oEPA
www.epa.gov/ord
science in ACTION
BUILDING A SCIENTIFIC FOUNDATION FOR SOUND ENVIRONMENTAL DECISIONS
EPA computer models predict exposures to arsenic that echo reality
Overview
Arsenic exists naturally in rocks, soil, water
and air. People's activities such as burning
coal, preserving wood and managing food
crops also contribute to arsenic in the
environment.
The U.S. Environmental Protection Agency
(EPA) has set the arsenic standard for
drinking water at 10 parts per billion (less
than one-tenth of a drop of water in a 53-
gallon tub) to protect consumers served by
public water systems from the effects of long-
term, chronic exposure to arsenic. Along with
protective limits for arsenic in drinking water,
arsenic exposures in occupational settings are
well described, and safeguards and limits
have also been established.
However, exposure to arsenic among the
general U.S. population from both food and
water ingestion is not well-defined because
the variables that impact people's exposures
to contaminants are difficult to measure, or
vary from person to person and community to
community, including the types and amounts
of food people eat, and water they drink.
Purpose of Research
A critical element of EPA's mission is to
protect public health by managing risks from
exposures to pollutants. To do this, EPA relies
on science-based decision-making and
innovative solutions.
EPA scientists recently conducted research to
improve EPA's capability to estimate
exposures to arsenic via two pathways
(drinking water and ingesting food) for the
general U.S. population.
Approach
Exposure estimates were developed using
state-of-the-science computer models to
combine available high-quality data on food
people eat and water they drink with
estimates of arsenic residues in foods and
drinking water. The accuracy of the modeled
estimates was then evaluated by comparing
them against physical measurements from
large-scale field surveys. Model development
and evaluation were performed using large
high-quality data sets from the U.S. Food and
Drug Administration (FDA), Natural Resources
Defense Council (NRDC), Centers for Disease
Control and Prevention (CDC), and EPA.
EPA scientists specializing in modeling
research used EPA's Stochastic Human
Exposure and Dose Simulation (SjHEDS)-
Dietary model in this research. SHEDS-
Dietary is a population-based model that
incorporates real-world food consumption and
chemical residue data, and applies advanced
statistical methods to simulate individual
exposures to chemicals over different time
periods (e.g., daily, yearly).
EPA's modeled predictions were compared to
arsenic intake from foods consumed by
people participating in EPA's National Human
Exposure Assessment Survey (NHEXAS)
duplicate diet study, as well as urinary
biomarker measurements provided by
participants in CDC's 2003-2004 National
Health and Nutrition Examination Survey
(NHANES).
To predict the dose from people's arsenic
consumption, SHEDS-Dietary estimates were
linked to the MENTOR-3P modeling system.
MENTOR-3P stands for Modeling ENvironment
for TOtal Risk with Physiologically Based
Pharmacokinetic Modeling for Populations.
Linking SHEDS' estimates of exposure and
ingestion, with MENTOR-3P's estimates of
internal dose, allowed researchers to
understand the variability between and
among individuals, and equipped EPA with the
ability to develop arsenic dose predictions
U.S. Environmental Protection Agency
Office of Research and Development
-------
that could be compared to biomarkers
measured in urine.
Conclusions and Impact
In this study, EPA scientists were able to use
the models to accurately predict dietary
exposures to arsenic. The study's modeled
estimates of arsenic exposures and doses for
the general U.S. population from food and
water were found to compare favorably with
the physical measurements and observations
from biomonitoring samples and dietary
studies.
Study results showed that the foods we eat
may be as or more important as sources of
The study is detailed in a November 2009
peer-reviewed journal article: Xue, J.,
Zartarian, V., Wang S-W, Liu, S.V.,
Georgopoulos P, 2009. Probabilistic Modeling
arsenic as the water we drink and cook with.
Food groups that have been reported to
contain arsenic include some fruits, fruit
juices, rice, beer, flour, corn, and wheat,
depending on the item and where it is grown.
The study found estimated arsenic exposures
from diet to be low — approximately one-fifth
the level set to protect consumers from the
effects of long-term, chronic exposure to
arsenic in drinking and cooking water alone.
This research strengthens EPA's capability to
accurately predict dietary exposures to
arsenic and other chemicals. It also expands
and enhances EPA's suite of predictive
modeling exposure assessment tools.
of Dietary Arsenic Exposure and Dose and
Evaluation with 2003-2004 NHANES Data.
Environ Health Perspectives, vol. 118, no. 3,
p. 345-350.
Technical Contact:
Roy Fortmann, Ph.D., Acting Director
Human Exposure and Atmospheric Sciences Division
National Exposure Research Laboratory (NERL)
Office of Research and Development
U.S. Environmental Protection Agency
fortmann. roy@epa. gov
Media Contact:
Emily Smith, Communications Director
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
smith, emily@epa.gov (919.541.5556)
May 2010
U.S. Environmental Protection Agency
Office of Research and Development
------- |