U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Contaminated Drinking Water at Camp Lejeune. Contaminated Water Supplies at Camp Lejeune: Assessing Potential Health Effects. Washington (DC): National Academies Press (US); 2009.

Cover of Contaminated Water Supplies at Camp Lejeune

Contaminated Water Supplies at Camp Lejeune: Assessing Potential Health Effects.

Show details


In the early 1980s, two water-supply systems on the Marine Corps Base Camp Lejeune in North Carolina were found to be contaminated with the industrial solvents trichloroethylene (TCE) and perchloroethylene (PCE). The water systems were supplied by the Tarawa Terrace and Hadnot Point water-treatment plants, which served enlisted-family housing, barracks for unmarried service personnel, base administrative offices, schools, and recreational areas. The Hadnot Point water system also served the base hospital and an industrial area and supplied water to housing on the Holcomb Boulevard water system (full-time until 1972 and periodically thereafter).

PCE was the primary contaminant found in the wells serving the Tarawa Terrace system. The chemical was used by an off-base dry cleaner (ABC One-Hour Cleaners), and the groundwater became contaminated with PCE as a result of spills and improper disposal practices. Contamination of the wells from that source is estimated to have begun as early as 1953, the year when dry-cleaning operations began. There were also other on-base sources of contamination in the Tarawa Terrace system that had a smaller impact on the water supply. The contamination of the Hadnot Point water supply was more complex and involved multiple sources and multiple contaminants. The primary contaminant found in those wells since monitoring began in the 1980s was TCE. It is likely that multiple sources contributed to the TCE contamination, including on-base spills at industrial sites and leaks from underground storage tanks and drums at dumps and storage lots. The Hadnot Point water-treatment plant began operating in 1943, but no estimates have yet been made of when the contamination began. Wells in both systems that were contaminated in the early 1980s were closed in the period November 1984–May 1985, and the entire Tarawa Terrace water-treatment plant was closed in 1987.

There has been considerable public controversy over the potential health consequences for former residents who were exposed to the contaminated water. TCE and PCE are known to have toxic effects in animals and in humans, so it is important to understand the scale and extent of exposure that occurred at the base to assess effects on the health of former residents. Only a few studies have been performed specifically on former residents of the base. To supplement those evaluations and to help to inform decisions about addressing health claims, the U.S. Navy was directed by Congress (Public Law 109-364, Section 318) to ask the National Research Council to address independently questions about whether any health outcomes are associated with past contamination of the water supply at Camp Lejeune. The National Research Council assembled a multidisciplinary committee of environmental scientists, toxicologists, epidemiologists, and biostatisticians to review the scientific evidence on associations between adverse health effects and historical data on prenatal, childhood, and adult exposures to contaminated drinking water at Camp Lejeune. The committee was asked to focus its attention on toxicologic and epidemiologic literature on TCE and PCE and to consider studies of Camp Lejeune residents and other populations exposed to the contaminants of concern and proposals for additional studies of Camp Lejeune residents.

To address its task, the committee divided its investigation into two major categories: assessing exposure to contaminants in the water supply and assessing the possible health effects associated with the contaminants. The reviews were then integrated to ascertain whether conclusions could be drawn about the likelihood that outcomes in people who lived or worked in the affected areas of the base were caused by the contaminated water supplies. The contribution of past and current studies of the Camp Lejeune population was evaluated, as was the potential contributions of future research on this population.


To understand the exposures that occurred because of the contamination of water supplies at Camp Lejeune, it is important to characterize the contamination—including its location, magnitude, duration, and variability—and the individual water-use patterns and other water-related behavior of the population that was exposed. The first component involves identifying the contaminants of concern, their sources, and their estimated concentrations in any particular water-supply system over time. The second component is to characterize how members of the population may have been exposed to the contaminated water supply at home, at work, and in other settings through water consumption, dermal contact, and inhalation of volatile compounds during showering, bathing, dishwashing, and other activities. Such factors are important determinants of exposure and are likely to vary widely in the population.

