Jump to main content.

Water Quick Finder

 

Region 6 Interim Strategy:
Arsenic - Freshwater Human Health Criterion
for Fish Consumption

Map of EPA Region 6 Arkansas Louisiana New Mexico Oklahoma Texas
National Information
General Information
Regulatory History
Purpose of Arsenic Interim Strategy
Criteria Development for Office of Water
Bioconcentration Factor
Inorganic Arsenic and Organic Arsenic
EPA Region 6 Recommended Criteria
References
Appendix A: Agency RFD and Cancer Assessment Process and Integrated Risk Information System (IRIS)
Appendix B: Health Effects of Arsenic



General Information

Arsenic is a ubiquitous, naturally-occurring element. Increased levels of arsenic in water and soil can be found in certain areas of the country as a result of leaching from rock into ground water, and possible geothermal activity. In addition, nonferrous mining and smelting operations, refining operations and now discontinued pesticide manufacturing facilities may add to increased levels of arsenic in water. Only very limited quantities of arsenic-containing pesticides are still manufactured and used under strict limitations in the U.S. They represent a minimal source of arsenic exposure.

Arsenic may exist in both an organic and inorganic form, either in the trivalent or pentavalent state. The inorganic form is associated with the toxicity described in this risk characterization. Arsenic usually occurs in waters as inorganic oxides in the pentavalent form. Trivalent forms of arsenic (inorganic and organic) are more toxic to humans and aquatic organisms and are usually only present under anaerobic conditions.

oxidation
trivalent-------------->pentavalent
forms<--------------forms
reduction

Sources of human exposure to arsenic compounds may include air, soil, water and food. Dietary sources may include dairy products, meat, poultry and fish, fruits and vegetable and grain products. Water quality criteria may be established to protect consumption of water and/or fish.

Regulatory History

The Environmental Protection Agency (EPA) Office of Water has established guidance or regulations for arsenic under the Clean Water Act (CWA) and the Safe Drinking Water Act (SDWA). Under the CWA, a water quality criterion for arsenic - fish consumption was established at 0.14 ug/L in 1992 using the hazard assessment in EPA's Integrated Risk Information System (IRIS) database according to the current methodology for developing ambient water quality criteria for human health. The criterion for water + fish consumption is 0.018 ug/L. [For further information concerning IRIS, see Appendix A]. These arsenic water quality criteria represent a one in one million (10-6) cancer risk level for arsenic exposures.

EPA Office of Water also has a drinking water standard, or maximum contaminant level (MCL) issued under the SDWA of 50 ug/L. This level was developed in the mid-1940s by the Public Health Service and is not based on a risk assessment as we know it. Since the SDWA was passed in 1974, EPA Office of Water has been in the process of reevaluating the MCL. More recent data and analyses raise questions about the adequacy of the current MCL to protect human health "with a margin of safety." In addition, EPA Office of Water has been in the process of developing a drinking water health advisory, which is a non-regulatory guidance document, since 1985. Both the MCL revision and health advisory have yet to be completed due to continuing uncertainties associated with the risk. The Safe Drinking Water Act Amendments of 1996 mandates that a revised regulation for arsenic be proposed by January 1, 2000. The 1996 amendments also require that EPA develop a comprehensive plan for study of health risks to support the revised MCL within 180 days of enactment of the amendments.

Purpose of Arsenic Interim Strategy

As discussed above, there is currently no agreement between the recommended criteria developed pursuant to the CWA and the MCL developed pursuant to the SDWA. There are also questions concerning the toxicity database used in developing the cancer potency slope. While these issues need resolution and are being addressed by EPA Headquarters, the purpose of this strategy is to determine a reasonable approach to interpreting the arsenic criterion recommended to protect human health via the exposure route of consuming fish. The confusion over the appropriate criterion for freshwater fish consumption arose during the promulgation of the National Toxics Rule. In 1991, EPA proposed the National Toxics Rule in the Federal Register. EPA received two comments (# 68 and # 74) which expressed concern over the different forms of arsenic in fish and the degree of carcinogenicity of different forms of arsenic. In the final National Toxics Rule, issued December 22, 1992, a footnote was added to the human health criteria for arsenic (fish consumption only and water and fish consumption) that stated EPA's criteria apply only to the inorganic form of arsenic (EPA, 1992). Criteria based on inorganic arsenic are difficult to regulate and require additional resources for water quality analyses. States continue to use arsenic in criteria and implementation without recognizing that most of the arsenic is in the organic form in freshwater finfish. Since most of the arsenic present in finfish is in the organic form, criteria can be modified to reflect this.

