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

Evaluation of Pubic Health Impact of PCB Contamination

Introduction

The CNMI Department of Public Health invited residents of Tanapag to participate in a health screening. A clinic facility was opened in the Village to conduct this screening. The main purpose of the clinic was to evaluate exposure to PCB and to examine residents for physical illness related to PCB exposure. Other goals of the clinic included providing residents with free medical evaluation for common diseases and risk factors for disease, evaluating the need for a community-based primary care clinic, and establishing a database of health information for future public health planning.

The clinic was open five days each week for most of the Spring and Summer of 2000. Any individual who had resided in the village at the time of the screening, or who had lived in the village for greater than three years was eligible to participate in the screening. Over 1200 individuals were interviewed for history of exposure to PCBs or PCB-contaminated soil. Physicians provided complete medical exams to individuals who visited the clinic, and participants were asked to provide a medical history to screen for health conditions and health risk behaviors. Information was collected on participants' health conditions, including cancer, diabetes, and dermatologic problems. The clinical examination included the collection of blood samples for a battery of routine clinical tests. A portion of each blood sample was sent out to specialized commercial laboratories to measure PCB levels in blood serum.

Medical Screening and Clinical Evaluation of PCB Exposure

What is BMI?

Body Mass Index, or BMI, is a mathematical calculation used to determine whether a patient is underweight, overweight, or obese. This calculation is based on an individual's height and weight. BMI alone is not diagnostic and other information, such as waist circumference, age, gender, diet, exercise habits, and smoking, must be considered in determining if a high BMI is associated with increased risk of obesity related diseases and conditions.

No clinical signs of PCB-related illness was observed in Tanapag Residents. However, data gathered at the Tanapag Clinic revealed a number of community health issues including hypertension, diabetes, and obesity. All of these individuals were informed of their health status and provided with recommendations for treatment and follow-up with their physicians. Of the adults seen in the clinic:

• 48% had hypertension and over half of these were newly diagnosed by the clinic. Of those with previously diagnosed hypertension, only one-third were adequately controlled.
• 20% had diabetes mellitus and over one-third of these were newly diagnosed by the clinic. Of those with previously diagnosed diabetes, only one-third were adequately controlled.
• Nearly 54% of the clinic population were considered obese with a body mass index (BMI) greater than 30. This is important because obesity is a risk factor associated with diseases such as hypertension and diabetes mellitus.

The Tanapag Clinic examined over 500 individuals from newborns to 19 years of age. 450 had vital signs recorded. Examinations revealed that signs of hypertension, elevated blood glucose, and obesity were already being seen in the teenagers. Also, 38 cases of impetigo and 31 cases of eczema were diagnosed among this age group. As with the adults, all of these individuals were informed of their health status and provided with recommendations for treatment and follow-up with their physicians.

Smoking and pack-years

Determining the risk of developing tobacco-related illness is based partially on the number of years smoked and the number of cigarettes smoked in a lifetime. A common equation for estimating this risk is by the number of "pack-years" smoked. A "pack year" is the equivalent of smoking a pack of cigarettes a day for a year. To calculate pack years, multiply the number of years smoked by the number of packs per day.

Smoking for less than ten pack-years generally places a person at lower risk for developing illnesses commonly related to tobacco use. If a person has smoked more than ten pack years, more serious tissue damage is to be expected. The earlier a person reduces or quits smoking, the lower the level of tissue damage.

Health risks are increased if someone has risk factors that make people more susceptible to smoking-related illnesses. Several of these high risk categories include:

1. Pregnant women.
2. Women over thirty who take birth control pills.
3. Members of families who are high risk for heart disease.
4. Smokers who already have a tobacco-related illness.
5. Smokers who are exposed to toxic agents in the workplace.
6. Smokers with high risk lifestyles (such as being overweight, a heavy drinker or low physical activity level).

Reducing or quitting smoking will greatly enhance health at all times, no matter how many pack years someone has smoked. During the clinical exams, physicians collected information related to certain behaviors that have an impact on health. During the clinical exams, physicians collected information related to certain behaviors that have an impact on health status. About 30-40% of the adults smoke and half of these are at least 20 pack-year smokers.

