PUBLIC HEALTH ASSESSMENT

ANDERSEN AIR FORCE BASE
YIGO, GUAM

APPENDIX C: ATSDR'S EXPOSURE EVALUATION PROCESS

Estimates of Human Exposure Doses and Determination of Health Effects

Deriving Exposures Doses

After identifying contaminants in site media above comparison values, ATSDR further evaluatesexposures to these contaminants considering information about exposures combined with scientificinformation from the toxicological and epidemiological literature. If necessary, ATSDR estimatesexposure doses, which are estimates of how much contaminant a person is exposed to on a daily basis.Variables considered when estimating exposure doses include the contaminant concentration, theexposure amount (how much), the exposure frequency (how often), and the exposure duration (howlong).

The estimated exposure doses can be used to evaluate potential noncancer and cancer effectsassociated with contaminants detected in site media. When evaluating noncancer effects, ATSDRcompares the estimated exposure dose to standard toxicity values, including ATSDR's minimal risklevels (MRLs) and the U.S. Environmental Protection Agency's reference doses (RfDs), to evaluatewhether adverse effects may occur. The chronic MRLs and RfDs are estimates of daily humanexposure to a substance that is likely to be without appreciable risk of adverse noncancer effects overa specified duration. The chronic MRLs and RfDs are conservative values, based on the levels ofexposure reported in the literature that represent no-observed-adverse-effects levels (NOAEL) orlowest-observed-adverse-effects-levels (LOAEL) for the most sensitive outcome for a given route ofexposure (e.g., dermal contact, ingestion). Uncertainty (safety) factors are applied to NOAELs orLOAELs to account for variation in the human population and uncertainty involved in extrapolatinghuman health effects from animal studies. ATSDR also reviews the toxicological literature andepidemiology studies to evaluate the weight of evidence for adverse effects.

When evaluating the potential for cancer , ATSDR uses a weight of evidence approach to determinewhether cancer effects are likely or not from an exposure to a toxic agent. In it's evaluation, ATSDRconsiders toxicologic and epidemiologic literature as well as cancer slope factors (CSF) that define therelationship between exposure doses and the likelihood of an increased risk of developing cancer overa lifetime. The CSFs are developed using data from animal or human studies and often requireextrapolation from high exposure doses administered in animal studies to lower exposure levelstypical of human exposure to environmental contaminants. The CSF represents a theoretical, upper-bound estimate of the probability of developing cancer at a defined level of exposure; therefore, theytend to be very conservative (i.e., overestimate the actual risk) in order to account for a number ofuncertainties in the data used in extrapolation. ATSDR also considers the cancer effect levels (CELs)reported in the literature. The CEL is the lowest dose of a chemical in a study, or group of studies,that was found to produce increased incidences of cancer (or tumors) in animals.

Estimating Exposure Doses from Ingesting Drinking Water from Base Wells

VOCs have been detected in Andersen AFB water supply wells MW-1, MW-2, and the Tumon-Mauiwell at concentrations greater than ATSDR comparison values and EPA's maximum contaminantlevels for drinking water. To determine whether exposure to these contaminants in the well water isrelated to adverse health effects ATSDR estimated exposure doses for people consuming watercontaining the highest measured concentrations in the wells (39 ppb of TCE and 10 ppb of PCE). Theestimated exposure doses were then used to estimate potential noncancer outcomes.

In estimating to what extent people might be exposed to contaminants, ATSDR used "conservative" orsafe assumptions about possible human exposure and any associated health effects. ATSDR assumedthat a person drank the most contaminated well water. ATSDR also used conservative assumptionsabout how often people drink water and how much they drink. For example, ATSDR assumed that atypical adult drank 2 liters of water each day and weighed 70 kilograms (kg) and that a child drank 1liter of water each day and weighed 10 kg. Because ATSDR does not know with certainty how longexposure may have occurred, ATSDR estimated an exposure period of 30 years for an adult and 6years for a child to calculate maximum exposure doses. These assumptions likely overestimate actualexposure because water from the wells was blended with other water in the distribution system andmilitary employees drinking the TCE- and PCE-contaminated water were unlikely to be exposed formore than a year or two due to the relatively short duration of military duty stations. Furthermore, theactual exposure period was likely much shorter than 30 or 6 years, because the contaminated well wastaken off line once the contamination was detected. The conservative assumptions, however, allowATSDR to estimate the highest possible exposure dose and determine the corresponding health effects.Although ATSDR expects that few individuals, if any, were exposed to the highest levels ofcontamination, the "conservative" estimates are used to protect public health. Also, as a reminder,Guam residents and the general public did not use drinking water from the military supply system.

Noncancer: The resulting adult and child exposure doses for PCE are lower than its ATSDR MRL of0.01 mg/kg/day. No chronic oral MRL or RfD is currently available for TCE. ATSDR recentlywithdrew the intermediate MRL and no chronic MRL or RfD exists for TCE. The study on which theintermediate MRL was based has been questioned because it contains certain flaws and limitations(e.g., the exact amount of TCE-contaminated water consumed by laboratory animals in the study isuncertain). For comparison, ATSDR reviewed the available toxicologic literature to determinepossible adverse effects associated with exposure at doses estimated for this pathway. On the basis ofthis review, the exposure doses estimated for TCE by ATSDR are several orders of magnitude lowerthan the lowest doses reported in the toxicologic literature capable of producing noncancer effects inexperimental animals administered oral doses of TCE (ATSDR 1997). Therefore, drinking watercontaining the highest detected levels of TCE and PCE reported in the wells is not likely to result inadverse noncancer effect.

Cancer: TCE and PCE have been shown to cause cancer in laboratory animals given large doses. Thelink between TCE or PCE and cancer in humans drinking water is controversial, however. Availablestudies are inconclusive and the data are inadequate to establish a link. EPA is currently reviewing thescientific literature pertaining to the carcinogenicity of TCE and PCE to determine its cancerclassification. Some studies have shown that individuals drinking TCE-contaminated water with up to220 ppb (a concentration over 24 times greater than the maximum level detected at Andersen AFB)suffered no increased incidence of cancer (Vartianinen et al. 1993; ATSDR 1997a, b).

ATSDR concludes that there is no apparent public health hazard associated with drinking water inthe past from MW-1, MW-2, and the Tumon-Maui well.

Estimating Exposure Doses from Ingesting of Local Biota

Metals, pesticides, and SVOCs were measured in samples of crops grown or game grazing atAndersen AFB. People regularly consume papaya fruit and other edible fruit grown on and off base.Some Guam residents recreationally hunt game around Andersen AFB. ATSDR evaluated potentialhuman health hazards associated with ingesting contaminated biota. Chemical exposure doses wereestimated using conservative consumption rates for a 70 kg adult (2 grams per day of Sambar deerand monitor lizard; 20 grams per day of wild pig; and 340 grams per day of papaya) and exposurefrequencies (365 days per year for a 30-year period).

Of all the chemicals analyzed, only arsenic (0.00044 mg/kg/day) exceeded its respective MRL(0.0003 mg/kg/day). No MRL or RfD exists for aluminum, an essential human nutrient. Aluminumconcentrations in Guam biota, however, appeared elevated compared to normal backgroundconcentrations. Aluminum concentrations in on-base deer and pig tissue were slightly above normalbackground concentrations in mammals (1 to 5 ppm) as reported by Puls (1989), but detectedconcentrations in off-base deer and pig tissue were within normal background levels. ATSDR furtherexamined the toxicologic literature to evaluate whether health effects were likely to occur at the detected levels of arsenic and aluminum.

Arsenic

ATSDR found that the estimated exposure doses for consumption of on-base produce wereapproximately 30 to 100 times lower than the lowest observed effect levels for chronic, oraldoses of arsenic in humans (ATSDR 1993). Additionally, several epidemiologic studies ofchronic, oral arsenic exposure report no health effects at average chronic doses of0.0004-0.01 mg/kg/day (Mazumder et al. 1988; Valentine et al. 1985; Cebrian et al. 1983;Southwick et al. 1981; Harrington et al. 1978). It should be noted that estimated exposuredoses and excess cancer risk were evaluated assuming that all produce (fruits and vegetables)consumed by Guam residents contained arsenic levels equal to the maximum arsenic leveldetected in on-base papaya. This assumption likely overestimates actual exposure. Access toAndersen AFB is restricted and no on-base areas are commercially farmed. Most residents,therefore, probably obtain their fruits and vegetables from areas not grown on Andersen AFBproperty. No off-base papaya samples had detectable levels of arsenic in their edible parts. Theconsumption of produce grown off-base, therefore, would not pose a public health threat. Theestimated exposure dose from ingesting arsenic in on-base produce (0.00044 mg/kg/day) onlyslightly exceeds ATSDR's chronic oral MRL of 0.0003 mg/kg/day.

Aluminum

Aluminum non- and off-base papaya slightly exceeded background concentrations inunprocessed produce (0.1 to 7.16 ppm) as determined by Schenck et al. (1989). All monitorlizard aluminum concentrations were elevated. Over three-fourths of exposure to aluminum inlocal biota results from the ingestion of monitor lizard tissue. No island residents claimed,however, to eat monitor lizard when asked during the Guam diet survey (EA Engineering1995). Therefore, exposure from the consumption of monitor lizard tissue appears unlikely,and, if it occurs at all, limited.