While there may be several approaches to account for the different forms of arsenic in finfish, Region 6 recommends that States and Tribes adopt an arsenic criterion based upon the inorganic fraction that would be found in edible fish tissue (no Region 6 State or Tribe is in the National Toxics Rule for arsenic). This approach does not reevaluate the toxicity information, but looks at bioconcentration of inorganic arsenic from the water body into freshwater finfish. The strategy can be used by any State or Tribe in Region 6 that is adopting a risk-based criterion to protect human health from consuming inorganic arsenic found in finfish.

Criteria Development for Office of Water

The EPA Office of Water develops drinking water criteria or regulatory levels to protect human health under the CWA and the SDWA. Criteria developed under the CWA are developed to restore and maintain the chemical, physical, and biological integrity of all navigable waters. States may adopt these criteria in developing their State water quality standards or develop their own criteria.

Under the SDWA, criteria are referred to as Maximum Contaminant Level Goals (MCLGs) and are developed as non-enforceable health goals that are protective of adverse health effects and incorporate a margin of safety. Maximum Contaminant Levels (MCLs) are enforceable standards and are based on risk characterization. Factors such as analytical methods, technology and costs, economic impact and regulatory impact may also be used in the development of MCLs. MCLs apply to the concentration of a contaminant at the tap. States must adopt Federal drinking water standards (MCLs) or develop more stringent levels in a certain time period following promulgation of a drinking water standard.

The methodology to develop criteria for each program has evolved separately and incorporates different policies and assumptions. As a result, criteria for the same contaminant under the CWA and the SDWA can be different, even though they protect human health through drinking water exposures. Information on arsenic and the uncertainties with the human health arsenic criteria are found in Appendix B. The following equation was used to develop human health criteria for fish consumption under the CWA: 

	Criterion =     RF x BW  
                     q1*[BCF x FC]

	where:	RF  = risk factor (dimensionless)
		BW  = body weight (kg)
		q1* = cancer potency factor (mg/kg/day)-1	
		BCF = bioconcentration factor (L/kg)
		FC  = fish consumption rate (kg/day)

Bioconcentration Factor

The amount of pollutant that will accumulate in fish/shellfish is important to estimate since this route of exposure is potentially significant to human populations. Bioaccumulation occurs both through uptake across the gill membranes and other external body surfaces (bioconcentration) and through ingestion of contaminated food (biomagnification). Arsenic does not appear to progressively accumulate through the food chain (Callahan et al., 1979 and EPA, 1982, 1983 in ATSDR, 1993). EPA generally recommends that fish fillets be used to measure chemical contamination as this is portion consumed by humans (EPA, 1989). In cases where whole fish is consumed by certain subpopulations or where toxicity to wildlife is of concern, measurements from whole fish are more appropriate (EPA, 1995).

EPA's current bioconcentration factor (BCF) for arsenic is found in Ambient Water Quality Criteria for Arsenic (EPA, 1980). The BCF was calculated from the geometric mean (weighted with consumption rates) of two species. Data from the eastern oyster (BCF=350, 112 day test) and bluegill (BCF=4, 28 day test) resulted in a BCF for arsenic of 44. The criteria document also stated that BCFs of 0 were obtained from tests with rainbow trout. The data for the rainbow trout was not used in the calculation of the BCF. EPA Region 6 believes that the use of the eastern oyster data most likely overestimates the health risks associated with freshwater finfish consumption.

The BCF from the bluegill test was obtained from whole body measurements, while the BCF for the eastern oyster was measured using soft parts. BCFs for muscle tissue, the edible portion, should be lower than results obtained with whole fish (Stephan, 1993). Literature also supports this theory. Azcue and Dixon (1994) measured arsenic in four tissues of the rock bass. The highest concentration was found in bone and scales, followed by (in decreasing concentration), intestines and contents, muscle and liver. A BCF of 0.71 was calculated using the mean concentration for muscle tissue (0.04 mg/kg) divided by the concentration in water (0.056 mg/L). In comparison, a BCF of 2.3 was calculated from the mean whole-body concentration of rock bass (0.128 mg/kg). So, the BCF based on the whole fish tissue was 3 times greater than the BCF based on muscle tissue.