Of the adults seen in the clinic, one-quarter have more that three drinks of alcohol per day. Physicians generally recommend that adults consume no more than two drinks of alcohol be per day. Women are encouraged to not consume alcoholic beverages during pregnancy.

Results from 1999 ethnomedical study and medical records review

In 1999, to respond to concerns from Tanapag residents about health effects associated with PCB contamination, the ACOE funded an ethnomedical study for the potential exposure period of 1968-1996. The goal of this project was to determine if medical records and related documents indicated if PCB contamination in Tanapag Village has had a significant impact on the health of residents in the area. An ethnomedical study involves gathering historical, medical, and ethnographic data on human exposure and this was done during a month long field study. This field study involved extensive interviews with village residents, community elders, health care practitioners and other observers who could provide ethnographic and ethnomedical information needed to supplement the medical data being gathered. Particular effort was made to locate the families who resided near the soil/sampling sites identified in earlier studies. Medical records were reviewed for a total of forty seven residents; death records were reviewed for the entire village.

Reviews of medical records indicate that various types of skin rashes are common in Tanapag, and that biomedical attention is frequently sought for treatment. Two other conditions-impetigo, which is caused by bacteria such as Streptococci, and psoriasis for which there is a genetic predisposition-also occur with some regular frequency. While skin rashes vary in symptoms (dry scaly itchiness to pustulated sores) and causes, together these affected nearly 29% of all Tanapag residents who sought out medical treatment at the Commonwealth Health Center between 1968 and 1996. It is difficult to compare incidence rates for Tanapag with the general population since skin rashes do not form a category which is routinely reported. Skin related ailments such as psoriasis, skin rashes, cellulitis and impetigo are very common in the islands (Prasad 1999).

With regards to mortality data, cerebrovascular disease (stroke) and myocardial infarction were the third and fourth leading causes of mortality for the CNMI in 1995; accidental death is the leading cause. Cancers, including lung, liver and cervical, are increasing in the overall CNMI population and are the second leading cause of death. Mortality rates due to lung and liver cancer are higher in Carolinians than in Chamorros (Prasad 1999).

While the history of possible exposure extends over several decades, the results show little direct exposure to PCBs as revealed in the medical records and interviews. With the exception of a couple of cases, morbidity data in general do not reflect the types of symptoms known or believed to be associated with PCB exposure. As a result, there appears to be no significant impact to human health from potential PCB exposures.

Collection and Analysis of Blood Serum

Serum sampling and analysis, May 2000

With ATSDR input, the CNMI DPH developed an exposure investigation protocol. ATSDR staff assisted the DPH in conducting the exposure investigation in May and June of 2000.

Methods

Target Population and Recruitment:

Any individual who currently resides in the village, or has lived in the village for greater than one year in the past, was invited to participate in the exposure investigation. Prior to conducting the investigation, ATSDR staff conducted a public meeting to discuss residents' questions and concerns.

Consent and Confidentiality:

Each adult and a parent or legal guardian of each minor participant was required to sign an informed consent/assent form, prior to testing. Confidentiality was protected in accordance with applicable Federal and Commonwealth laws.

Blood serum sampling and analysis:

A 10 ml tube of blood was drawn from each individual participating. After clotting and spinning, 4 ml of serum was sent for analysis of whole weight total PCBs by high resolution capillary gas chromatography. Serum lipids were measured and used to interpret individual results when reported back to patients.

Results

The mean (average) serum PCB level for this population was 2.0 mg/L (ppb) with a 95% upper confidence level of 5.7 mg/L (ppb). A value of one half the detection level was used for all reported non-detects. Figure 6 illustrates the distribution of serum PCB concentrations found in this exposure investigation. Of the1059 individuals who had PCB results reported from the lab, 892 had non-detectable (< 3 ppb) levels. Another 152 individuals had levels between 3 and 9.9 ppb, 9 had levels between 10 and 19.9, and 6 had levels greater than 20 ppb. The highest level was 36 ppb.

