In Vivo Relative Oral Bioavailability of Arsenic from CCA-Affected Soils in the Primate
Dr. Stephen M. Roberts, University of Florida, Gainesville, FL      
Raymond J. Bergeron, University of Florida, Gainesville, FL       

USEPA SHEDS Model: Methodology for Exposure Assessment for Wood Preservatives
Haluk Özkaynak, U.S. EPA, Office of Research and Development, Washington, DC
Valerie Zartarian, U.S. EPA, Office of Research and Development, Research Triangle Park, NC
Jianping Xue, U.S. EPA, Office of Research and Development, Research Triangle Park, NC  
Winston Dang, U.S.EPA, Office of Pesticide Programs,  Washington, DC

Exposure and Risk Assessment for Arsenic from CCA-Treated Wood Playground Structures
Dr. Kristina Hatlelid, U.S. Consumer Product Safety Commission, Washington, DC
David Cobb, U.S. Consumer Product Safety Commission, Washington, DC
Dwayne Davis, U.S. Consumer Product Safety Commission, Washington, DC
Mark S. Levenson, U.S. Consumer Product Safety Commission, Washington, DC
Jonathan D. Midgett, U.S. Consumer Product Safety Commission, Washington, DC
Treye A. Thomas, U.S. Consumer Product Safety Commission, Washington, DC
Patricia M. Bittner, U.S. Consumer Product Safety Commission, Washington, DC

Comparison of a Probabilistic/Mechanistic (SHEDS) Approach to a Deterministic/Empirical Approach for Evaluating Exposures to CCA-Treated Wood
Dr. Barbara D. Beck, Gradient Corporation, Cambridge, MA
Catherine Petito Boyce, Gradient Corporation, Seattle, WA
Eric M. Dube, Gradient Corporation, Cambridge, MA

Background Inorganic Arsenic Exposures in Children     
Dr. Joyce S. Tsuji, E^xponent^®, Bellevue, WA    

Relationship Between Childhood Arsenic Exposures and Cancer Risks in the U.S. and Findings of an Ecological Arsenic Health Effects Study
Floyd Frost, Ph.D., Lovelace Respiratory Research Institute, Albuquerque, NM
Kristine Tollestrup, Ph.D, University of New Mexico, Dept. of Family Medicine
Lucy Harter, Washington State Dept. of Health
Melissa Roberts, Lovelace Respiratory Research Institute, Albuquerque, NM

Arsenic Epidemiology & Cancer Slope Factor     

Dr. Steven H. Lamm, Consultants in Epidemiology and Occupational Health, Inc., Washington, D.C.

In Vivo Relative Oral Bioavailability of Arsenic from CCA-Affected Soils in the Primate

Stephen M. Roberts, Center for Environmental & Human Toxicology, University of Florida, Box 110885, Gainesville, FL 32611, Tel: 352 392-4700 ext. 5500, Fax: 352 392-4707
Raymond J. Bergeron, Department of Medicinal Chemistry, University of Florida, Box 100485, Gainesville, FL 32610, Tel: 325 846-1956

The relative oral bioavailability of arsenic in soil from a former CCA wood treatment facility was measured in the Cebus monkey.  In preliminary experiments, sodium arsenate in solution was administered intravenously and orally to five male Cebus monkeys, and blood, urine, and feces were collected.  The disappearance of arsenic from blood following an intravenous dose in the monkeys was similar to that reported previously in humans.  Further, the fraction of dose excreted in urine and feces after both intravenous and oral administration were comparable to values reported in humans.  Relative bioavailability of arsenic from soil was measured by comparing arsenic excretion following a single oral dose of soil with arsenic excretion following a similar oral arsenic dose as sodium arsenate in water.  The relative bioavailability of arsenic from the soil sample was found to be 16.3 ± 6.5 % (mean ± SD) in five monkeys.  This research was supported through a contract with the Florida Department of Environmental Protection.

USEPA SHEDS Model: Methodology for Exposure Assessment for Wood Preservatives

Haluk Özkaynak, U.S. EPA, Office of Research and Development, Ariel Rios Building, 1200 Pennsylvania Avenue, NW, MC-8601 D (room 400W22), Washington, DC 20460, Tel: 202-564-1531, Fax: 202-565-0075
Valerie Zartarian, U.S. EPA, Office of Research and Development, 109 TW Alexander Drive, Mail Drop E205-02, Research Triangle Park, NC 27711, Tel: 617-918-1541, Fax: 617-918-0541
Jianping Xue, U.S. EPA, Office of Research and Development, 109 TW Alexander Drive, Mail Drop E205-02, Research Triangle Park, NC 27711, Tel: 919-541-7962, Fax: 919-541-9444  
Winston Dang, U.S.EPA, Office of Pesticide Programs, Office of Prevention, Pesticides and Toxic Substances, MC-7510C, 1200 Pennsylvania Avenue, N.W., Washington, DC 20460-0001, Tel: 703-308-6216  Fax: 703-308-6466                         