Due to restricted access to base properties, public exposure to contaminants from local biota ismore likely to result from the consumption of off-base food sources. The estimated cumulativeexposure to aluminum from analyzed off-base biota sources (0.0036 mg/kg/day) is 10 timeslower than exposure from on-base biota.

The toxic potential of aluminum is extremely low compared to many other metals (ATSDR1997c). If aluminum is chronically ingested, it may interfere with the body's up-take ofcalcium and phosphorous, however, the retention of aluminum in healthy mammals,specifically those without kidney dysfunction, is minimal. Even when dietary levels are high,aluminum concentrations in tissues do not reflect this increased exposure because mostaluminum is excreted in the feces (ATSDR 1997c). Therefore, ATSDR concludes thataluminum is not likely to pose a noncancer public health hazard.

The EPA has not classified aluminum for human carcinogenicity, but ATSDR found nostudies regarding cancer in humans after oral exposure to aluminum or aluminum compounds(ATSDR 1997c). Therefore, ATSDR concludes that exposure to aluminum in Guam biotawill not pose a carcinogenic public health hazard.

Using current toxicologic information, ATSDR concludes that there are no apparent health hazards(past, current, or future) associated with consumption of local biota.

Evaluation of Radon Exposure

According to EPA, exposure to radon is a national environmental health problem and is not isolated toGuam (EPA 1992). Elevated radon levels have been discovered in virtually every state (Air Chek1998). The EPA estimates that as many as 8 million homes throughout the country have elevatedlevels of radon. To date, state surveys show that 1 out of 5 homes in the United States has elevatedradon levels (above 4 pCi/L) (Air Chek 1998).

Toxicologic studies report that radon exposure causes no adverse health effects from short termexposure. The primary health concern associated with residential radon exposure is lung cancer,although there is currently no clear evidence that radon exposure causes lung cancer. A recent reportfrom the National Research Council estimates that approximately 1 in 7 of all lung cancer deaths canbe attributed to radon exposure, independent of smoking status, though these estimates are uncertain (BEIR VI 1999).

Many factors influence the risk of lung cancer resulting from radon exposure. Among these are theradon level, the duration of exposure, the time since initiation of exposure, the age of an exposedindividual, and the individual's smoking habits. The combined effects of cigarette smoking and radonexposure place current and former smokers at particularly high risk for lung cancer.

Epidemiologic studies show that individuals working in certain industries susceptible to radonreleases are at greatest risk, because they are often exposed to high levels of radon over an extendedperiod of time. In one study, uranium miners exposed to radon levels of 50 to 150 pCi/L in air forabout 10 years have shown an increased frequency of lung cancer (ATSDR 1990), though this studysuffers from several weaknesses including lack of control for exposures to other agents that couldcontribute to lung cancer, such as silica and smoking. In the past, some housing units at AndersenAFB contained radon levels above 50 pCi/L, but the duration of exposure to these levels was probablycloser to 2 years (not 10 years), the average tour length on Andersen AFB.

The most effective methods of lung cancer prevention are to reduce radon exposure and to modifyother risk factors, such as smoking (ATSDR 1992). Fortunately, the Air Force has been activelymitigating on-base housing levels of radon since 1989. Andersen AFB plans to continue its radontesting and mitigation of residential units in the future, as well as expand its base program to other,lower priority buildings.

ATSDR was unable to fully assess potential health hazards (if any) associated with past radonexposure at Andersen AFB. First, the full extent of past radon exposure at Andersen AFB remainsunknown due to limited historical sampling data. Second, ATSDR does not have health-basedcomparison values for radon and EPA has not identified an inhalation reference concentration forradon. Moreover, EPA's carcinogen assessment summary for radon (formerly determined a humancarcinogen) has been withdrawn pending further review. ATSDR found no clear evidence that long-term exposure to radon at levels that are normally present in the environment (1 to 3 pCi/L for average outdoor air levels) is likely to result in harmful health effects.

APPENDIX D: BIOTA TABLES

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million
NA = not applicable

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.
** Copper and lead values are biased since the animals were shot with copper-plated lead pellets.

ppm = parts per million

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.
** Copper and lead values are biased since the animals were shot with copper-plated lead pellets.

ppm = parts per million
ND = not detected

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million

Source: EA Engineering 1995

* The mean was calculated using censored data (nondetects were assumed to be equal to 50% of the reported detection limit) when the frequency of detection was less than 100%.

ppm = parts per million

APPENDIX E: RESPONSE TO PUBLIC COMMENT

The Agency for Toxic Substances and Disease Registry (ATSDR) released the Andersen Air ForceBase (Andersen AFB) Public Health Assessment (PHA) on September 27, 2001, for public reviewand comment. That public comment period ended November 30, 2001. The following documentsATSDR's response to comments and questions received during the public comment period.