Discussion

Based on the results of this exposure investigation, it does not appear that the population tested in Tanapag Village experienced unusually high PCB exposure. Detected serum PCB concentrations appear to be in the range of typical U.S. background levels. Although we do not have a reference population specific to Saipan, the Division of Laboratory Sciences at the National Center for Environmental Health, Centers for Disease Control and Prevention considers a level up to 5 ppb as a "non-occupationally exposed" level typical of U.S. background levels. (NCEH 2001) No children or young adults (up to age 30) had detectable levels of PCB ( 3 ppb), so it appears that no unusual exposure has taken place during the past several years. The Tanapag population tested tended to have slightly increasing serum PCB levels with increasing age. This is similar to findings in other investigations, and reflects the normal accumulation of organic substances (like PCBs) in the body over time (see body burden in appendix 14.3). Figure 7 shows the average serum PCB levels in Tanapag residents by age and gender.

It is difficult to associate detection of serum PCB with a particular exposure source. From the results it is not possible to predict whether there was exposure to PCBs in capacitor fluid, soil or local foods in past decades or whether PCB body burdens were previously higher. Measurement of serum PCB levels is not a medical diagnostic tool and not predictive of current or future illness. In spite of these limitations, it is the best measure of cumulative PCB exposure from all environmental sources and serves as a useful comparison between the exposure levels of one population versus another.

Public health professionals can make inferences about the level of exposure and public health impact of PCB contamination by comparing the serum PCB levels in a population of interest (Tanapag) with the serum PCB levels in other groups who have not had unusual PCB exposures (such as accidents, working with PCBs or eating contaminated fish). The majority of the population in Tanapag had serum PCB levels in the normal background range. The few individuals who had elevated levels were below the range where harmful health effects were observed in individuals who had high serum PCB levels resulting from accidental or work place exposures.

Some consideration was given to testing fat tissue samples for PCB, however serum PCB level was chosen as the appropriate biological marker of PCB exposure in Tanapag Village residents. Serum PCBs are accepted as strong indicators of PCB exposure (Dekoning and Karmaus 2000). Blood sampling is also less invasive, less prone to complications (e.g. infection, scarring), and less expensive than fat tissue sampling. Additionally, data exist on serum PCB concentrations in other exposed populations that simplify comparison and interpretation of the serum PCB levels in Tanapag.

It is ATSDR's determination that the residents of Tanapag Village who were tested in the exposure investigation have levels of PCB in their serum which would not be expected to cause clinically significant adverse health effects.

Phase II serum sampling and PCB congener-specific analysis, October 2001

CNMI DPH requested that ATSDR provide technical and financial assistance to provide congener-specific serum PCB analysis for the Tanapag residents testing in the highest percentile for serum PCB in the original May 2000 exposure investigation. CNMI DPH was asked by the local community to provide an independent confirmation of the initial serum PCB sampling results.

Some members of the Tanapag community expressed concern about the methodology used in the original May 2000 exposure investigation and whether the results were reliable. Community members asked why the PCB mixture Aroclor 1260 was used as the standard in the original laboratory analysis of the serum, rather than Aroclor 1254. Aroclor 1260 is normally used when analyzing and reporting total PCB concentration in biological specimens like blood serum. Aroclor 1260 is a mixture of individual forms of the PCB molecule called congeners (see "What is a PCB congener" on next page). The mixture of congeners in Aroclor 1260 is similar to the pattern of PCB congeners found in the human body. Aroclor 1260 is routinely used as the analytical standard for comparison of the PCB conger pattern obtained from blood serum and for quantifying the amount of PCB in blood serum since most general population PCB congener profiles are similar to the profile of Aroclor 1260. Given that Aroclor 1260 has been generally used as the pattern reference for determining levels in serum and given the similarity between the Aroclors (especially those with the higher chlorination patterns such as 1254 and 1260), use of 1260 as the standard when the exposure may have been 1254 would not lead to significantly different values. It is ATSDR's determination that the analysis that was performed using the Aroclor 1260 standard would provide an accurate and reliable analysis for total serum PCBs.

Method

The CNMI DPH and Tanapag Village residents jointly determined which individuals were invited to participate. Those who were identified as having the highest serum PCB values in the previous serum sampling were included in the second round of sampling. As a public health investigation, the results were only applicable to these individuals and could not be generalized to other populations.

What is a PCB congener?