A physically-based, Monte Carlo probabilistic model (SHEDS-Wood: Stochastic Human Exposure and Dose Simulation model for wood preservatives) has been applied to assess the exposure and dose of children to arsenic (As) and chromium (Cr) from contact with chromated copper arsenate (CCA)-treated playsets and residential decks.  Short-term (for As and Cr), intermediate-term (for As and Cr), and lifetime (for As only) absorbed doses are estimated for: dermal contact with playset or deck residues; dermal contact with soil concentrations around treated playsets or decks; ingestion of CCA-containing soil near treated playsets or decks; and ingestion of wood residues via the hand-to-mouth pathway.

SHEDS-Wood calculates the predicted exposure and dose to As and Cr using age and gender representative time-location-activity diaries from EPA’s Consolidated Human Activity Database (CHAD). Based on user-specified inputs, exposure days and exposure events within a day are simulated.  The time series of exposure and dose are computed using pathway-specific exposure equations and the real-time diary activities for which a contact event is possible.  Model inputs, represented as analytical distributions (e.g., lognormal, beta) include: fraction of outdoor time and days per year a child plays on/around playsets and decks; As and Cr residue and soil concentrations on or near CCA treated playsets or decks; and various exposure factors such as residue-to-skin transfer efficiencies, soil-to-skin adherence factor, saliva and bathing removal efficiency, daily incidental soil ingestion rate, fraction of hand skin surface area contacting soil, frequency of hand to mouth activity, maximum dermal loading, and dermal and GI absorption rates.  Model results for the simulated population are analyzed to determine the dominant pathways influencing predicted exposures and dose. Both deterministic and statistical methods are used to assess the sensitivity of results to key input variables. Uncertainty analyses are performed using the 2^nd Stage Monte-Carlo simulation results and a nonparametric bootstrap methodology.

Disclaimer: This work has been funded wholly by the United States Environmental Protection Agency.  It has been subjected to Agency review and approved for publication.

Exposure and Risk Assessment for Arsenic from CCA-Treated Wood Playground Structures

Kristina M. Hatlelid, Consumer Product Safety Commission, Washington, D.C. 20207, Tel: 301-504-7254, Fax: 301-504-0079
David Cobb, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel: 301-424-6421 x106, Fax: 301-413-0000
Dwayne Davis, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel: 301-424-6421 x107, Fax: 301-413-0000
Mark S. Levenson, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel: 301-504-7408, Fax: 301-504-0081
Jonathan D. Midgett, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel:  301-504-7692, Fax: 301-504-0124
Treye A. Thomas, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel:  301-504-7738, Fax: 301-504-0079
Patricia M. Bittner, U.S. Consumer Product Safety Commission, Washington, D.C. 20207, Tel:  301-504-7263, Fax: 301-504-0079

In June, 2001, the U.S. Consumer Product Safety Commission (CPSC) docketed a petition from the Environmental Working Group and the Healthy Building Network that requested that the Commission enact an immediate ban of chromated copper arsenate (CCA)-treated wood for use in playground equipment.

CPSC staff completed toxicological reviews of CCA components, chromium, copper, and arsenic, but chose to focus on arsenic because it is the most potent of the three chemicals. The staff considers that the principal exposure to arsenic from CCA-treated wood occurs through transfer of wood surface arsenic residues to a child’s hands and fingers and subsequent direct (e.g., thumb-sucking) and indirect (e.g., handling of toys or food) hand-to-mouth transfer of the residues.

CPSC staff conducted laboratory and field studies to estimate the amount of arsenic to which a child might be exposed during a typical playtime on CCA-treated wood playsets. The staff measured the dislodgeable arsenic levels on the surface of CCA-treated wood structures in the Washington, D.C. metropolitan area using 8 CCA-treated wood decks and 12 CCA-treated wood playsets, representing a variety of ages, wood treatments, and manufacturers. The study used adult volunteers using their bare hands, as well as surrogate materials, e.g., polyester cloth, that could be used in place of the bare hands, to rub the surface of the wood structures.

Based on the field studies, the staff estimated that the mean dislodgeable arsenic levels from handling CCA-treated wood playground equipment is about 7.6 micrograms. Using this information and other assumptions in a deterministic risk assessment model, the CPSC staff concluded that a young child who plays primarily on CCA-treated wood playground structures in early childhood has an increased lifetime risk of 2 to 100 per million of developing lung or bladder cancer. This is an increased risk above the risk of cancer due to other factors during one’s lifetime.  (The opinions expressed by the authors do not necessarily represent the views of the Commission.  As this abstract was prepared by the authors in their official capacity, this abstract is in the public domain and may be freely copied or reprinted.)

