Skip directly to search Skip directly to A to Z list Skip directly to site content


Arsenic Soil Clean-up Levels



In response to recommendations made in previous health consultations, the U.S. EnvironmentalProtection Agency Region 6 (EPA-6) contracted with the College of Veterinary Medicine at theUniversity of Missouri, Columbia (CVMUM) to assess the relative bioavailability of arsenic in soilfrom the El Paso remediation area. Using data from this study, scientists from the TexasCommission of Environmental Quality (TCEQ) and EPA-6, proposed a residential soil clean-uplevel of 46 milligrams per kilogram (mg/kg). The EPA has asked the Texas Department of Health(TDH) and the Agency for Toxic Substances and Disease Registry (ATSDR) for an independentassessment to evaluate whether the proposed residential surface soil clean-up level for arsenic would be protective of public health.


Arsenic is a naturally occurring element in the earth's crust and is usually found in combination withother elements. Arsenic compounds can be classified into three main groups: 1) inorganic arseniccompounds, 2) organic arsenic compounds, and 3) arsine gas. In the environment, arsenic is mostoften found as inorganic arsenic, which is formed when arsenic combines with other elements suchas oxygen, sulfur, and chlorine. Organic forms of arsenic, which result when arsenic combines withcarbon and hydrogen, are generally considered less toxic than the inorganic forms. Backgroundconcentrations of arsenic in soil range from 1 to 40 mg/kg with an average value of about 5 mg/kg(ATSDR, 2000).

Analysis of the toxic effects of inorganic arsenic in soil is complicated by several factors. First,arsenic can exist in different valence states. A number of studies have noted differences in therelative toxicity of these compounds, with trivalent arsenites tending to be somewhat more toxic thanthe pentavalent arsenates (Byron, 1967; Gaines, 1960; Maitani, 1987, Sardana, 1981; Whillhite,1981). For the purposes of this consultation, we have not emphasized the difference in toxicitybetween the arsenates and the arsenites because in most instances, this difference is reasonably small(about 2-3 fold) and is often within the bounds of uncertainty regarding the No Observed AdverseEffect Levels (NOAELs) or Lowest Observed Adverse Effect Levels (LOAELs). In addition, thedifferent forms of arsenic may be interconverted, both in the environment and in the body and as inmost cases involving human exposure, the precise chemical speciation will not always be known.

Second, although both inorganic and organic arsenic compounds in water generally are well absorbed(over 90 percent of an ingested dose of inorganic trivalent or pentavalent arsenic in water isabsorbed), most scientists would agree that arsenic in soil probably is not as well absorbed. Whilewe often do not know the actual bioavailability of arsenic in soil, the average bioavailability ofarsenic in the soil from this area (as determined by CVMUM) was 40%. TCEQ used thisbioavailability factor to calculate a proposed residential soil clean-up level of 46 mg/kg.

Third, there is some evidence that the mechanisms that the body uses to distribute, metabolize, andexcrete arsenic may result in the ability of the body to detoxify much of the arsenic. Most organicand inorganic arsenic leaves the body in urine within a few days of exposure, although some remainsin the body for several months or longer. The metabolism of arsenic in humans has been wellstudied and involves two basic processes: 1) oxidative/reduction reactions that interconvert arsenateand arsenite, and 2) methylation reactions, which convert arsenite to the organic formsmonomethylarsonic acid (MMA) and dimethylarsinic acid (DMA). Human exposure to eitherinorganic arsenites or arsenates results in increased urinary levels of inorganic arsenite, inorganicarsenate, MMA and DMA (ATSDR, 1993). The relative proportions of these metabolites in urinecan vary depending upon the type of arsenic administered, the time after exposure, the route ofexposure, and the dose of arsenic administered. In humans, the relative proportions are usually about40-60% DMA, 20-25% inorganic arsenic, and 15-25% MMA. The liver is apparently the main siteof the methylation process and the enzymes involved apparently require arsenite as a substrate. Thus, although the methylation products generally predominate, arsenate is not methylated unlessit is first reduced to arsenite.