PCBs are a mixture of many different forms of the PCB molecule. Each form is called a congener. While there are theoretically 209 potential congeners, only around 36 typically observed in the environment are relevant to human exposure. In blood serum, PCBs are usually measured and reported on a whole weight basis. This approach is an accurate and reliable method for quantifying total PCB in human serum. A more refined approach involves analyzing and reporting the serum concentration as the sum of individual congeners. This approach is more expensive and time consuming, however it is the most accurate method to quantify PCB concentration.

Prior to sampling, CNMI DPH staff met with participants to discuss their questions and concerns. Each adult and a parent or legal guardian of each minor participant was required to sign an informed consent/assent form prior to testing. Confidentiality was protected in accordance with applicable Federal and Commonwealth laws.

A 10 milliliter (ml) tube of blood was drawn from each of the 13 participating individuals. After clotting and spinning, the blood serum was sent to the laboratories of the National Center for Environmental Health, Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia. At the CDC's labs, PCB congener-specific analysis was performed by high resolution gas chromatography/isotope dilution high resolution mass spectrometry.

Serum lipids were extracted and measured by gravimetric analysis.

Results and Discussion

Table 10 provides a comparison of the total whole weight PCB analytical results and the congener-specific analysis. Statistical analysis of the results indicate there is no significant difference between the lipid adjusted total PCB and lipid adjusted congener-specific results. Figure 8 shows the correlation between the May 2000 and October 2001 serum PCB results.

There are many uncertainties and limitations in comparing and interpreting serum PCB measurements. Since PCBs are bound to serum fat molecules (lipids), the results were corrected for serum lipid (blood fat) content. Adjusting serum PCB concentration is important since two individuals could be exposed to similar amounts of PCB and report very different serum PCB concentrations because the serum lipid concentrations were different. Many factors influence blood lipid content, such as lifestyle, age, gender, drug effects and individual genetic variation. It is not scientifically justifiable to compare serum PCB concentration of different individuals or of the same individual at different times unless these variables are properly addressed. This can be problematic given the limited information available on genetic variation of different individuals within populations and more specifically within individuals in populations.

In addition to genetic variations leading to differences in serum lipid concentrations are differences in the metabolism of the individual PCB congeners. These congeners may not be metabolized, excreted or stored in body tissues at the same rate in different individuals. This means that even if two people exposed to the same PCB mixture at identical concentrations it can't be assumed that their average serum PCB concentrations would be the same.

In spite of limitations and uncertainties in interpreting serum PCB levels, normalizing the serum PCB concentrations by adjusting for serum lipids is the most widely accepted method for comparing analytical results. The serum PCB sampling results of the May 2000 sampling and the October 2001 sampling for the 13 individuals are closely correlated (Figure 8). This reinforces the validity of the original May 2000 serum PCB sampling results.

Exposure History Collection and Evaluation

In conjunction with the blood sampling, an exposure history questionnaire was designed to identify the residents who were most likely to have been exposed to PCBs. Specific questions addressed the person's residential history, occupational history, and dietary habits. The questionnaire also involved many questions about the general health status of individuals. ATSDR provided qualified staff to assist DPH staff in collecting the exposure histories.

Methods

The exposure history questionnaire was administered by clinic staff. The exposure history questionnaire can be found in appendix 14.5. Collected information included an individuals' current weekly intake of several food items that were raised or caught in Tanapag. Information was also collected on individuals' history of previous contact with capacitors or areas known to be contaminated with PCBs. The exposure history questionnaire was designed to identify the residents who are most likely to have been exposed to PCBs. These individuals could include (1) residents who live in close proximity (1/4 mile or less) to a known PCB contaminated area, (2) people who engaged in activities that could lead to exposure, e.g., digging graves, gardening, or playing in contaminated areas, and (3) people who had direct contact with leaking fluid from a capacitor or transformer. Individuals' were also asked about their current weekly intake of land crab. Based on environmental sampling, land crab was the only local food to have consistently measurable amounts of PCB. Other foods such as fish, shellfish, eggs, yam and taro were found to have either non-detectable or insignificant concentrations of PCB (Table 6). Thus, consumption of land crab was the only food source that was evaluated for an association with individuals' serum PCB concentrations.