Comparison of a Probabilistic/Mechanistic (SHEDS) Approach to a Deterministic/Empirical Approach for Evaluating Exposures to CCA-Treated Wood

Barbara D. Beck, Gradient Corporation, Cambridge, MA
Catherine Petito Boyce, Gradient Corporation, Seattle, WA
Eric M. Dube, Gradient Corporation, Cambridge, MA

Quantifying potential exposures and risks to inorganic arsenic (As_i) releases from wood treated with chromated copper arsenate (CCA) presents several risk assessment challenges.  For example, standardized algorithms are not available in typical risk assessment guidance for quantifying transfer of dislodgeable arsenic from wood surfaces into the body.  Consequently, scenario-specific assumptions for certain exposure parameters are  being refined, e.g., by conducting relevant scientific studies.  This presentation compares two methodologies  for conducting human health risk assessments (HHRAs) for As_i from treated wood: (1) a probabilistic, mechanistic methodology as exemplified by the U.S. Environmental Protection Agency’s Stochastic Human Exposure and Dose Simulation model as applied to a wood preservative exposure scenario (SHEDS-Wood) and (2) a deterministic, empirical methodology with a focused sensitivity analysis (DEM/FS), as exemplified by recent HHRA’s conducted by the U.S. Consumer Product Safety Commission (CPSC) and Gradient Corporation.  Strengths and limitations of each model are described, and data needs are highlighted for each approach.  Among other observations, the SHEDS-Wood approach allows specific activity patterns to be incorporated in the model and more explicitly characterizes variability and uncertainty in the model results.  By contrast, the lack of available data for many of the necessary input assumptions and for validating the model results adds uncertainty when interpreting estimates derived using the SHEDS-Wood model.  The DEM/FS approach relies on an empirical approach based on soil ingestion studies to estimate intake of surface residues from CCA-treated wood, an approach which provides a benchmark for evaluating the validity of estimates for this pathway.  The focused sensitivity analysis component of the DEM/FS approach allows analysts to readily identify critical assumptions and to assess the conservatism of the analyses.  The DEM/FS approach cannot directly incorporate information regarding certain behavioral variables, however, such as variations in numbers of hand-to-mouth contacts, and does not provide quantitative information regarding the distribution of potential risk levels.  Certain critical data needs (e.g., loadings of surface residues) are shared by both methods, whereas other data needs (e.g., the fraction of the hand inserted into the mouth during outdoor activities) are more relevant to one approach (i.e., the SHEDS methodology for this specific example).  The risk assessment implications of the model features and data needs are also discussed in this presentation.

Background Inorganic Arsenic Exposures in Children

Joyce S. Tsuji, Ph.D., DABT, Lisa Yost, MPH, DABT, Exponent, Bellevue, WA 
Leila Barraj, Ph.D, Exponent, Washington, DC

Understanding where arsenic exposures fall with respect to natural background is important for communicating the relative magnitude of arsenic exposure and risk in relation to everyday exposures.  Arsenic is naturally ubiquitous in the environment.  The primary background sources of inorganic arsenic to the general population are food, water, and to a lesser extent, soil.  Using reported inorganic arsenic data for food types and the USDA food intake survey for children ages 1–6 years, the probabilistically modeled average inorganic arsenic intake was 3.2 µg/day (95th percentile of 6.2 µg/day; 99th percentile of 9.5 µg/day).  Foods contributing the most inorganic arsenic were rice, other grains, and fruit.  One bowl of rice (1 cup cooked) contributes nearly 4 µg of arsenic.  Inorganic arsenic intake by Asian or other populations with much greater rice consumption compared with the typical U.S. population would be considerably higher (e.g., 14 µg/day for Japanese adults versus 3.2 µg/day in U.S. adults; dietary arsenic estimates are not available for Japanese children).  At an average arsenic level in drinking water in the U.S. around 1–2 µg/L, drinking water adds an additional 1–2 µg/day assuming 1 L/day consumption.  EPA’s new drinking water standard for arsenic (10 µg/L) would contribute 10 µg/day.  By comparison, ingestion of soil containing arsenic at a background level of 20 mg/kg would result in a reasonable maximum estimate of 2 µg/day (assuming 50 percent relative bioavailability or arsenic in soil) and an average that is below 1 µg/day.  Thus, total background arsenic intakes are available for placing arsenic exposures and risk in context with naturally occurring sources.