1. Comment: The commenter asked about health risks associated with eating coconut crab, apopular wildlife food trapped in the study area. The commenter is concerned thatcontaminants in soil, such as polychlorinated biphenyls (PCBs), accumulate in portions of thecrab (the fatty tissue) that are eaten by local residents.

Response: No data on coconut crabs were available for ATSDR's review during thepreparation of this public health assessment. Despite the lack of data, ATSDR believes thatpeople are not at risk of exposure to unhealthy levels of Andersen AFB-related contaminantswhen they consume coconut crabs. In making this determination, ATSDR reviewed relevant,supplemental information, including bioaccumulation data for other native biota, areas of soilcontamination, and possible harvesting areas.

Both on-site and off-site wildlife data were available for Sambar deer, wild pig, snakes, andlizards. Although not specific to coconut crab, these data help us assess whether wildlife, ingeneral, are accumulating contaminants from Andersen AFB and, if so, whether contaminantsare accumulating at levels that might be harmful to the consumer. For all cases, ATSDRfound no apparent health hazards associated with the wildlife as a source of food.

Several areas of soil contamination were noted at Andersen AFB, but PCBs were most oftenreported in soil at the Main Base. Access restrictions are enforced at sites on the Main Base,thus essentially preventing people from harvesting crabs in the areas of likely PCBcontamination.

Together, this information suggests that the wildlife are not accumulating contaminants atlevels of concern and the soil contamination and harvesting areas are not co-located. Still, ifbiota sampling is conducted in the future, ATSDR recommends as a prudent measure that theprogram include sampling of coconut crabs.

2. Comment: The commenter expressed concern about high levels of trichloroethylene (TCE)and tetrachloroethylene (PCE) in groundwater and the potential for future exposure and healthrisk. The commenter also questioned how much was known about the extent of groundwatercontamination and whether actions will be taken in the future to further evaluate the extentand to remove contamination.

Response: TCE and PCE have been detected in groundwater beneath Andersen Air ForceBase, but ATSDR believes that sufficient regulatory mechanisms are in place to trackgroundwater contamination and prevent potential harmful exposures in the future.

• The Air Force, Guam Environmental Protection Agency (GEPA), and U.S.Environmental Protection Agency (EPA), agencies overseeing remediation at the base,are well aware of the groundwater contamination issue at Andersen AFB. With GEPAand EPA oversight, Andersen Air Force Base (Andersen AFB) routinely collectsgroundwater samples from a network of groundwater monitoring wells to trackgroundwater contamination. Through this effort, Andersen AFB has identified thehighest levels of TCE within base property, at the northwest side of the MARBOAnnex (in the Yigo Subbasin) near the Waste Transfer Stations. Additionally, elevatedlevels of PCE have been measured in monitoring wells adjacent to the MARBOAnnex laundry. Andersen AFB has committed to a long-term groundwater monitoringprogram to ensure that contamination does not threaten local or base water supplies inthe future.


• Andersen AFB has removed contaminated soil (and a potential source of TCE andPCE to the underlying groundwater) from certain areas of the MARBO Annex.


• Andersen and municipal suppliers are required by law to test water supplies forpollutants and to comply with GEPA and EPA drinking water standards. Pastmonitoring efforts revealed that volatile organic compound (VOC)- contamination hadentered three Andersen base water supply wells. In response to this finding, Andersenwithdrew two of the wells from service indefinitely. Air stripping units are in place toremove VOC contamination, however, if and when these wells are restored to service.(VOCs concentrations in the third well had dropped to levels well below ATSDR'scomparison values and EPA's maximum contaminant levels.) Except for these wells,no other drinking water wells have been or are likely impacted by VOC contaminationbecause either: (1) contamination is not present upgradient of the well or (2)contamination, though present upgradient of the active well, is at relatively low levels.Water from all other operating base wells and from nearby municipal water wells willcontinue to be tested to ensure that the water delivered to customers is safe fordrinking.


• GEPA maintains a wellhead protection program to prevent contamination fromentering drinking water wells. Under this program, GEPA curtails development withina 1,000- foot radius of a drinking water well and restricts installation of water supplywells on property impacted by TCE and PCE.



ATSDR believes that these collective measures will help protect the aquifer beneath Andersen AFB and safeguard the quality of drinking water for future use.

Table of Contents