A draft version of the Great Lakes Initiative proposed a BCF of 1 (Stephan, 1993). The Final Water Quality Guidance for the Great Lakes Initiative, Federal Register March 23, 1995, pp 15366 - 15425, did not contain human health criteria for arsenic. Additional information presented below indicates that the bioconcentration factor for arsenic in freshwater fish is considerably lower than the BCF of 44.

States and Tribes may wish to use data from whole body measurements for derivation of BAF rather than fish fillets if subpopulations exist which consume skin, heads or other parts of the fish. Differences in cooking preparation may affect the bioavailablity of contaminents.

A water quality study in the Middle Rio Grande is currently underway with the City of Albuquerque, the Pueblo of Isleta, the New Mexico Environment Department, the U.S. Geological Survey and EPA Region 6. One of the objectives of the project is to calculate a site-specific bio-accumulation factor for arsenic based on fish species in the middle Rio Grande. A joint report from the study cooperators is scheduled to be released by the participants in 1997. EPA Region 6's Interim Arsenic Strategy may be amended following completion of this study. In the meantime, Region 6 recommends that a BCF of 1, as proposed during the Great Lakes Initiative, be used in the calculation of human health criteria.

Inorganic Arsenic and Organic Arsenic

As discussed under General Information, arsenic may be present in surface and groundwater as inorganic or organic compounds. Fish tissue also contains inorganic and organic compounds. Most analyses of arsenic related to fish consumption have focused on marine organisms and have determined that arsenic in marine organisms is primarily present as organic compounds (arsenobetaine and arsenocholine, in certain species). Various articles have noted that absence of arsenic related food poisonings from seafood throughout history (Kaise et al. 1985, Yamauchi et al., 1986, Edmonds and Francesconi, 1993). Several studies have measured inorganic arsenic in freshwater fish and other food groups and were reviewed during development of the Interim Strategy.

The organic forms are considered much less toxic, but data are insufficient to characterize the risk. Edmonds and Francesconi (1993) reported that, to the best of their knowledge, there is no evidence to indicate the demethylation of arsenic (organic inorganic) in animals. Several studies have examined the effects of organic arsenic in animals and humans. Four articles are summarized below.

The joint study of the Middle Rio Grande is measuring inorganic and organic arsenic in fish. The results of these analyses may also be used to amend EPA Region 6's Strategy. Although most literature reported inorganic arsenic as less than 10%, the data was either unpublished or did not list individual species. Therefore, EPA Region 6 recommends that a conservative estimate of 30% inorganic arsenic from Lawrence et. al (1986), be used in the calculation of the arsenic human health criterion.

EPA Region 6 Recommended Criteria

Based on the above information, EPA Region 6 recommends that human health criterion for freshwater fish consumption be calculated by the formula presented below. The criterion can be further modified by changing the assumed values for body weight, fish consumption, or risk factor as designated in State or Tribal Water Quality Standards. The value of 6.5 g/day for fish consumption is based on market survey data gathered in 1973-1974. EPA is evaluating more recent surveys for revision of the default value. Guidance on developing site-specific surveys to determine fish consumption is available from EPA. 

	Criterion =       RF x BW
            	      q1*[BCF x FC x I]

    0.0205 mg/L  =          10-6 x 70 kg
                  1.75 (mg/kg/day)-1[1 L/kg x 0.0065 kg/day x .30]
			
          or
		  
     20.5 ug/L

  where:	10-6		 RF  = risk factor
	   	70 kg	         BW  = body weight 
	     	1.75 mg/kg/day	 q1* = cancer potency factor
		1 L/kg		 BCF = bioconcentration factor
		0.0065 kg/day    FC  = fish consumption rate
		0.30	 	 I   = percent of total 
                           arsenic in fish tissue 
                           that is inorganic

References

Agency for Toxic Substances and Disease Registry (ATSDR). 1993. Toxicological Profile for Arsenic. Atlanta, GA. U.S. Department of Health and Human Services, Public Health Services.

Azcue, J.M. and D.G. Dixon. 1994. Effects of Past Mining Activities on the Arsenic Concentration in Fish from Moira Lake, Ontario. J. Great Lakes Res. 20(4):717-724.

Borum, D.R. and C.O. Abernathy. 1994. Human Oral Exposure to Inorganic Arsenic. Arsenic Exposure and Health. Northwood, England. W.R. Chappell, C.O. Abernathy and C.R. Cothern, eds. Science and Technology Letters. (Chapter 2).