Since the methylated metabolites of arsenic are less toxic than inorganic arsenic, methylation maybe an important detoxification mechanism. Because the methylation of arsenic is an enzymaticprocess that could be saturated, the dose at which this saturation occurs is an important issue. Limited human data suggest that doses up to approximately 200 µg/day are sufficiently methylated,but that doses above this level may begin to saturate the system, leading to inorganic arsenicaccumulation in tissues. These observations are uncertain because they are based on limited data. Since the dose rate at which the methylating capacity becomes saturated is uncertain and cannot beprecisely defined with current data it would be inappropriate to base public health decisions on thesedata. In addition, the pattern of arsenic methylation following chronic low-level exposure to arsenichas not been adequately studied. Although under normal conditions, the availability of methyldonors (i.e., methionine, choline, and cysteine) apparently does not affect the rate of detoxification;dietary deficiencies such as low choline, low methionine, and low protein may reduce the ability todetoxify arsenic.

Non-Carcinogenic Effects

By the oral route, the effects most likely to be of human health concern are gastrointestinal irritation,decreased production of red and white blood cells, abnormal heart function, blood vessel damage,impaired nerve function causing a "pins and needles" sensation in the hands and feet, and a groupof skin diseases, including hyperkeratosis. Most of the non-cancer effects begin to occur at similaroral exposure levels. Ingestion of food or water with 0.3 to 30 ppm of arsenic can cause stomachand intestinal irritation. The minimal dose at which these effects usually are observed in humansafter chronic ingestion of arsenic ranges from 0.012 milligrams of arsenic per kilogram of bodyweight per day (mg/kg/day) to 0.1 mg/kg/day.

Although there is no evidence to suggest that arsenic can injure pregnant women or their fetuses,studies of animals show that doses large enough to cause illness in pregnant females also may causelow birth weight, fetal malformations, or fetal death. One of the most characteristic effects oflong-term oral exposure to inorganic arsenic is a pattern of skin changes that includes a darkeningof the skin and formation of hyperkeratotic warts or corns on the palms, soles, and torso. Currentlythis end-point is considered the most appropriate basis for establishing a chronic oral Minimal RiskLevel (MRL) or Reference Dose (RfD). However, other end-points (liver injury, vascular disease,and neurological effects) also appear to have similar thresholds (ATSDR, 2000).

In one study of a very large population, Tseng (1968), found no adverse effects in any person withan average total daily intake of inorganic arsenic (water plus food) of 0.0008 mg/kg/day. This studyhas served as the basis for both ATSDR's MRL and EPA's RfD, both of which are 0.0003 mg/kg/day. Both the RfD and the MRL were derived by dividing the 0.0008 mg/kg/day NOAEL by anuncertainty factor of three (3) to account for both the lack of data on reproductive toxicity and toaccount for some uncertainty as to whether the NOAEL accounts for all sensitive individuals. Thereis not a clear consensus among scientists regarding the oral RfD. Arguments for various valueswithin a factor of 2-3 of the recommended RfD value have been made.

To evaluate the proposed clean-up level of 46 ppm arsenic in surface soil, we independentlyevaluated the exposure that children might receive from all sources (soil ingestion, aboveground andbelowground vegetables, dermal absorption, and inhalation). We applied the bioavailability factorof 40% to the ingestion scenarios. We used children because they constitute the sub-populationmost likely to experience the highest levels of exposure to arsenic in soil (due to play activities andnormal hand-to-mouth exposure) and because their dose of arsenic relative to body weight is higherthan that of adults (ATSDR, 1995). Assuming a maximum soil arsenic concentration of 46 mg/kgand combined exposure from all sources, the estimated daily dose that a child would receive(0.000313 mg/kg/day) is equivalent to EPA's RfD of 0.0003 mg/kg/day. Therefore, with respect tonon-carcinogenic health effects the proposed cleanup level of 46 mg/kg for arsenic in soil is adequateto protect children(1) (See Appendix A for pathway-specific parameters and results).