Serum PCB concentrations were evaluated for associations with exposure history information. To evaluate the exposure history variables, individuals were placed in one of two groups, those having serum PCB concentrations less than 5 ppb and those having serum PCB concentrations greater than or equal to 5 ppb. Odds ratios were calculated to characterize the likelihood of having serum PCB concentrations of 5 ppb or greater for a given exposure characteristic. For example, an odds ratio of two for a given exposure variable would indicate that individuals with that exposure are two times more likely to have serum PCBs concentration of 5 ppb or greater.

See appendix 14.6 for a more detailed description of the data analysis methods and an explanation of 95% confidence intervals.

Exposure and medical history information for minors was collected from parents or legal guardians. Each adult and a parent or legal guardian of each minor participant was required to sign an informed consent/assent form, prior to testing. Confidentiality was protected in accordance with Federal and Commonwealth laws.

Results

There were over 1,200 individuals who attended the clinic with ages ranging from less than one year to 93 years of age. Slightly over one-half of participants (51.5 percent) were women. Approximately 11 percent of participants (n = 134) did not have valid serum PCB results or refused to participate in the blood collection portion of the screening activity. Approximately 6.5 percent of the participants (n = 71) had serum PCB concentrations at 5 ppb or greater. There were 15 individuals with serum PCB concentrations at 10 ppb and above, and 6 of those individuals had serum PCB concentrations at 20 ppb and above. All participants under 18 years of age had serum PCB concentrations below 3 ppb. In general, serum PCB concentrations were higher among older individuals and among those with higher measurements of blood lipids (e.g., triglycerides and cholesterol). The average serum PCB concentration of adult men was slightly higher than adult women (Figure 9).

The distribution of serum PCB concentrations were evaluated for each group according their reported weekly consumption of land crab. Table 8 presents the average serum PCB concentration and the high values for each group (e.g., the 90th and 95th percentile). Both the averages and the upper percentiles increase with increasing reported consumption of land crab.

The results for several types of self-reported past exposure are presented in Table 9. Individuals who reported that they consume three or more land crab meals weekly were two times more likely to have serum PCB concentrations of 3 ppb or greater. However, after adjusting for age, sex, and cholesterol, the risk estimate was slightly reduced and it cannot be concluded with sufficient confidence that the observed risk estimate reflects a real excess. For a discussion on statistical significance and 95% confidence intervals, see appendix 14.6. Those reporting some past contact with capacitors or the oil from capacitors and those who reported having dug a grave in Cemetery 2 also had slightly elevated risk estimates. However, none of these self-reported factors demonstrated a statistically significant elevated risk for having serum PCB concentrations of 5 ppb or greater. As presented in the discussion that follows, these odds ratios were not calculated to assess individuals' health risks.

Discussion

The large proportion of non-detects resulted in a skewed distribution of PCB concentrations. By separating the individuals into groups that were above and below a specified level of serum PCB (5 ppb), we were able to assess exposure history variables using logistic regression. There was not sufficient comparison data to determine the level of serum PCB that would be considered elevated in this Pacific island community. The cut point of 5 ppb PCB was based on previous knowledge of serum PCB concentrations among non-occupationally exposed populations. Factors that were found to be associated with having a detectable serum PCB concentration does not imply that these factors are associated with adverse health effects, nor should it be implied that individuals with serum PCB concentrations above 5 ppb are at elevated risk for adverse health effects. This analysis is limited to describing any reported past and/or current exposures that resulted in an increased likelihood of having serum PCB concentrations at or above the defined cut point of 5 ppb.

Associations that are observed in a cross-sectional survey must be interpreted with caution because uncertainties exist. For example, information collected on land crab consumption reflected current eating habits, which may represent an alternation in land crab consumption behaviors over recent years because of concerns about environmental contamination. As such, associations between consumption of some food items and serum PCB concentrations should be considered within the context of recent sampling of the local biota and do not necessarily indicate that clear link between consumption of local foods and serum PCB concentrations exist.

The data also indicated an association between some of the variables reflecting contact with capacitors, capacitor oils, or contaminated soil and having detectable serum PCB levels. There is not adequate information on the nature of these contacts, and it is possible that some individuals' past exposures may be misclassified.