Brown, R.M., D. Newton, C.J. Pickford and J.C. Sherlock. 1990. Human Metabolism of Arsenobetaine Ingested with Fish. Human & Experimental Toxicology. 9:41-46.

Callahan M.A., M.W. Slimak, and N.W. Gabel, et al. 1979. Water-related Environmental Fate of 129 Priority Pollutants. Vol I. Introduction and Technical Background, Metals and Inorganics, Pesticides and PCBs. Washington D.C. Report to U.S. Environmental Protection Agency, Office of Water Planning and Standards, by Versar Incorporated, Springfield, VA. EPA 440/4-79-029a. Cited in ATSDR, 1993.

Cullen, W.R. and K.J. Reimer. 1989. Arsenic Speciation in the Environment. Chem. Rev. 89:713-764.

Edmonds, J.S. and K.A. Francesconi. 1993. Arsenic in Seafoods: Human Health Aspects and Regulations. Marine Pollution Bulletin. 26(12):663-674.

EPA (Environmental Protection Agency). 1980. Ambient Water Quality Criteria for Arsenic. Washington D.C. U.S. Environmental Protection Agency, Office of Water Regulations and Standards. EPA 440/5-80-021.

- 1982. An Exposure and Risk Assessment for Arsenic. Washington D.C. U.S. Environmental Protection Agency, Office of Water Regulations and Standards. EPA 440/4-85-005. Cited in ATSDR, 1993.

- 1983. Treatability Manual. Vol. I. Treatability data. Washington D.C. U.S. Environmental Protection Agency, Office of Research and Development, I.4.2.1-I.4.2.6. EPA-600/2-82-001a. Cited in ATSDR, 1993.

- 1985. Ambient Water Quality Criteria for Arsenic. Washington D.C. U.S. Environmental Protection Agency, Office of Water Regulations and Standards. EPA 440/5-84-033.

- 1988. Special Report on Ingested Inorganic Arsenic. Skin Cancer; Nutritional Essentiality. Washington, D.C. Risk Assessment Forum, U.S. Environmental Protection Agency, EPA 625/3-87/013. Cited in Borum and Abernathy, 1994.

- 1989. Assessing Human Health Risks from Chemically Contaminated Fish and Shellfish: A Guidance Manual. Washington D.C. U.S. Environmental Protection Agency, Office of Water Regulations and Standards. EPA 503/8-89-002.

- 1992. Water Quality Standards; Establishment of Numeric Criteria for Priority Toxic Pollutants; States' Compliance Final Rule. Washington, D.C. U.S. Environmental Protection Agency, Office of Science and Technology. Federal Register 57:60848, December 22, 1992.

- 1995. Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories. Volume 1 - Fish Sampling and Analysis. 2nd ed. Washington D.C. U.S. Environmental Protection Agency, Office of Science and Technology. EPA 823-R-95-007.

Kaise, T., S. Watanabe, and K. Itoh. 1985. The Acute Toxicity of Arsenobetaine. Chemosphere 14(9):1327-1332.

Lacayo, M.L., A. Cruz, S. Calero, J. Lacayo and I. Fomsgaard. 1992. Total Arsenic in Water Fish and Sediments from Lake Xolotlán, Managua, Nicaragua. Bull. Environ. Contam. Toxicol. 49:463-470.

Lawrence, J.F., P. Michalik, G. Tam, and H.B.S. Conacher. 1986. Identification of Arsenobetaine and Arsenocholine in Canadian Fish and Shellfish by High-Performance Liquid Chromatography with Atomic Absorption Detection and Confirmation by Fast Atom Bombardment Mass Spectrometry. J. Agric. Food Chem. 34:315-319.

Mayer, E.R., W. Kosmus, H. Pogglitsch, D. Mayer, and W. Beyer. 1993. Essential Trace Elements in Humans. Serum Arsenic Concentrations in Comparison to Healthy Controls. Biol. Trace Elem. Res. 37:27-38.

NRC. National Research Council. 1989. Recommended Dietary Allowances. 10th ed. Report of the Food and Nutrition Board. National Academy of Sciences, National Press Academy.

Stephan, C.E. 1993. Draft - Derivation of Proposed Human Health and Wildlife Bioaccumulation Factors for the Great Lakes Initiative. Duluth, MN. U.S. Environmental Protection Agency, Office of Research and Development.