Short-Term or Sporadic Pica Behavior

Soil pica behavior (ingestion of more than 1.0 gram of soil per day) may occur in a sizable portionof children throughout the year (Calabrese, 1997). While any individual child may only exhibit picabehavior infrequently, the behavior is not limited to a small subset of the population. It has beenestimated that approximately 62% of children will ingest >1.0 gram of soil on 1-2 days/year. Additionally, 42% of children will ingest >5 grams of soil and 33% will ingest >10 grams of soil on1-2 days per year. For some contaminants periodic pica episodes potentially could result in acuteintoxication (Calabrese, 1997). To explore the potential public health significance of pica behaviorat this site, we considered the scenario of a 15 kg child who ingested 5,000 mg of soil per day for14-21 days. At a soil arsenic concentration of 46 mg/kg and a bioavailability factor of 40%, the dailydose of absorbed arsenic during the pica events would be approximately 0.006 mg/kg/day - wellbelow the acute LOAEL for serious effects (0.050 mg/kg/day) reported by Mizuta, 1956. Indeed,for such a child, the soil arsenic level would have to exceed 375 mg/kg in order for the child toabsorb a dose exceeding the acute LOAEL. Alternatively, the child would have to ingest 40.8 gramsof soil per day for 14-21 days at the proposed clean-up level of 46 mg/kg in order to exceed the acuteLOAEL for arsenic. The effects associated with this acute LOAEL include nausea, vomiting,diarrhea, occult blood in the feces and gastric and duodenal juice, and abnormal electrocardiogram(Mizuta, 1956). Based on these data, we would not expect to see any children exibiting signs orsymptoms of acute toxicity from arsenic as a result of short-term or sporadic pica behavior (See Appendix B for pathway-specific parameters and results).

Carcinogenic Effects of Arsenic

A large number of epidemiologic studies and case reports provide evidence that ingestion of arsenicincreases the risk of developing cancer. The most common effect is increased risk of multiple skincancers. Some of the skin cancers develop from hyperkeratotic warts or corns characteristic ofchronic arsenic exposure. Multiple basal cell carcinomas may also occur, usually from cells notassociated with hyperkeratinization. In most cases, skin cancer develops only after prolongedexposure; however, several studies have reported skin cancer in people exposed for less than oneyear. Liver, bladder, kidney, and lung cancer also have been associated with oral exposure to arsenic(Smith, 1992), but these associations are less well established and currently not suitable for inclusionin risk estimates.

Based on the epidemiological studies, the EPA has classified arsenic as a Group A "known human"carcinogen. This classification is based on consistent evidence of increased risk of lung cancer inworkers exposed to airborne arsenic contaminated dust (EPA, 1984) and on the clear dose-dependentrelationship between ingested arsenic and skin cancer (Tseng, 1968 & 1977).

Using the dose-response data from the 1968 Tseng study, the EPA derived an oral cancer slope factorof 1.5 (mg/kg/day)-1 for inorganic arsenic using a model that assumes the dose-response curve islinear at low doses. The fact that the body may detoxify arsenic through methylation suggests thatthe dose-response curve for arsenic may be non-linear at low doses. Thus, the slope factor basedon the linear model may over-estimate cancer risks at low doses. The EPA has concluded thatalthough the current slope factor might overestimate low dose risk, data are too limited to permit aquantitative adjustment of the slope factor (EPA, 1988). Based on the above slope factor, weestimated the excess lifetime risk for developing cancer, associated with exposure from soilingestion, above- and below-ground home-grown vegetable ingestion, soil dermal contact, andair/dust inhalation combined to 46 ppm arsenic in surface soil from all pathways for 30 years, to be5.74 x 10-6. Qualitatively, we would interpret this risk level as no apparent increased lifetime riskfor developing cancer (See Appendix C for pathway-specific parameters and results).


Based on what we know about the effects of arsenic on the human body under normal exposureconditions and after applying the results of the arsenic bioavailability study conducted by CVMUMto the ingestion exposure scenarios, TDH and ATSDR conclude the following:

  1. Considering all pathways of exposure to inorganic arsenic in surface soil, we would not expect a soil clean-up level of 46 mg/kg to result in adverse non-carcinogenic health effects in children or adults.

  2. The proposed clean-up level of 46 mg/kg of arsenic in surface soil is not expected to be a problem (with respect to serious or less serious adverse health effects) even in children with short-term or sporadic pica behavior who may eat up to 5,000 mg soil per day for up to 14 days in any one year period.

  3. Long-term exposures (30 years) associated with the proposed clean-up level of 46 mg/kg of arsenic in surface soil would result in no apparent increased lifetime risk of developing cancer.

  4. Based on conclusions 1, 2, and 3 above, the proposed cleanup level of 46 mg/kg for arsenic in surface soil at this site would pose no apparent public health hazard.