Given that the primary focus of this effort was to offer a public health service to the community, there are some factors that may influence the interpretation of the collected data. First, the need to offer the clinic universally to all former and past residents of Tanapag village resulted in a high level of participation, but there may be some residents and former residents that were not represented in the screening effort. Although there is no evidence that specific geographically- or demographically-defined sub-populations of the village were under represented in the data collection effort, we can not exclude this possibility. Second, the media coverage and educational outreach associated with the clinic operations served a necessary purpose in heightening the community's awareness of the past environmental contamination. Such heightened awareness could affect individuals' recall of past exposures or activities that they perceive as adversely affecting their health. However, we have no reason to believe that over-reporting of exposures would be differential with respect to serum PCB status, and individuals did not know their respective serum PCB concentrations at the time of the exposure history interview.

In conclusion, there is insufficient evidence to indicate that contact with PCBs or PCB-contaminated items in Tanapag may have contributed to the PCB burden of some individuals. As described elsewhere, we cannot infer that these contacts resulted in an added health risk. Moreover, we are unable to assess the relative contribution to PCB body burden of these variables relative to other sources of PCB exposure.

Public health implications of land crab contamination

ATSDR was asked to provide a public health consultation on the public health significance of PCB contamination in land crab tissues collected in the Tanapag Village area. This health consultation evaluated the public health impact of exposure to PCBs in land crabs and made recommendations on actions that local agencies and Saipan residents can take to reduce PCB exposure from contaminated crab. The health consultation specifically addressed the land crab sampling that occurred in Tanapag Village in May and December 2000.

EPA and CNMI DEQ collected 92 land crabs from four harvest areas near Tanapag and a background (reference) location near Smiling Cove (Figure 9). Sampling indicated that PCBs in land crabs ranged from 0.0007 to 0.9600 mg/kg, with an average of 0.0242 mg/kg. A few of the crabs were tested for metals. Three metals (iron, aluminum, and manganese) were slightly elevated in some areas. The metals concentrations in land crabs were below levels of health concern.

ATSDR estimated the amount of exposure to PCB from eating large amounts of land crabs for many years. From a comparison of estimated exposures to conservative health guidelines and the toxicologic and epidemiologic literature, it appears that the PCB exposure from eating land crabs is too low to result in harmful health effects, even to sensitive groups such as children.

Blood tests are the best measure of long term exposure to PCB from all environmental sources, including crab. The average blood serum PCB level among residents of Tanapag who eat crab are below the U.S. average in people with no known source of exposure, and are below levels where harmful health effects would be expected in people.

How can I remove PCB from land crabs?

The easiest way is to remove the liver (the part that has most of the PCB) and to grill, steam or boil the crab(s). Throw away the cooking liquid (PCB comes out of the crab into the cooking liquid).

ATSDR concluded that a ban on eating land crabs does not appear necessary. Although the PCB in Tanapag land crabs does not appear to be a public health hazard, people can be exposed to small amounts of PCB from eating crab. As a precaution and because some data gaps exist regarding PCB exposures and adverse health effects, CNMI DPH and DEQ recommended that Saipan residents stop harvesting and consuming land crabs from the Tanapag area for the time being. In spite of this, people may still choose to harvest and consume Tanapag land crab.

Although ATSDR does not believe that a ban on eating land crabs is necessary, reducing PCB exposures from land crabs is prudent because several data gaps exist regarding PCB exposures and adverse health effects. To reduce PCB exposure, people who choose to eat crab, especially pregnant women, women of reproductive age, and families of small children, should take steps to prepare and cook crab to remove PCB (See "How can I remove PCB from land crabs?").

ATSDR released a health consultation in July 2001 evaluating the health impact of land crab contamination. The complete document is included in Appendix 14.7.

Summary determination of adverse health effects from exposure to PCB in Tanapag Village

The available evidence indicates that the PCB contamination in Tanapag has not adversely affected the health of village residents. This determination is based upon several facts; 1) From the results of the exposure investigation, it appears that the residents have levels of PCB in their serum which would not be expected to cause clinically significant adverse health effects; 2) The medical screening did not observe any clinical signs of illness that could be associated with PCB exposure; 3) Evaluation of the exposure history database revealed that there is insufficient evidence to indicate that contact with PCBs or PCB-contaminated in Tanapag items may have contributed to elevated serum PCB levels.