Tam, G.K.H., S.M. Charbonneau, F. Bryce and E. Sandi. 1982 Excretion of a Single Oral Dose of Fish-Arsenic in Man. Bull. Environm. Contam. Toxicol. 28:669-673.

Tetra Tech. 1996. Assessing Human Health Risks from Chemically Contaminated Fish in the Lower Columbia River. Report prepared for the Lower Columbia River Bi-State Program. Redmond, WA. Cited in EPA, 1995.

Tseng, W.P., H.M. Chu, S.W. How, J.M. Fom, C.S. Lin, and S. Yeh. 1968. Prevalence of Skin Cancer in an Endemic Area of Chronic Arsenism in Taiwan. J. Natl. Cancer Inst. 40:453-463.

Uthus, E.O. 1994. Estimation of Safe and Adequate Daily Intake for Arsenic. Risk Assessment of Essential Elements. Eds W. Mertz, CO Abernathy, and SS Olin, ILSI press, Washington, DC, pp 273-282.

Weiler, R.R. 1987. Unpublished data. Ministry of the Environment, Report No. 87-48-45000-057, Toronto, Ontario. Cited in EPA, 1988.

Yamauchi, H., T. Kaise, and Y. Yamamura. 1986. Metabolism and Excretion of Orally Administered Arsenobetaine in the Hamster. Bull. Environ. Contam. Toxicol. 36:350-355.

Appendix A: Agency RFD and Cancer Assessment Process, and Integrated Risk Information System (IRIS)

For chemicals suspected of being carcinogenic to humans, the risk assessment consists of the weight of evidence of carcinogenicity in humans, using bioassays in animals and human epidemiology studies and other pertinent information such as mutagenicity studies. The objectives of the assessment are to determine the level or strength of evidence that the substance is a carcinogen and to provide an upperbound estimate of the possible risk of human exposure to the substance in drinking water. Substances are placed in one of five categories reflecting weight of evidence:

Group A - Human carcinogen based on sufficient information in human studies;

Group B - Probable human carcinogen based on limited evidence in human (Group B1), or sufficient evidence in animals in the absence of human information (Group B2);

Group C - Possible human carcinogen based on limited evidence of carcinogenicity in animals in the absence of human data.

Group D - Not classifiable based on lack of data or inadequate evidence in animal studies;

Group E - No evidence of carcinogenicity in humans.

If toxicological evidence leads to the classification of a contaminant as a carcinogen, mathematical models are used to estimate an upperbound excess cancer risk associated with lifetime exposure through drinking water. Several models are available to extrapolate data. EPA generally uses the linearized multistage (LMS) model which fits linear dose-response curves to low doses. It is consistent with a no-threshold model of carcinogenesis, ie., exposure to even a very small amount of the substance theoretically produces a finite increased risk of cancer. The LMS model uses dose response data from an appropriate study to calculate a carcinogenic potency factor (q1*) which is then used to determine concentrations of water that are associated with the theoretical upperbound excess lifetime cancer risk of one in ten thousand (10-4) to one in one million (10-6).

The risk assessment for noncancer health effects can be characterized by the reference dose (RfD). The oral RfD (in mg/kg/d) is an estimate, with an uncertainty spanning perhaps an order of magnitude, of a daily exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious health effects during a lifetime. The RfD is derived from a no- or lowest observed adverse effect level (called the NOAEL or LOAEL, respectively) that has been identified from appropriate studies in humans or animals. The NOAEL or LOAEL is divided by an uncertainty factor(s) to derive the RfD. Although the RfD is expressed as a single number, it is actually a range with an inherent uncertainty of an order of magnitude.

Uncertainty factors are used to estimate the comparable no- effect level for a large heterogeneous human population. The use of uncertainty factors accounts for several data gaps including intra- and interspecies differences in response to toxicity, the small number of animals tested compared to the size of the population, sensitive subpopulations and the possibility of synergistic action between chemicals. Uncertainty factors may vary from 1 to 30,000. There is very little confidence in an assessment where the uncertainty factor is greater than 3,000.

Appendix B: Health Effects of Arsenic

In evaluating the risks posed by arsenic exposure, risk analysts are fortunate to have a large human database. Humans appear to be more sensitive to the carcinogenic effects of arsenic than animals. Ingestion of arsenic contaminated water and soil can result in both carcinogenic and noncarcinogenic health effects.

Carcinogenic Effects:

A number of epidemiology studies have been conducted which report an increased incidence of skin cancer in exposed populations. Studies have been conducted in Taiwan, Mexico, Chile, Hungary, and Argentina. The studies conducted outside of Taiwan evaluated populations much smaller in size than the Taiwan study. The Taiwan study also reported a weak association between arsenic ingestion and increased incidence of internal cancers such as cancers of the liver, bladder, kidney, and lung. Studies conducted in the US have not demonstrated an association between arsenic in drinking water and skin cancer. While there was no demonstrated elevated cancer incidence in these U.S. populations, the population sizes studied were too small and/or the lengths of exposure time too short to expect to see an effect.

The largest epidemiology study is the Taiwan study (Tseng etc al., 1968), which also serves as the basis for the EPA risk assessments for arsenic discussed below. In this study, an increased prevalence of skin cancer was observed in 40,000 Taiwanese consuming arsenic contaminated drinking water (0.05 to >1 mg/L) from artesian wells as compared with 7,500 residents of Taiwan or Matsu consuming "arsenic - free" (0-0.017 mg/L) drinking water. The number of people with skin cancer was reported to increase in association with increasing concentrations in the water they consumed.

Related to carcinogenicity, arsenic is a genotoxic agent that induces chromosomal aberrations, micronuclei and sister chromatid exchange in mammalian cells and neoplastic transformations in Syrian hamster embryo cells. Arsenic does not cause cancer in experimental animals. Thus, no animal model is available for studying arsenic-induced cancer.

EPA and other organizations such as the International Agency for Research on Cancer (IARC) have classified arsenic as a human carcinogen. (See Appendix A for further discussion of cancer hazard assessment.) This classification is based on sufficient evidence of carcinogenicity from human data involving occupational and drinking water exposures. EPA used the evidence of skin cancer reported in the Taiwan study as the basis for the arsenic hazard and dose response assessment. Using a time- and dose- dependent multistage model which assumes that any exposure to a compound such as arsenic could result in a cancer response, the cancer potency (q1*) estimated for ingested arsenic is 1.75 mg/kg/day. This potency can then be used to estimate concentrations of arsenic in water associated with an increased cancer risk of one in ten thousand (10-4) to one in one million (10-6):

10-4 = 1.8 ug/L
10-5 = 0.18 ug/L
10-6 = 0.018 ug/L

Although this assessment has been adopted by the Agency and extensively peer reviewed, there are a number of issues and uncertainties that make the skin cancer risk assessment conclusions incomplete and uncertain (see Issues and Uncertainties below) and thus, complicate risk management decisions.

Noncarcinogenic Effects:

In addition to the cancer effects observed in epidemiological studies, arsenic exposure has also been reported to result in adverse health effects other than cancer in human and animal studies. Dermal changes including variations in skin pigments, thickening of the skin (e.g.,hyperkeratosis)and ulcerations, peripheral neurotoxicity, peripheral vascular effects, cardiac effects, gastrointestinal and liver effects, and diabetes have been observed.

Acute high-dose oral exposure (e.g., poisonings) to arsenic typically leads to gastrointestinal irritation accompanied by difficulty in swallowing, thirst, abnormally low blood pressure, and convulsions.

The most common signs of long-term arsenic exposure from drinking water are dermal changes such as variations in skin pigments, hyperkeratoses, and ulcerations. Blackfoot disease, a peripheral vascular disease, has been associated with chronic arsenic exposure in Taiwan; however, the exact etiologic mechanism is unknown. Studies in Canada and US report neurological effects after chronic exposure from drinking water containing arsenic. Enlargement of the liver was observed in populations in India that were exposed to arsenic in drinking water. An association between ingested arsenic and ischemic heart disease (low oxygen supply to the heart muscle)and diabetes mellitus (hyperglycemia or diabetes) have been reported in the area of Taiwan where Blackfoot disease is endemic.

There is no clear consensus among Agency scientists on the reference dose (RfD) workgroup to estimate and oral RfD. Applying the Agency's RfD methodology (see Appendix A for further discussion on RfDs), the EPA Risk Assessment Council determined that strong scientific arguments could be made for various values within a factor of 2 or 3 of the currently recommended RfD value, i.e., 0.1 to 0.8 ug/kg/d. The Taiwan data are used to estimate the RfD of 0.3 ug/kg/d, at which no adverse effects (hyperpigmentation, keratosis and possible vascular complications) were observed. It should be noted, however, that the RfD methodology, by definition, assumes that for each estimate there is an inherent uncertainty spanning perhaps an order of magnitude. The arsenic summary on IRIS is the first to articulate this point.