  5. This assessment does not evaluate exposure risks associated with significant chronic pica behavior.


Not applicable


Not applicable


Report Prepared by

Richard Beauchamp, MD
Senior Medical Toxicologist
Environmental Epidemiology and Toxicology Division

John F. Villanacci, PhD
Environmental Epidemiology and Toxicology Division

ATSDR Regional Representatives
Jennifer Lyke
Regional Representative
ATSDR Region 6

Karl Markiewicz, Ph.D.
Regional Representative
ATSDR-Region 6

Robert Knowles, MS, REHS
Environmental Health Scientist
Division of Health Assessment and Consultation
Superfund Site Assessment Branch
State Programs Section


Agency for Toxic Substances and Disease Registry (ATSDR). 1995. Child health initiative. Atlanta, GA: U.S. Department of Health and Human Services.

ATSDR. 2000. Toxicological Profile for Arsenic. Atlanta, GA: U.S. Department of Healthand Human Services.

Byron, WR., Bierbower, GW., Brouer, JB., et al. 1967. Pathological changes in rats anddogs from two-year feeding of sodium arsenite or sodium arsenate. Toxicol. Appl.Pharmacolo. 10:132-147.

Calabrese, E.J., Stanek, E.J., James, R.C., and Roberts, S.M. Soil Ingestion: A concern for acutetoxicity in children. Environ. Health Perspect. 105: 1354-1358 (1997).

U.S. Environmental Protection Agency (EPA). 1984. Health assessment document forinorganic arsenic. Final Report. Research Triangle Park, NC: EPA-6008/83-021F.

Gaines, TB. 1960. The acute toxicity of pesticides to rats. Toxicol. Appl. Pharmacol. 2:88-99.

Maitani, T., Saito, N., Abe, M., et al. 1987. Chemical form-dependent induction of hepaticzinc-thionein by arsenic administration and effect of co-administered selenium in mice. Toxicol.Lett 39:63-70.

Mizuta, N., Mizuta, M., Ito, F., et al. An outbreak of acute arsenic poisoning caused by arsenic-contaminated soy-sauce: A clinical report of 220 cases. Bull Yamaguchi Med Sch 4(2-3):131-149(1956).

Sardana, MK., Drummond, GS., Sass, S. et al. 1981. The potent heme oxygenase inducingaction of arsenic in parasiticidal arsenicals. Pharmacology 23:247-253.

Shils, ME. and Young, VR. 1988. Modern Nutrition in Health and Disease. Seventh ed. Lea& Febiger. Philadelphia. PA.

Smith, AH., Hopenhayn-Rich, C., Bates, MN., Goeden, HM., Hertz-Picciotto, I., Duggan,HM., Wood, R., Kosnett, MJ., Smith, MT. 1992. Cancer risk from arsenic in drinkingwater. Environ. Health Perspect. 97:259-267.

Tseng, WP., Chu, HM., How, SW., Fong, JM., Lin, CS., Yeh, S. 1968. Prevalence of skincancer in an endemic area of chronic arsenism in Taiwan. J. Natl. Cancer Inst. 40:453-463.

Tseng, WP. 1977. Effects and dose-response relationships of skin cancer and Blackfootdisease with arsenic. Environ. Health Perspect. 19:109-119.


This public health consultation was prepared by the Texas Department of Health (TDH) under aCooperative Agreement with the Agency for Toxic Substances and Disease Registry(ATSDR). It is in accordance with approved methodology and procedures existing atthe time the health consultation was initiated.

Robert Knowles
Technical Project Officer, SPS, SSAB, DHAC, ATSDR

The Division of Health Assessment and Consultation, ATSDR, has reviewed this health consultation and concurs with its findings.

Roberta Erlwein
Chief, State Programs Section, SSAB, DHAC, ATSDR



Click here to view Appendix A in PDF format (PDF, 299KB)


Click here to view Appendix B in PDF format (PDF, 180KB)


Click here to view Appendix C in PDF format (PDF, 327KB)

1 This consideration is in accordance with ATSDR's Child Health Initiative (ATSDR, 1995).

Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Centers for Disease Control and Prevention   1600 Clifton Road Atlanta, GA 30329-4027, USA
800-CDC-INFO (800-232-4636) TTY: (888) 232-6348 - Contact CDC–INFO

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #