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

(a/k/a Tanapag Village (Saipan))
Tanapag Village, Saipan, Commonwealth of the Northern Marianas Island
EPA Facility ID: MPD982524506
August 31, 2004

Oral bioavailability of polychlorinated biphenyl (PCB) residues in soil

Importance of bioavailability in exposure assessment

In evaluating exposure pathways involving PCB contamination in soil, ingestion typically contributes the most to estimated exposure dose, with predicted exposures to children being the most significant. In order for a chemical in soil to produce a hazard, it must be desorbed from the soil matrix and absorbed into the systemic circulation where it can be transported to the site of action (Figure 1).

Bioavailability is defined as the percentage of an external exposure mass that reaches the systemic circulation (internal dose). The bioavailability of PCBs is governed, in part, by the interactions between the organic compound and the soil. The nature and extent of these interactions depend upon the chemical and physical properties of the soil and the organic compound, as well as the conditions present at the point of contact with a receptor (NEPI, 2000).

Figure 1.

(Source: NEPI, 2000)

When evaluating exposure to contaminants in soil, an important consideration is the bioavailability of the external exposing mass relative to the bioavailability of the chemical compound under the conditions used to derive the toxicity criteria. This is especially important when the conditions of exposure are different than those in the toxicology studies (Hrudley et al., 1996). For the sake of simplicity and conservatism, many of the toxicity criteria developed by USEPA and ATSDR do not reflect the fraction of the external dose that is absorbed (Paustenbach et al., 1997). Failure to adjust the toxicity criteria to incorporate bioavailability may result in a substantial over prediction of exposure and subsequent health harm due to ingestion of contaminated soil.

Practical bioavailability of PCBs comes from actual field experience in humans living near hazardous waste sites. Serum PCB levels were within background ranges in persons at highest risk of non-occupational exposure to PCBs at 10 different contaminated sites, even though soil PCB levels were as high as 330,000 ppb in soil samples (Stehr-Green et al., 1988). At two other sites, where average blood levels were elevated, it was subsequently determined that occupational exposures and consumption of PCB-contaminated fish had also occurred. These data indicate that, in contaminated environments where food contamination is not an issue, humans usually do not accumulate additional body burdens of PCBs (Kimbrough, 1995).

Estimates of oral bioavailability of PCB in soil

Bioavailability considerations can be incorporated into an exposure assessment by measuring the bioavailable fraction of a compound in soil. This can be done using either animal studies (i.e., in vivo tests) or physical/chemical studies (i.e., in vitro tests) that mimic biological systems in animal models.

A challenge to understanding the bioavailability of PCBs in soil is the variability of chemical and biological factors critical to absorption compounded by the variability ensuing from the weathering of soils. Studies that evaluate bioavailability under the laboratory environmental may not be representative of the mixture under environmental conditions.

Precise estimates of the oral bioavailability of PCB resides in soil are not available. In vivo and in vitro studies indicate that the bioavailability of PCB in soil is significantly less than the bioavailability of PCB in corn oil used to derive toxicity criteria for PCB (Fries et al., 1989; Fries, 1985; Oomen et al., 2001; Hack et al., 1996). From controlled feeding studies involving weathered soil, bioavailability appears to be higher than in vitro studies that used an artificial digestive tract model (Table 1).

Table 1.

Percent PCB bioavailability in soil.
Citation Range of Bioavailability Experimental model
(Fries, 1985) 40-65% In vivo controlled feeding study in ruminants
(Oomen et al., 2001 30-47% In vitro artificial digestive tract combined with intestinal epithelial cell absorption model
(Hack & Selenka, 1996) 33-64% In vitro artificial digestive tract model

Absorption of halogenated hydrocarbons is high and often exceeds 90% of the dose when administered in corn oil (Matthews et al., 1978), however PCBs in soil are adsorbed to organic matter and clay particles and are not as bioavailable (Fries, 1985). This is consistent with in vivo studies in rats demonstrating that intestinal absorption of PCBs administered via spiked soil is lower than that of PCBs ingested via corn oil (Fries et al., 1989).

In vivo studies

Lambs fed soil spiked with polybrominated biphenyls (PBB) in their diet were found to have a net absorption of 40-65% as measured in fecal matter (Fries, 1985). Fries varied the organic content of the soil and discovered decreasing PCB bioavailability with increasing organic content.

In vitro studies

Eighteen PCB-contaminated soils and materials were evaluated in an artificial digestive tract model to simulate in vivo digestion of soil-bound contaminants (Hack and Selenka, 1996). Thirty-three percent of PCBs in soil were mobilized in the gastrointestinal assay, while 64% were mobilized when dry whole milk was added. The presence of lipids and micelles were attributed to the increased mobilization. The amount mobilized represented the fraction of chemical in the gut that has desorbed from soil and is available to be absorbed. Absorption across the intestinal lumen was not evaluated in this assay.

The bioaccessible fraction of four PCB congeners in soil ranged from 30-47% in another artificial digestive tract model (Oomen et al., 2000). This assay utilized an in vitro artificial digestive system followed by an intestinal epithelial cell model to simulate mobilization from soil and absorption and uptake from the gastrointestinal tract. The assay included sequential addition of saliva, gastric juice, duodenal juice, and bile salts.

Practical considerations

Only a fraction of the chemical that is measured in a soil may be truly available to impact the health of humans. To more accurately estimate oral exposure to PCBs in contaminated soil, several elements that can affect the assessment need to be considered. First, an accurate estimate of soil ingestion is needed, typically ranging from 50-200 mg/day (Calabrese et al, 1997; van Wijnen et al., 1990; Davis et al., 1990). Secondly, site-specific factors need to be weighed that may affect the degree of PCB mobilization from soil. The oral bioavailability of PCBs in soil varies with soil type, age, and chlorination level of the compound, with bioavailability decreasing with aging in the soil and with greater levels of chlorination (NEPI, 2000). PCBs in soils with higher organic content appear to be less bioavailable (Fries 1985).

Mobilization of PCBs during digestion is the critical factor determining the oral bioavailability of PCBs in soil. PCBs mobilized from soil should be regarded as available for intestinal absorption. In studies using in vitro digestion models, the amount of contaminant that is mobilized from soil represents the maximum amount available for intestinal absorption (Oomen et al., 2001).

Finally, the degree of intestinal absorption can be affected by additional factors such as dietary lipids and the presence of soil co-contaminants (such as oil). Because PCBs are highly fat soluble, increasing dietary lipid levels can increase mobilization of PCB from soils and absorption from the digestive tract. It is less clear what the effect of constituents such as bile proteins and fatty acids in intestinal fluid may have on absorption and PCB bioavailability, however more PCBs were mobilized from soil when more bile or protein was added during the artificial digestion (Oomen et al, 2000).


Although no precise estimate of PCB bioavailability in soil is available, it appears that the range is around 40-65%. The Fries study defined an upper range of 65% in vivo, which is supported by the in vitro model used by Oomen et al. (Table 1). This in vitro model simulated both mobilization and absorption, attempting to account for the impact of other constituents in the digestive tract on bioavailability. Given the variability of chemical and biological factors that affect bioavailability, it is important to take into account site-specific environmental conditions such as soil type, age, organic content and dietary lipids when selecting an appropriate bioavailability factor. In lieu of this information, using the upper range of bioavailability is recommended.

For further information on the bioavailability of organics in soils, see the National Environmental Policy Institute (NEPI) report entitled "Assessing the Bioavailability of Organic Chemicals in Soil for Use in Human Health Risk Assessments" available at


Calabrese, E. J., Stanek, E. J., James, R. C., & Roberts, S. M. (1997). Soil ingestion: a concern for acute toxicity in children. Environ.Health Perspect., 105, 1354-1358.

Davis, S., Waller, P., Buschbom, R., Ballou, J., & White, P. (1990). Quantitative estimates of soil ingestion in normal children between the ages of 2 and 7 years: population-based estimates using aluminum, silicon, and titanium as soil tracer elements. Arch.Environ.Health, 45, 112-122.

Fries, G. F. (1985). Bioavailability of soil-borne polybrominated biphenyls ingested by farm animals. J.Toxicol.Environ.Health, 16, 565-579.

Fries, G. F., Marrow, G. S., & Somich, C. J. (1989). Oral bioavailability of aged polychlorinated biphenyl residues contained in soil. Bull.Environ.Contam Toxicol., 43, 683-690.

Hack, A. & Selenka, F. (1996). Mobilization of PAH and PCB from contaminated soil using a digestive tract model. Toxicol.Lett., 88, 199-210.

Kimbrough, R. D. (1995). Polychlorinated biphenyls (PCBs) and human health: an update. Crit Rev.Toxicol., 25, 133-163.

National Environmental Policy Institute (NEPI) (2000). Assessing the Bioavailability of Organic Chemicals in Soil for Use in Human Health Risk Assessments. Bioavailability Project Policy, Organics Task Force. Available at

Oomen, A. G., Sips, A. J., Groten, J. P., Sijm DTHM, & Tolls, J. (2000). Mobilization of PCBs and Lindane from soil during in vitro digestion and their distribution among bile salt micelles and proteins of human digestive fluid and the soil. Environ Sci Technol, 34, 297-303.

Oomen, A. G., Tolls, J., Kruidenier, M., Bosgra, S. S., Sips, A. J., & Groten, J. P. (2001). Availability of polychlorinated biphenyls (PCBs) and lindane for uptake by intestinal Caco-2 cells. Environ.Health Perspect., 109, 731-737.

Paustenbach, D. J., Bruce, G. M., & Chrostowski, P. (1997). Current views on the oral bioavailability of inorganic mercury in soil: implications for health risk assessments. Risk Anal., 17, 533-544.

Stehr-Green, P. A., Welty, E., & Burse, V. W. (1988). Human exposure to polychlorinated biphenyls at toxic waste sites: investigations in the United States. Arch.Environ.Health, 43, 420-424.

van Wijnen, J. H., Clausing, P., & Brunekreef, B. (1990). Estimated soil ingestion by children. Environ.Res., 51, 147-162.

Exposure history survey form

Commonwealth of the Northern Mariana Islands, Department of Public Health, Medical Evaluation for PCB Exposure form   [PDF:228KB]

Data Analysis Methods for Exposure History Survey

A priori confounders were screened with a cut point of p < 0.10 for inclusion in multivariate analyses using Spearman non-parametric correlation analysis for continuous variables and assessing the difference between least-squares means of the logged PCB values for categorical variables. These a priori confounders included age, triglycerides, cholesterol, body mass index, and gender.

For land crab consumption, distribution of serum PCB levels were characterized by the four levels of reported consumption (no consumption, one to two meals per week, three to five meals per week, and more than five meals per week). Since average PCB concentrations were heavily weighted by the large number of non-detects, we evaluated the upper tail of the distribution (the 90th and 95th percentiles) by level of reported consumption.

To evaluate the exposure history variables, we compared those with serum PCB concentrations below 5 ppb to those with serum PCB concentrations at 5 ppb and above. The level of 5 ppb was chosen as the cut point for this analysis because it represents the upper range of background exposures in previously studied communities (NCEH 2001). Risk estimates for having 5 ppb or greater serum PCB were calculated for the different levels of reported food consumption and for reported contact with the capacitors, the capacitor oil, or contaminated soil. Odds ratios and 95% confidence intervals (CI) from logistic regression were used to approximate risk estimates after adjusting for age, sex, triglycerides, and cholesterol.

The 95% CI is the range of estimated values that has a 95% probability of including the true risk estimate for the population. The confidence interval is a statistical measure of the precision of the risk estimate. For example, if the confidence interval does not include 1.0 and the interval is below 1.0, then the variable being considered demonstrated significant "protection" for the given outcome. Similarly, if a confidence interval does not include 1.0 and the interval is above 1.0, then variable being considered demonstrated a significant elevated risk for the given outcome. If the confidence interval includes 1.0, then the true ratio may be 1.0, and it cannot be concluded with sufficient confidence that the observed risk estimate reflects a real excess or deficit. The width of the confidence interval also reflects the precision of the risk estimate. For example, a narrow confidence interval (e.g., 1.1-1.5) indicates that the study's sample size was sufficiently large to generate a fairly precise risk estimate. A wide interval (e.g., 0.7-4.5) indicates far less precision, and more uncertainty, in the calculated risk estimate.

Health consultation: Evaluation of Land Crab Contamination

Health Consultation  [PDF:440KB]

Response to Public Comments

The Agency for Toxic Substances and Disease Registry (ATSDR) received the following comments/questions during the public comment period (July 28 to September 9, 2003) for the Tanapag Village (Saipan) Public Health Assessment (PHA) (July 28, 2003). For comments that questioned the validity of statements made in the PHA, ATSDR verified or corrected the statements. The comments presented below do not include editorial comments concerning such things as word spelling or sentence syntax.

Polychlorinated Biphenyl (PCB) and Exposure-Related Comments:
  1. Comment: Would PCB serum levels be higher if samples were collected when exposures first occurred? Older people may have higher PCB serum levels than younger people because of gradual accumulation over time, but could the higher PCB levels in older people also be a result of higher exposures occurring when PCBs were first released? Would exposure to PCB vapors when the capacitors broke open result in increased exposures?

    Response: In Section 7.0, Evaluation of Public Health Impact of PCB Contamination, ATSDR discusses some of the limitations of the PCB serum level data. PCB serum level data cannot be used to identify PCB sources (e.g., fish, land crabs, capacitor fluid), predict possible public health impacts, or determine past PCB serum levels. Nonetheless, these data are the best available information for measuring PCB exposures from all possible environmental sources and comparing Tanapag resident exposures to exposures occurring in other populations. In conducting comparisons between populations, public health officials can draw only general conclusions about possible exposures.

    Although past PCB serum levels cannot be determined from the available data, the levels reported in Tanapag residents is consistent with studies of PCB serum levels in populations with no known PCB exposures. Studies of non-occupationally exposed individual who did not consume fish from PCB-contaminated waterbodies identified mean PCB serum levels ranging from 3 to 17 parts per billion (ppb), with individual concentrations ranging from non-detect to 60 ppb. The average PCB serum level in Tanapag residents was 2.0 ppb. The vast majority of Tanapag residents had non-detect levels of PCBs (892 individuals or 84% of the tested population). Another 152 individuals (14%) had levels between 3 and 10 ppb. Only 15 individuals (2%) had PCB serum levels above 10 ppb, with the highest level reported at 36 ppb. Studies of individuals known to be exposed occupationally and through fish consumption found average PCB serum levels from 4 to 400 ppb (individual levels ranged from non-detect to 2,200 ppb) and from 5 to 72 ppb (individual levels ranged from non-detect to 360 ppb), respectively (ATSDR 1997). Because the Tanapag data are most consistent with data from studies of non-exposed populations, ATSDR can infer that exposures occurring in Tanapag are within normal ranges and would not be expected to result in adverse health effects.

    Results from Tanapag are also similar to studies of PCB levels reported by age group. Several studies of PCB levels in tissue have found a significant correlation between PCB concentrations and age. As age increased, so did the levels of PCBs. In addition, PCBs are known to bioaccumulate and biomagnify in animals. Under these processes, PCBs collect in animal tissue faster than the PCBs can breakdown, resulting in higher PCB concentrations in older animals (ATSDR 1997). The same processes apply to humans, as reflected by the increased blood serum levels found in older Tanapag residents. Again, ATSDR can infer that exposures occurring in Tanapag are consistent with the general population.

    PCB exposures most likely occur through ingestion of contaminated media (e.g., soil, sediment, or foodstuff), dermal contact with these media, or inhalation of contaminated dust, as described in detail in Table 7. Inhalation of PCB vapors is expected to be minimal compared to these other exposure routes. The vapor pressure, which is a measure of a chemical's ability to move to the air, for PCBs is very low and indicates that PCBs will move to the air slowly and in small concentrations. The vapor pressure for Aroclor-1254 is only 0.0000771 millimeters of mercury at 25 degrees Celsius (mmHg at 25 oC) and the vapor pressure for Aroclor-1260 is 0.0000405 mmHg at 25 oC. In comparison, the vapor pressures for several chemicals that readily move to the air are 181 (acetone), 75 (benzene), and 58 (trichloroethylene) mmHg at 25 oC, or about 1,000,000 times that of Aroclor-1254 and -1260 ( 2004).

  2. Comment: Are garment workers included in the Tanapag Village population numbers and how would considering or not considering this population affect the PHA findings? Based on population data, apparently only half the population (1,200 individuals from a population of 3,318) underwent health screening. How are the PHA findings affected by this participation rate?

    Response: Based on agreements with the Commonwealth of Northern Mariana Islands (CNMI) Department of Public Health (DPH), ATSDR included population data for permanent residents of Tanapag Village only. Temporary workers, such as those employed at the garment factory, were not considered in the assessment. Permanent residents may live in Tanapag for decades, whereas temporary workers only reside in Tanapag for several years. As such, permanent residents would be more likely to contact PCBs for longer durations then temporary workers. Long exposure durations could be associated with higher PCB serum levels, as indicated by study results finding higher PCB serum levels in older people. Considering only permanent residents, therefore, allowed the CNMI DPH and ATSDR to focus their efforts on the population segment most likely to have the highest exposure levels. As such, results may be biased high, meaning that the reported PCB serum concentrations may be higher than if both permanent residents and temporary workers were considered in evaluations.

    The population data reporting 3,318 individuals residing in Tanapag are from the 2000 census. Approximately 1,200 individuals (36% of the population) attended the open clinic in Spring and Summer 2000 for evaluation of PCB exposures. Arguments can be made that this participation rate could lead to data that are biased high (reported PCB serum levels are higher than actually occur in the population) or that the data are biased low (reported PCB serum levels are lower than actually occur in the population). For example, residents voluntarily participated in the exposure study. Data could be biased high if volunteers participated in the study because they felt they were exposed to PCBs more often than others in the community (e.g., people who visited the cemeteries daily or consumed land crabs daily). Data could be biased low, however, if a segment of the population did not participate in the study (e.g., older people who may believe that they have lead a long life without unusual health problems).

    To address potential biases, ATSDR conducted statistical analyses of the data and adjusted results to account for individual variations in age, gender, serum triglyceride, and serum cholesterol (factors which affect PCB serum levels). In addition, the study population included people with varying exposures based on categories such as land crab consumption rate, reported contact with capacitors or capacitor contents, and contact with soil in Cemetery 2. Although weakly positive associations between PCB serum levels and land crab consumption rates and age were found, the data did not conclusively indicate if these associations were indicative of adverse health effects. Because ATSDR considered potential biases in its evaluation, even if the other 64% of the Tanapag Village population were included in the study, sufficient data would not likely exist to conclusively determine if an association existed between exposures, PCB serum levels, and adverse health effects.

  3. Comment: Does the conclusion that banning land crab consumption is unnecessary contradict the information provided on how to remove or minimize PCBs in land crabs? Are land crabs safe to eat? Would understanding the land crab lifecycle provide further information to support or refute conclusions?

    Response: ATSDR concluded in this PHA and in the July 2001 Health Consultation that exposures to PCBs from eating large amounts of land crabs for many years are too low to cause adverse health effects in people, including sensitive individuals such as children. Because data gaps exist regarding PCB exposures and adverse health effects, ATSDR describes actions that people can take to minimize PCB exposures when consuming land crabs. The CNMI DPH and Department of Environmental Quality (DEQ) further recommended that people minimize or eliminate consumption of land crabs harvested from Tanapag Village. These recommendations simply reflect precautions people can take to minimize PCB exposures and do not signify that a public health concern exists.