Issues and Uncertainties:

As noted above, there are a number of issues and uncertainties associated with the cancer and noncancer health effects and subsequent risk assessment for arsenic. Some of these relate to arsenic in general and some are specific to use of the Taiwan study to quantify the risk.

General:

1. Arsenic may be either inorganic or organic. The inorganic form is associated with the health effects described above. The organic form, which is often found in fish and some seafoods, appears to be less toxic.

2. The typical US diet contains 10 - 20 ug inorganic arsenic/day.

3. Arguments have been made that arsenic is an essential element (NRC, 1989) with a nutritional requirement of 12 - 25 ug/d (Uthus, 1994). Studies in rats, hamsters, minipigs, goats and chicks indicate that it is essential for normal growth and development.

4. Recent human data from dialysis patients suggest that patients with lower serum levels of arsenic are more susceptible to CNS and vascular disease and cancer (Mayer et al., 1993). However, more studies are needed in this area.

5. Arsenic can be detoxified at low exposure levels. Genetic, dietary or other lifestyle factors may enhance or inhibit potential arsenic detoxification processes in humans. Preliminary studies in Mexico indicate that people chronically exposed to high concentrations of arsenic (>400 ug/L) will metabolize it differently than those exposed to low levels (<20 ug/L).

6. The available ecologic-type epidemiology studies do not determine a cause and effect relationship between arsenic exposure in drinking water and skin cancer. In addition, lack of an animal model also precludes use of a controlled laboratory experiment to make such a determination.

7. Unlike many other types of cancer, skin cancer is treatable.

8. Considering the genotoxicity data (chromosomal and DNA changes), exposure information, lack of animal cancer data and incidence of skin cancer, the multistage model may not be the appropriate model for estimating cancer potency. Available data can be used to support a plausible threshold (there is a dose below which a cancer effect would not occur) model.

Taiwan Study:

1. Arsenic concentrations consumed by Taiwanese are not known. In the Taiwan study, arsenic concentrations are determined by village, some of which contained more than one well. One measurement was taken from each well, with the results averaged to estimate arsenic concentrations for the village. Thus, individual doses can not be determined.

2. There is some question about the analytical method used in the Taiwan study to measure arsenic levels in water. A recent analysis of the method used (Nataleson, spectrophotometry with molybdenum blue) demonstrated that it is only sensitive in the 50 to 100 ug/L. Reported concentrations less than that are not accurate. This directly impacts the risk estimates discussed above.

3. The water consumed in Taiwan contained other possible cancer causing substances in addition to arsenic.

4. The nutritional status of the Taiwanese is not known. In general, their diets tended to be deficient in protein. This may compromise their ability to detoxify arsenic. In addition, there were other sources of arsenic in the Taiwanese diet such as sweet potatoes and rice. There have been studies linking poor nutrition, eg, sweet potato consumption, with increased risk from drinking water containing high arsenic levels.

5. EPA scientists have attempted for more than a decade to obtain the original data to assess quality and accuracy of the reported results with no success. Thus, reliance is placed on published results only.

History of Risk Management Guidance:

Risk assessors and managers have been evaluating the problem of arsenic for a long time. In the Office of Water, this issue has continued for over 20 years. In 1988, the Agency's Risk Assessment Forum published a book "Special Report on Ingested Inorganic Arsenic: Skin Cancer; Nutritional Essentiality." This book articulated Agency risk assessment conclusions concerning carcinogenicity and essentiality. It was reviewed by the Agency's Science Advisory Board and endorsed by the Risk Assessment Council. As a result of the issues and uncertainties associated with arsenic risk estimation as discussed in the book, Lee Thomas, EPA Administrator, issues guidance to risk managers on use of the cancer risk estimates to assist with decision-making. Concerning the cancer hazard assessment, the memo stated:

When the Risk Assessment Council reached a decision on the RfD, guidance to risk managers state "Risk managers should recognize the considerable flexibility afforded them in formulating regulatory decisions when uncertainty and lack of clear consensus are taken into account."

Back to Water Quality Standards Program Page

Local Navigation

Jump to main content.