Water-Supply Contamination

The Tarawa Terrace and Hadnot Point water-supply systems began operating in 1952 and 1943, respectively. From a conceptual standpoint, their operations were similar. Water-supply wells collected groundwater and pumped it to a water-treatment plant. The wells were “cycled,” meaning that only a subset of wells pumped water to the treatment plant at any given time. A few wells on both systems were contaminated. When those wells were operating, they delivered contaminated water to the treatment plant, where it was mixed with water from other wells and processed before being distributed on the base. Over the years, wells were added and some were taken temporarily offline or were closed for various reasons. Thus, concentrations of contaminants to which people were exposed varied substantially on a short-term and long-term basis.

The residential areas served by the two water systems were primarily enlisted family housing and barracks for unmarried service personnel. Thus, many of the exposed were young families and people of reproductive age. The population was also transient, with some people living on the base for a few months for training or for a few years for longer assignments.

Tarawa Terrace

The committee reviewed the available data on the exposures that occurred at Camp Lejeune. For Tarawa Terrace, the Agency for Toxic Substances and Disease Registry (ATSDR) performed a historical reconstruction of contamination scenarios and used its model to estimate the concentrations of chemical contaminants that occurred during different periods. ATSDR’s historical reconstruction involved investigation into operations of the off-base dry cleaner, on-base operations, operation of water-supply wells and water-treatment plants, water-monitoring data, groundwater flow, and other data relevant to providing a chronology of events related to the contamination. The primary contaminant identified as present at Tarawa Terrace is PCE. PCE is typically degraded by natural processes in the soil and groundwater to TCE, trans-1,2-dichloroethylene (1,2-DCE), and vinyl chloride. Groundwater models were used to reconstruct the migration of PCE from the dry cleaners to the water-supply wells serving Tarawa Terrace, and then mixing models were used to predict monthly concentrations of PCE and its degradation products in finished water (groundwater that was treated at a water-treatment plant for delivery to residences) from 1957 to 1985. Because the models were based on several simplifying assumptions and were calibrated by using a small number of water-quality measurements taken during a narrow window (1980-1985) of the total contamination period, considerable uncertainty is associated with the predictions. Some of the uncertainty was characterized when ATSDR performed statistical analyses to calculate the probability that its exposure estimates were reasonable. To gain some perspective on its estimates, ATSDR compared its monthly estimates with the U.S. Environmental Protection Agency (EPA) maximum contaminant level (MCL) for PCE in drinking water of 5 μg/L that was established in 1985. The model estimated that starting in No vember 1957, the concentration of PCE delivered to residents exceeded that MCL and remained well above it until the wells were closed in 1985.

The committee concluded that ATSDR applied scientifically rigorous approaches to address the complex groundwater-contamination scenario at Tarawa Terrace. The outcome of the modeling was monthly estimates of the concentrations of contaminants in the water supply to which people could have been exposed. Although ATSDR recognized and tried to account for the limitations and uncertainties associated with its models, the committee judges that—because of the sparse set of water-quality measurements, the need to make unverifiable assumptions, and the complex nature of the PCE source—it is virtually impossible to estimate exposure to historical levels of PCE and its degradation products accurately. Reporting precise values based on model predictions gives the misleading impression that the exposure of the former residents and workers at Tarawa Terrace during specific periods can be accurately defined. It is the committee’s judgment that ATSDR’s model is best used for estimating exposure categories qualitatively. From that perspective, a single exposure category of “exposed” appears to be applicable to persons who resided or worked at Tarawa Terrace during 1957-1985.

Hadnot Point

The water-supply contamination scenario for Hadnot Point is much more complex than that for Tarawa Terrace because there were multiple sources and contaminants. The extent of contamination has not yet been characterized, inasmuch as historical reconstruction or groundwater modeling has not yet been performed for Hadnot Point. The committee therefore relied on site descriptions of source areas, laboratory reports and other documentation of supply-water sampling, and results of monitoring of groundwater wells that were installed as part of remedial investigations to characterize likely exposures. Numerous sites have been identified as possibly contributing to the contamination of the groundwater, including an industrial area, a drum dump, a transformer storage lot, an industrial fly-ash dump, an open storage pit, a former fire training area, a site of a former on-base dry cleaner, a liquid-disposal area, a former burn dump, a fuel-tank sludge area, and the site of the original base dump. TCE appears to be the primary contaminant of concern on the basis of measurement data from the 1980s, but many other chemicals had the potential to contaminate the water supply, given the nature of activities at sites near the supply wells. Other chemicals measured in the water supply included PCE, vinyl chloride, 1,1-DCE, 1,2-DCE, methylene chloride, benzene, and toluene. Sampling performed in the early 1990s as part of remedial investigations also detected metals in monitoring wells, but little if any metal analysis was conducted for the timeframe of interest (1943-1985), and the committee did not review such data. Qualitative evidence suggests that the potential magnitude of groundwater contamination appears to have been much higher at Hadnot Point than at Tarawa Terrace.