    Gaining a deeper understanding of the land crab lifecycle is unlikely to change conclusions regarding potential health impacts from consumption of PCBs in land crabs. To assess concerns about consumption of land crabs, the US Environmental Protection Agency (EPA) and DEQ collected land crabs from areas identified as having the highest PCB concentrations in soil and from areas identified by Tanapag residents as common harvesting locations. EPA and DEQ also collected land crabs similar in size to those that Tanapag residents would harvest and consume. As such, PCB concentrations reported in these samples were considered representative of potential PCB exposures for Tanapag residents. Regardless of how PCBs entered and accumulated in the land crabs during their lifecycle, the available data provide information about likely exposures. Gathering information about the land crab lifecycle, therefore, is unlikely to change assessment conclusions.

    General Comments:

  4. Comment: Exposures to contaminants other than PCBs are not considered in this PHA. What are the possible adverse health effects from exposures to these chemicals?

    Response: ATSDR prepared this PHA in response to specific concerns regarding PCB exposures and potential adverse health effects. As such, ATSDR focused assessments and evaluations on PCB exposures and related public health concerns.

    Nonetheless, ATSDR reviewed available sampling data gathered for other contaminants to ensure that concerns beyond PCB exposures were also identified. Additional data reviewed included volatile organic compounds, semivolatile organic compounds, and metals data collected from the Former Tanapag Fuel Farm and Former Tanapag Military Disposal Site (discussed in Section 5.0 Nature and Extent of Environmental PCB Contamination). Additionally, metals data from land crab sampling (discussed in Appendix 14.7 Health Consultation: Evaluation of Land Crab Contamination) was considered in the evaluation. This data review identified only one sample collected from the center of the disposal site that contained elevated levels of arsenic, chromium, and iron. No other elevated levels of contaminants were identified. The data, therefore, do not indicate that impacts from contaminants other than PCBs are a pervasive problem in Tanapag Village. Regardless, remediation efforts to remove and treat soil containing PCBs would also address other contaminants present in the soil.

  5. Comment: A grid survey of the entire village to identify soil contamination has been repeatedly requested by Tanapag Village residents. In addition, more capacitors may exist in Tanapag Village because the number of capacitors shipped to Saipan has never been verified. What actions are being taken to address these concerns?

    Response: ATSDR forwarded these comments to EPA and DEQ for their review. ATSDR can only recommend actions and has no legal authority to enforce actions. As such, comments regarding environmental sampling, site characterization, and remedial actions are best addressed by EPA and DEQ, the regulatory agencies that have legal authority to require further sampling and remediation efforts.

  6. Comment: What could be a natural source of thallium in groundwater and are the elevated levels above EPA drinking water standards sufficient to warrant groundwater treatment?

    Response: Soil, which is partially composed of rocks that have been broken down over time, is a natural source of thallium in groundwater. Thallium is one of many chemicals and elements that can be found in rocks and soil and is available for transport to groundwater. Regardless of the thallium source, groundwater underlying Tanapag Village is too saline (salty) to serve as a daily water supply. People typically use bottled or rain water for drinking purposes. As such, ATSDR expects exposures to thallium at concentrations above the EPA drinking water standard to be minimal. If groundwater remains an undesirable source of drinking water, ATSDR believes that groundwater treatment for thallium is unnecessary.

  7. Comment: Is the Note of Explanation provided on the inside front cover the legal definition of a PHA referred to in the Foreword?

    Response: Yes. The Note of Explanation summarizes the regulatory framework and process under which ATSDR completes PHAs. The information provided in the Note of Explanation and the Foreword refers to the overall PHA process and does not provide site-specific information. ATSDR provided details of the site-specific methodology used to complete the PHA for Tanapag Village in the main text of the document.


ATSDR. 1997. Toxicological Profile for Polychlorinated Biphenyls (Update). Agency for Toxic Substances and Disease Registry. U.S. Department of Health and Human Services, Public Health service. September, 1997. 2004. Database and Internet Searching. Downloaded from .

Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

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. #