposure from residential use of treated wood are likewise within background intakes. These examples provide context on how predicted risks compare to typical exposures within the U.S. population, thereby providing perspective for risk communication and regulatory decision-making on arsenic in the environment and in consumer products.

Introduction

Arsenic occurs naturally in the environment and is ubiquitous in water, food, and soil (ATSDR, 2000). Consequently, understanding how arsenic exposures from human activities compare to natural background exposures is important for communicating the relative magnitude of calculated risks in perspective with everyday exposures. Because risk assessments predict risks greater than one per million even for background exposures to arsenic, definition of de minimis risk at this level is less useful for indicating public health concern. Risk communication for arsenic will become even more critical if calculated risks for arsenic increase based on reanalyses (Morales et al., 2000, NRC, 2001; 66 CFR 6976–7066).

Recent attention has also been directed at children’s exposure to arsenic (CPSC, 2003, U.S. EPA, 2005a) and potential health risks (CPSC, 2003, Tsuji et al., 2004a, Tsuji et al., 2004b). In particular, concerns have been expressed regarding childhood exposure to inorganic arsenic from playsets and residential decks made of wood treated with chromated copper arsenate (CCA), because risk estimates exceed one per million (Patch and Maas, 2006). Although CCA-treated wood is no longer sold for residential uses, existing residential wood has not been recalled. Thus, context on how such exposures would compare to natural sources of inorganic arsenic from diet and typical water exposures would be helpful for risk communication and public health decision-making.

Background inorganic arsenic intake in the U.S. has been investigated recently for children exposed through diet (Yost et al., 2004), and previously for all ages exposed through diet (Schoof et al., 1999b), and for diet, water, soil, and air (Meacher et al., 2002). This paper builds on that earlier work to conduct probabilistic analyses of background inorganic arsenic intake for children (ages 1–6 years) and the general U.S. population (all ages) from the primary background sources, diet and water (Meacher et al., 2002). In comparison to the dietary intake modeling of Meacher et al. (2002), the more comprehensive dietary modeling reported by Yost et al. (2004) is used to estimate dietary and water intakes for the general U.S. population and for children ages 1–6 years. Our modeling also incorporated reduction in arsenic drinking water concentrations in the U.S. as a result of the compliance with the arsenic standard reduction from 50 to 10 μg/L.

As risk comparison examples, the results of this analysis are compared to arsenic intakes from soil exposure based on default EPA risk assessment assumptions and to estimated arsenic exposures from CCA-treated wood used for playsets and residential decks (U.S. EPA, 2005a, Zartarian et al., 2006).