ATSDR plans to perform a historical reconstruction of estimates of the concentrations of water-supply contaminants at Hadnot Point similar to the one performed for Tarawa Terrace. On the basis of its review of Hadnot Point water-system contamination, the historical groundwater modeling performed for Tarawa Terrace, and ATSDR’s preliminary plans for historically reconstructing exposures that occurred at Hadnot Point, the committee recommends that simpler models be used instead of complex groundwater models. In particular, the use of conceptual models based on hydrogeologic characterization studies coupled with mass-balance calculations or analytic models should be given serious consideration because they can be performed relatively quickly and can be used to achieve a crude characterization of the degree and timeframe of contamination of the aquifer. Groundwater-modeling studies using public-domain MODFLOW-family tools should be performed only after establishing a clear need for a study. To support further analyses, the committee also recommends that the Marine Corps create and maintain a comprehensive public database of water-quality measurements for all environmental media samples collected across the base in the course of investigating the nature and extent of contamination at Camp Lejeune. The database should include information on where samples were taken, sampling dates, analytes meas ured, laboratory quality-control information (including limits of detection), and other information relevant to exposure assessment.

Water-Use Patterns and Behavior

Places and dates of residence are key determinants of likely exposure at Camp Lejeune, but individual behaviors also affect the magnitude of exposure. Such behavior includes water consumption, showering or bathing patterns, and other water-related behavior (such as dishwashing). Such information is not available in archival records, and it is far too remote in time for accurate recall. A study in progress evaluating birth defects and childhood cancers is collecting self-reported water-use information from surviving mothers of offspring in the study, but the data are not yet available. The contaminated water systems also supplied nonresidential areas of the base, including schools, workplaces, recreational areas, and a hospital. Water-use patterns and behavior in those setting are expected to differ substantially from residential uses and behavior. In addition, the residential and nonresidential exposures could overlap, and people could have been exposed to contaminated water at multiple locations.


The committee considered a wide spectrum of potential health effects that are known or suspected to be associated with TCE and PCE by surveying the scientific literature on the contaminants and the health problems reported by former residents and workers of Camp Lejeune. The scientific literature reviewed included reports of toxicologic experiments with the solvents in laboratory animals; of epidemiologic studies of workers and communities exposed to TCE, PCE, and mixed solvents; and of studies of the Camp Lejeune population. Studies on how the chemicals are processed and distributed in the body of laboratory animals and humans were also reviewed and compared. Those lines of research were considered separately and then considered together to determine the health outcomes that were of greatest concern. The health effects on which there was convergent information from the toxicologic and epidemiologic literature, even if not perfectly concordant, were considered by the committee to be of most interest.

Epidemiologic Evidence

In evaluating the epidemiologic literature, the committee adopted a categorization scheme developed by the Institute of Medicine (IOM) for determining whether data indicate a statistical association between chemicals and various health outcomes. IOM’s approach was developed to evaluate exposure of veterans of the Vietnam War and the Gulf War and is used by the Department of Veterans Affairs to make decisions about compensation. The five categories in the scheme are limited/suggestive evidence of no association, inadequate/insufficient evidence to determine whether an association exists, limited/suggestive evidence of an association, sufficient evidence of an association, and sufficient evidence of a causal relationship. Among the five categories, only two were judged to be applicable to the literature on TCE and PCE: limited/suggestive evidence of an association and inadequate/insufficient evidence to determine whether an association exists. In the category of limited/suggestive, the evidence suggests an association between exposure to a chemical and a specific health outcome in human studies, but the body of evidence is limited by the inability to rule out chance and bias, so there is incomplete support of any association and insufficient basis for inferring a causal association. In the category of inadequate/insufficient, the available evidence is of insufficient quantity, quality, or consistency to support a conclusion about the existence of an association.