It is possible that past exposures could have been higher. ATSDR is unable to determine the degree of past exposure and determine the public health impact with much certainty. Analysis of serum PCB is the best measure of cumulative exposure from all sources. Results indicate that over the past several years, Tanapag residents have not experienced unusually high exposure. On average, serum PCB levels in Tanapag are within the range of U.S. background levels. A few residents have elevated serum PCB levels, indicating a higher degree of past exposure than the rest of the village population. It is not known with much certainty what the specific source of exposure was or when it occurred. The elevated blood levels are not high enough to result in adverse health effects.

There may be several reasons why there is no evidence of high exposure in spite of significant soil contamination. The areas of high soil contamination were isolated in a few locations (Figure 5). Human contact with contaminated soil was likely intermittent and infrequent. Many contaminated areas were under vegetative growth and people did not have the opportunity to come into contact with them on a regular basis. Past removal of capacitors and contaminated soil reduced the degree of further contamination and the potential for exposure.

Exposure activities and the concentration of the contaminant both play an important role in determining the amount of PCBs to which a person is exposed. However, a variety of other factors are involved that determine whether environmental contamination will result in significant exposure. Some of these factors include:

• duration of exposure: when the contamination occurred and how long residents have lived there.
• frequency of exposure: how often the person has contact with the soil.
• area of contamination: does the person come into contact with the highest level of PCBs all the time?
• bioavailability (potential for absorption from the gastrointestinal tract).

The main source of PCB exposure for people who do not work with PCBs is from eating foods containing small amounts of PCB (ATSDR2000a). No significant PCB contamination of local foods was detected. Except for a few individuals, past PCB intake appears to be within the normal range of the rest of the mainland U.S. population.

In order for a chemical in soil to produce a hazard, there must first be contact, or exposure. It then must get into the body. Only a fraction of the chemical that is measured in a soil may be truly available to impact the health of humans. If someone were to eat contaminated soil accidentally or on purpose, only a small fraction of the PCB in contaminated soil is typically absorbed into the body through the digestive tract. 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. A discussion of the oral bioavailability of PCBs in soil is presented in appendix 14.4.

Practical bioavailability of PCBs comes from actual field experience in humans living near hazardous waste sites. Studies involving populations living on or near areas of soil contamination in the absence of consuming contaminated fish have not found any signs of PCB-induced illness or any documented elevation in serum PCB. Poor skin and digestive system absorption of PCBs in soil and little or no direct contact may explain the lack of PCB uptake from contaminated soils. 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 the soil was highly contaminated with PCB(Stehr-Green 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).

Actions taken to investigate and remove PCB contamination in Tanapag were appropriate and necessary to protect public health and to reduce the potential for current and future PCB exposure. The fact that some residents' blood tests revealed elevated serum PCB levels indicated that in the past, exposures occurred and the PCB contamination posed a public health hazard.

Contaminated soil has been removed from Tanapag Village, eliminating a major exposure source. As a result of the clean up, the PCB contamination does not pose a current or future health hazard to Tanapag Village residents. Some small degree of exposure may continue if people choose to consume contaminated land crab, however this exposure would not be likely to result in health harm.

Community Concerns

Tanapag Village residents expressed several health concerns related to PCB contamination in their community. ATSDR and CNMI DPH responded to their concerns both verbally and in writing at public meetings, during private consultations and in letters, fact sheets and health consultation documents. The section below summarizes community concerns expressed to ATSDR and the response provided to the community.

How long ago did you have to be exposed for PCBs to show up in blood? Tissue? Breast milk?

With the exception of a few highly chlorinated biphenyl congeners, the elimination half-life (the time for half of a substance to be eliminated) in people ranges from 1 to 6 years. Elimination half-lives for total higher-chlorinated PCBs can range from 8 to 24 years.

If blood results are out of the normal range, is it safe to breast feed my baby?

Although exposure is possible from ingesting breast milk, currently there is no convincing evidence that PCB exposure via breast milk harms a nursing infant. PCBs detected in breast milk (or blood) are not necessarily an indication that breast feeding should be stopped. Benefits of breast feeding include fewer ear infections, higher immunity from diseases, and improved nutrition. In most cases, the benefits to a baby from breast feeding outweigh the risks of PCB exposure from small amounts in breast milk.

If blood results are out of the normal range, is it safe to have children?