Relationship Between Childhood Arsenic Exposures and Cancer Risks in the U.S. and Findings of an Ecological Arsenic Health Effects Study

Floyd Frost, Ph.D., Lovelace Respiratory Research Institute
Kristine Tollestrup, Ph.D, University of New Mexico, Dept. of  Family Medicine
Lucy Harter, Washington State Dept. of Health
Melissa Roberts, Lovelace Respiratory Research Institute

Arsenic exposure studies conducted by the Washington State Department of Health (WSDOH) in 1974 found that children in the Ruston area near the ASARCO copper smelter excreted arsenic in the urine at approximately the same concentration as workers in the smelter. Urinary arsenic levels as high as 0.68 ppm were observed in these children in 1974.  WSDOH studies also found that: 1) younger children excreted higher levels of arsenic than older children; 2) the level of arsenic in the urine of children was proportional to the distance of the residence from the smelter; and 3) most of the exposure likely resulted from inhalation of particles.  Other studies found that urinary arsenic levels in children were lower during times of the day when the winds minimized distribution of arsenic from the smelter, as well as during a smelter strike.  These finding suggest that some of the arsenic exposure might also have been due to dust in the environment.

We conducted a cohort study to determine whether childhood exposure to arsenic increases overall mortality and mortality due to lung and bladder cancer.  Cohort members were children who lived within 2.5 miles of the ASARCO copper smelter and arsenic refinery for at least two years from 1907-1932.  The cohort included 1,827 boys and 1,305 girls identified from school census records.  Based on local school census records and the U.S. Bureau of  Census records, we determined the duration of residence at each location for the period 1907-1932.  Exposure intensity was computed as the total number of years the child lived at residences <1.0 mile from the smelter stack from 1907-1932.  Most of the arsenic exposures in this area were from low-level emission from the processing plant rather than the stack.  Deaths were identified from church or cemetery records, state death records and, for more recent years, from the National Death Index.  Overall follow-up was 58% for males and 48% for females.  A total of 712 males and 361 females were found to be deceased.  Of these 46 males and 16 females had died from lung cancer.  In only one exposure intensity group (10 or more years at <1.0 mile from the smelter) were Cox proportional hazards ratios significantly higher than 1.00:  all causes of death (1.52), ischemic heart disease (1.77), and external causes (1.93).   For girls, hazard ratios were not significantly elevated for any cause of death in any exposure intensity group.

We have also conducted a nationwide exposure assessment for drinking water arsenic in the U.S. and have related arsenic exposure levels to lung, bladder, liver and kidney cancer death rates for the U.S. for 1950-2000.  The analysis (MlwIN) relates exposure levels to outcomes, adjusting for socio-economic and employment characteristics of the county population.

Arsenic and the Cancer Slope Factor 

Steven H. Lamm, M.D., D.T.P.H., Consultants in Epidemiology and Occupational Health, Inc., Washington DC, Tel: 202-333-2364

Background: Arsenic is a strange carcinogen in that it is well accepted as a human carcinogen and shows little evidence of carcinogenicity in other species.  Inhaled arsenic clearly causes lung cancer, and ingested arsenic clearly causes skin cancer.  The standard risk model assumes that the study population is one population and that the risk, proportional to the dose, is the same for all people within that population.  Some analyses of the accepted studies suggest otherwise.

Materials and Methods:  A number of studies in the last ten years have suggested that arsenic exposure causes bladder cancer.  The major study for analysis has been the 42-village study of Wu et al. (1989) from the Blackfoot disease (BFD) endemic area of SW Taiwan.  The Morales et al. (2000) analysis has served as the basis of the quantitative risk analysis by both the EPA and the National Research Council.  Review of the well data for the villages reveals two populations of villages based on water source dependency [artesian vs. non-artesian].  Separate analyses have been carried out for each population. US male bladder cancer mortality rates have also been analyzed by median arsenic levels in drinking water supply.  

Results:  There is no change in bladder cancer mortality risk with increasing arsenic exposure for villages not dependent upon artesian wells.  There is a sharp dependency on arsenic level for villages dependent on artesian wells.  The x-intercept for the artesian well dependent villages is just above 200 ug/L.  These analyses are consistent with US data where no change in bladder cancer mortality risk with increasing arsenic exposure is seen for US counties over the range of 3 – 60 ug/L.

Conclusion: US and Taiwan data both show no increase in bladder cancer mortality rates with drinking water arsenic levels up to the 100s of ug/L.  The data are consistent with a risk model for arsenic as a high-dose carcinogen or as a co-carcinogen.  The two largest epidemiological studies show no basis for proposing a positive arsenic cancer slope factor at US exposure levels, i.e., below a number of hundreds of micrograms per liter. 

Top