Overall, the committee did not find sufficient evidence to justify causal inference for any of the health effects it reviewed. The committee concluded that there was limited/suggestive evidence of an association between chronic exposure to TCE or PCE and cancers of the breast, bladder, kidneys, esophagus, and lungs. The epidemiologic literature was also judged to provide limited/suggestive evidence of an association between TCE or PCE and hepatic steatosis and acute tubular necrosis related to chronic exposure at high concentrations but not to chronic exposure at low concentrations. Studies also showed some evidence of an association between solvent exposure and acute glomerulonephritis. Findings of human studies were not sufficiently consistent to draw any firm conclusions about reproductive outcomes, but a few studies showed a potential association with male infertility, and there was a suggestion of an association between solvents in general and reduced fermale fecundability (the ability to conceive). The epidemiologic evidence provides some indication that solvent exposure during but not before pregnancy is associated with increased risk of miscarriage but not with preterm birth or reduced birth weight, and there is no direct evidence on perinatal mortality. The epidemiologic evidence on paternal exposure to TCE and adverse pregnancy outcome was inadequate/insufficient to determine whether an association exists. Human evidence on chronic exposure to TCE or PCE and the risk of congenital malformations was also judged to be inadequate to support conclusions about associations. Overall, there was limited/suggestive evidence of an association between principally inhalation exposure to solvents and neurobehavioral outcomes, with the most support for effects on visuomotor and motor function, fatigue, headache, and deficits in concentration; most of these effects were reported concurrently with exposure, and there has been little study of whether effects persist after exposure ceases. Epidemiologic studies have provided some support of two immunologically mediated outcomes—chronic glomerulonephritis and scleroderma. In each case, there is limited/suggestive evidence of an association with mixed solvent exposure and, for scleroderma, some indication of an association specifically with TCE.

Toxicologic Evidence

Animal cancer studies of TCE at maximally tolerated doses revealed liver and lung cancers in mice and kidney and testicular cancers in male rats. Similar cancer studies of PCE exposure revealed liver cancers in mice and mononuclear-cell leukemia and kidney cancer in male rats. These tumors were in most instances species-, gender-, and strain-specific. Malignant liver tumors were seen in only one strain of one sensitive species, the B6C3F1 mouse. Studies revealed that metabolic and mechanistic similarities between rodents and humans are such that highly exposed workers might develop TCE- and PCE-induced kidney tumors but appear to be much less susceptible than rats.

Review of noncancerous health outcomes in studies of TCE and PCE exposure indicated increased lung toxicity in mice, and hepatic and renal toxicity was reported after high exposure in rodents. Metabolism of TCE and PCE in rodents is qualitatively similar to that in humans but is quantitatively different and results in greater susceptibility of rodents to these compounds. Other studies revealed that rodent liver, kidney, and lung cells are more sensitive than equivalent human cells. Toxicologic studies reported adverse effects on indicators of male fertility in rats and mice exposed to TCE and PCE, respectively, at high doses, but there was little evidence of female infertility even at high concentrations. The toxicologic data constitute strong evidence that neither solvent is associated with congenital malformations in rats. Adverse pregnancy outcomes were not seen in toxicologic studies of maternal exposure to TCE in rats. A reduction in number of litters and increased perinatal mortality were observed in studies of mating pairs of rats and mice. Pregnancy outcomes after maternal inhalation exposure of rats to PCE indicate a reduction in intrauterine growth. Auditory deficits, reduction in performance of tasks, and other neurologic effects were reported in rats exposed to high TCE concentrations. Changes in visual evoked potentials in rabbits and decreased wakefulness in rats were reported in response to inhalation exposure to TCE. A few studies have reported neurobehavioral changes and altered brain neurochemistry in rats in response to inhalation exposure to PCE. TCE caused allergic sensitization in animal studies, including contact dermatitis and exacerbation of asthma. Toxicologic studies have shown exacerbation of autoim mune diseases in a genetically modified mouse model and immunosuppression after TCE exposure. Inhalation of PCE reduced innate bactericidal activity in mice subjected to inhaled microorganisms, but little information was available on the potential of PCE to suppress the immune system or to induce autoimmune diseases.

Integrated Consideration of the Epidemiologic and Toxicologic Evidence

Convergence of the epidemiologic and toxicologic evidence was considered to identify health outcomes of greatest interest and plausibility as potential consequences of exposure to TCE and PCE in the water supply. This approach supplemented IOM’s categorization approach by explicitly considering how the toxicologic evidence adds to the weight of evidence in characterizing health risks posed by TCE and PCE. The complementary strengths and weaknesses of the two bodies of literature provide important information on outcomes that are most deserving of attention. Review of epidemiologic studies of cancer outcomes provides limited/suggestive evidence of an association between chronic exposure to TCE or PCE and cancers of the breast, bladder, kidneys, esophagus, and lungs. Among those outcomes, positive concordance with the toxicologic evidence was strongest for kidney cancer observed in workers exposed to TCE, sometimes at doses where acute neurotoxicity was observed.

For noncancer outcomes, some convergence was found for toxic effects on the liver and kidneys of rodents and humans. Rodents exposed to high concentrations of TCE and PCE exhibited hepatic damage and renal tubular-cell damage. Epidemiologic studies also found limited/suggestive evidence of an association with hepatic steatosis (fatty accumulation in the liver) and sensitive measures of acute renal tubular necrosis. Such damage was associated with chronic high-level exposure to solvents but not with chronic low-level exposure.

Separate toxicologic evidence and epidemiologic evidence of associations between exposure to solvents and reproductive outcomes were found, but there was little convergence for specific reproductive outcomes. For example, toxicologic studies of high doses have reported adverse effects on indicators of male fertility in rats exposed to TCE and mice exposed to PCE; human studies were not consistent enough to support any firm conclusions, but a few studies showed a potential association with male infertility. The human data on female fertility were suggestive of an association between solvents and the ability to conceive, but there was little evidence of an association in the toxicologic literature to support female infertility even at high doses. Although the epidemiologic evidence of an association between chronic exposure to TCE or PCE and congenital malformations was judged to be inadequate to support conclusions, the toxicologic data provide strong evidence that neither solvent is associated with congenital malformations in rats. Reduction in fetal weight after maternal exposure of rats to PCE was observed in one toxicologic study; this outcome is considered somewhat analogous to the human outcome of “small for gestational age” (SGA), for which the epidemiologic data are inadequate/insufficient for determining whether an association exists.

Toxicologic studies report effects of exposure to high doses of TCE on the nervous system, such as central nervous system depression, attention deficits, alterations in visual evoked potentials, and other neurologic outcomes. Neurologic effects in toxicologic studies of PCE include anesthetic effects at high doses and changes in behavior and neurochemical markers at lower doses. Epidemiologic studies provide limited/suggestive evidence of an association between inhalation exposure to solvents and neurobehavioral effects; most of the reported effects were concurrent with exposure, and there has been little study of whether neurobehavioral effects persist after exposure ends.

Regarding effects on the immune system, toxicologic studies in sensitive strains of mice indicate that TCE can act as a skin sensitizer, modulate existing asthma, produce immunosuppression, and influence autoimmune diseases. Immunotoxic data on PCE are less abundant, with only a suggestion of effects on allergic sensitization and immunosuppression. Epidemiologic studies show limited/suggestive evidence of an association between mixed solvent exposure and two immunologically mediated outcomes, chronic glomerulonephritis and scleroderma. There is some indication of a specific association between TCE and scleroderma.

The committee is aware that some other health outcomes reported by former residents of the base (for example, male breast cancer and second-generation effects) are not cited above. The absence of inclusion of specific health outcomes does not mean that such effects should be excluded from further consideration of the Camp Lejeune population. Rather, it indicates that those outcomes have not been specifically investigated, or if they were considered, the studies were too small or of insufficient quality to support inferences.

Exposure Estimates in the Context of the Toxicologic and Epidemiologic Evidence

Perspective is needed in evaluating the exposures that occurred at Camp Lejeune. For example, some exposures are described as being “high” and others as being “low.” To understand the meaning of those descriptors, it is important to understand what is being compared. For example, ATSDR compared exposures with EPA’s MCL of 5 μg/L for PCE. In 1985, EPA classified PCE as a probable human carcinogen, and its policy is to assign a public health goal of zero exposure for such chemicals. The analytic feasibility of measuring PCE was considered in the setting of the MCL, and 5 μg/L was selected because it was judged to be the lowest concentration that could be reliably detected. Thus, the MCL is not based on toxicologic or epidemiologic data.