Regardless of PCB exposure, a health care provider should be consulted before making this decision. Exposure to PCB in Tanapag appears to be within normal ranges to the majority of the population. There is no evidence that exposure to PCBs in Tanapag would result in clinically significant harm to an unborn baby.

Health effects of PCB exposure to unborn babies that have been studied include:

Birth Defects

- No studies have reported structural birth defects in humans.

Birth Weight

- Highly exposed workers showed a slight, but clinically insignificant effect on birth weight.

Gestational Age

- Highly exposed workers showed a slight, but clinically insignificant effect on gestational age.

Developmental Delays

- Substantial data suggest that PCBs play a role in neurological development, but the observed alterations are subtle. In one study, the delays resolved by age 4 and in the other, lower full-scale and verbal IQ scores persisted even at 11 years of age. Further research is ongoing

Immune System

- Human studies provide evidence of immune system toxicity in infants exposed to PCBs in the womb and/or via breast milk. However, this evidence is limited because of mixed chemical exposures and insufficient information on exposure-response relationships.

Thyroid hormones

- Human studies provide some evidence of a slight, but clinically insignificant effect on thyroid hormones. Further research is ongoing.

If blood results are out of the normal range, should breast milk or fat tissue be sampled?

PCBs will be present in blood, adipose (fat) tissue, and breast milk. Detecting PCBs out of the normal range does not mean there will be health effects. Additionally, there is no comparison between blood, tissue, and breast milk levels so additional biological results would not be medically meaningful.

If physicians are interested in talking with other physicians either involved in exposure investigations or who have had to consult with patients concerned about interpreting their blood levels, they can contact ATSDR's Exposure Investigation Section toll-free at 1- 888-42-ATSDR.

Is there treatment available for PCB exposure?

No, not at this time. People with high levels should have a careful exposure history taken and increase efforts to identify and eliminate any current sources of PCB exposure. As sources of exposure are eliminated, the amount of PCB in the body will decrease over time.

If my children play outside, will they get sick from PCB exposure?

No. For most people, exposure to PCBs occurs primarily through eating contaminated foods. Contact with contaminated soils are unlikely to be an important contributory source to an individual's PCB body burden (ATSDR 2000a).

To reduce the possibility of exposure, it is a good idea to wash your hands before touching the face or eating, and cleaning your feet or leaving shoes outside before coming into the house. To reduce the possibility of exposure to soil contamination, children should avoid areas marked as hazardous and avoid touching stained, oily soils.

The Army Corps of Engineers has removed contaminated soils in Tanapag Village and the cemetery and has determined that previously remediated areas are safe.

Is it safe to visit the cemetery?

Yes. The Army Corps of Engineers located the soil contamination in the cemetery and has completed the removal and treatment of soil containing greater than 1 ppm of PCBs (ECC 2003).

You can further reduce the possibility of exposure to PCBs after visiting or working in the cemetery by washing your hands before eating or touching something that will go into your mouths and cleaning the soles of your feet or shoes before going inside homes or buildings.

Is it safe to eat fish and crabs?

Fish are a healthy, nutritious food and are an important part of the diet of the Tanapag people. ATSDR has determined that the fish and crabs safe to eat. Because there are many benefits to eating them, fish and crabs should continue to be part of a healthy diet.

Families can do many things to reduce the possibility of exposure to PCBs in fish and crabs, including:

• Select younger, smaller fish and crab.
• Remove the skin and fatty tissue in the belly and along the sides.
• Bake or broil the fish, and throw away the fatty juices and drippings (PCBS are in the fat).
• Avoid eating the liver and other internal organs of the fish and crabs.

Are foods like taro and yam safe to eat?

It is unlikely that taro and yam would contain substantial levels of PCBs because PCBs strongly stick to soil particles and plants do not easily take up PCB from the soil. Several studies have shown that plant uptake of PCBs is negligible. Just to be sure, EPA sampled plant tissues to determine if these foods were contaminated with PCBs. Tiny amounts of PCB were detected however they were far below levels of health concern.

People harvesting and consuming taro and yam can reduce the potential for exposure by washing the vegetables carefully and peeling before eating. Washing hands after contact with the soil would also reduce the possibility of accidentally eating PCBs and getting them on the skin.

Next Section    Table of Contents