In epidemiologic studies, “high” exposures tend to occur in occupational situations where TCE and PCE are used routinely. Inhalation is usually the primary route of exposure in occupational scenarios, with skin exposure a less important route. Exposure tends to be much lower in community studies than in occupational studies and to involve exposure by the oral, dermal, and inhalation routes.

In toxicologic studies, exposure is usually expressed in terms of vapor concentration for inhalation exposure (parts per million) and dose for oral exposure (milligrams per kilogram of body weight per day). Lowest-observed-adverse-effect levels (LOAELs) were identified from the animal toxicologic studies for different adverse health effects. In some cases, a no-observed-adverse-effect level was also identified. The committee compared LOAELs with a range of estimated daily intakes that may have occurred at Camp Lejeune. Adverse health outcomes used in the evaluation were renal toxicity, renal cancer, neurotoxicity, and immune-related health effects—adverse outcomes in animals judged to be most relevant to humans on the basis of metabolic, mechanistic, and epidemiologic studies. Because of known variation in contaminant concentrations at Camp Lejeune, the range of exposures considered included the highest measured concentrations of TCE and PCE in finished water, half those concentrations, and twice those concentrations. Results of a toxicologic hazard evaluation1 indicate that the lowest doses that elicited adverse health effects in animals are much greater than the doses to children and adults that may have occurred, as estimated from the highest measurements taken of the Camp Lejeune water supplies. Thus, in the context of human occupational and animal studies, potential exposure of human populations at Camp Lejeune is described as being “low.” Although such comparisons afford a general frame of reference, they should be considered as just one facet of the health-effects evaluation. There are limitations in extrapolating the results of toxicologic studies, in which laboratory animals are exposed to high concentrations under controlled conditions, to human exposure scenarios where exposure varies in concentration and duration. Even community studies cannot be directly extrapolated to the Camp Lejeune population, because the Camp Lejeune population was much more transient than the nonmilitary populations studied in the other scenarios; moreover, other contaminants or other risk factors were probably present in both cases.

Past and Current Studies of the Camp Lejeune Population

Two analyses of the Camp Lejeune population have been completed by ATSDR, both of which focused specifically on health risks to children who were exposed in utero and considered measures of fetal growth and duration of gestation. No clear associations were found between exposure and mean birth weight, preterm birth, and SGA, although one study conducted a subgroup analysis and reported an increased risk of SGA in infants born to older mothers or mothers who had prior fetal losses. Weaknesses in both studies limit the ability to draw definitive conclusions—most important, weaknesses in exposure assessment. Place of residence at the time of birth was used to categorize people as exposed or unexposed despite the potential for migration in or out over the course of pregnancy. It was discovered after the study was completed that an area that was considered unexposed (Holcomb Boulevard) had received water from a contaminated system (Hadnot Point) for the first 4 years of the study period, so the study results became invalid. ATSDR plans to reanalyze its study with corrected exposure information; the committee views this as a useful effort that can be completed rapidly without awaiting water-modeling results.

An ATSDR study of the effect of prenatal exposure on birth defects and childhood cancers is under way. In addition to many of the same methodologic concerns as in the studies of fetal growth and pre-term birth, the current study has limited statistical power to detect associations with congenital defects or childhood cancer, and it does not consider exposures in infancy or early childhood. The results of that study await completion of ATSDR’s water modeling at Hadnot Point. As noted above, the committee recommends that simpler or conceptual groundwater modeling be performed for the analysis of Hadnot Point and that the results of that effort be applied to the completion of the case-control study of congenital defects and childhood cancer.

Future Studies of the Camp Lejeune Population

ATSDR has evaluated the feasibility of conducting three additional studies of the Camp Lejeune population, including a health survey and studies that would evaluate deaths from all causes and cancer incidence among former residents and workers. ATSDR identified some of the same diseases and disorders identified in the committee’s review as being of interest. These included kidney cancer, lung cancer, breast cancer, scleroderma, liver disease, kidney disease, and spontaneous abortion. ATSDR also identified additional outcomes of possible interest for its study.

The proposed health survey was generated in response to a congressional directive. The survey would seek information on residential history and various health outcomes, and could be used to support the other two studies. The survey’s success depends on getting adequate participation (at least 60%). Even if satisfactory participation is achieved, there are concerns that there could be bias in the reported data, because people who have experienced disease or illness are more likely to participate in the survey.

There are a number of difficulties with performing the mortality and cancer incidence studies, including identifying, locating, and recruiting the study participants and obtaining reliable health information on them in an efficient manner. The committee found that although ATSDR considered the major issues bearing on the feasibility of the studies and proposed reasonable approaches to address them, there remain serious, unresolved questions about the feasibility and ultimate value of the studies. For example, it is not clear that the cancer incidence study could be performed successfully, because it is contingent on the cooperation of many state cancer registries. Even with cooperation, the statistical power to compare groups of interest across the range of outcomes has yet to be assessed. Statistical power is also an issue with the mortality study. The quality of exposure assessment remains problematic as well. On the basis of information reviewed, the committee considers it unlikely that the proposed studies, even if the notable uncertainties about feasibility are all resolved favorably, will produce results of sufficient certainty to resolve the question of whether Camp Lejeune residents suffered adverse health effects from contaminated water.



  • The available scientific information does not provide a sufficient basis for determining whether the population at Camp Lejeune has, in fact, suffered adverse health effects as a result of exposure to contaminants in the water supplies. On the one hand, several lines of scientific reasoning suggest such effects are unlikely to have occurred. The evidence includes a substantial body of research on the toxicology of TCE and PCE that indicates that the exposures required to cause adverse effects in laboratory animals were much larger than the highest measurements available on the Camp Lejeune water supplies; evidence that humans have lower sensitivity to TCE and PCE than rodents; epidemiologic data largely from occupational settings with higher, longer-term exposures to TCE and PCE that has not generated compelling evidence of adverse health effects; and the relatively short-term, intermittent nature of the exposures incurred at Camp Lejeune. On the other hand, the possibility that health effects may have been produced by the contaminant exposures at Camp Lejeune cannot be ruled out. Some effects of TCE or PCE exposure might have occurred below the level of detection in toxicologic studies, which focused on single contaminant exposures at high doses, used genetically homogeneous animal strains, and necessarily involved extrapolation across species. In addition, the population exposed at Camp Lejeune is more diverse and possibly more susceptible than those that have been exposed to TCE and PCE in occupational settings, and the actual concentrations of PCE and TCE and the presence of additional water contaminants are poorly documented and could thus be higher or more complex than the limited historical measurements suggest. There were divergent views among the committee members about the probability that each would assign to whether adverse health effects have in fact occurred, but there was consensus among them that scientific research is unable to provide more definitive answers to that question.
  • Additional research on potential health effects of water contamination at Camp Lejeune are unlikely to provide definitive information on whether exposure to it resulted in adverse health effects. Limitations in population size, data availability, and data quality cannot be overcome. Those limitations are due in part to the lack of documentation of exposure and the difficulty in assessing the health events that residents experienced after they were exposed. Even if ATSDR’s planned work goes forward successfully, the outcome of the efforts is unlikely to determine conclusively whether Camp Lejeune residents were adversely affected by exposure to water contaminants.
  • Because of the historical and complex nature of the contamination that occurred at Camp Lejeune and the availability of few empirical data on concentrations in water supplies, only crude estimates of exposure can be obtained. Even with the use of reasonable and, in some cases, advanced approaches, limitations in data availability and quality cannot be overcome. Thus, only a general conclusion can be drawn that the Tarawa Terrace and Hadnot Point water-supply systems were contaminated and that residents and workers were exposed to the contaminants in a highly variable manner. Additional work should make it possible to assign exposure categories of exposed and unexposed based on time and residence with reasonable certainty.


Additional research on the affected population should be only one of several potential responses by the Marine Corps to the water-contamination at Camp Lejeune. Given the likelihood that such studies would extend for many years and their expected inability to deliver definitive information on whether the water-supply contamination at Camp Lejeune caused adverse health effects, efforts to address and resolve the concerns associated with the documented contamination should not be deferred until such research is completed. Policy changes or administrative actions that would help to resolve the controversy should proceed in parallel with the studies (if they are continued) rather than in sequence.



A dissenting viewpoint on the conduct of this evaluation is provided in Chapter 4.

Copyright 2009 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK215286


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (6.9M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...