in exposure in the study population while minimizing exposure misclassification (error). Misclassification of exposure is of particular concern in environmental-epidemiology studies because the health effects of environmental exposures tend to be small, and it is usually difficult to accurately estimate exposure to environmental contaminants, which can occur by multiple pathways and in multiple locations. Furthermore, environmental exposures are often at low concentrations, which make biases due to exposure misclassification more likely to affect epidemiologic results. If misclassification of exposure is not differential by health outcome, it commonly biases risk estimates toward the null (that is, toward finding no association) and can cause associations to be missed (Copeland et al. 1977; Flegal et al. 1986). To evaluate the degree of misclassification in an epidemiologic study, it is important to consider the ability of an exposure metric to correctly classify the magnitude of exposure in the study population and to differentiate between those who are exposed at magnitudes that could result in adverse health effects (sensitivity) and those who are exposed at lower magnitudes (specificity). It is important to maximize specificity when the prevalence of exposure in the study population is low and to maximize sensitivity when the prevalence of exposure is high (Nuckols et al. 2004).
Exposure assessment for epidemiologic studies of the effects of water-supply contamination includes two components. The first is estimation of the magnitude, duration, and variability of contaminant concentrations in water supplied to consumers. An important consideration is hydrogeologic plausibility: an association between a contaminant source and exposure of an individual or population cannot exist unless there is a plausible hydrogeologic route of transport for the contaminant between the source and the receptor (Nuckols et al. 2004). The second component is information on individual water-use patterns and other water-related behaviors that affect the degree to which exposures occur, including drinking-water consumption (ingestion) and dermal contact and inhalation related to the duration and frequency of showering, bathing, and other water-use activities. Water use is an important determinant of variability of exposure to water-supply contaminants, particularly if it varies widely in the study population. Ideally, exposure-assessment strategies include both components, but in practice it may be difficult to obtain either adequately.
A number of approaches have been used to assign exposures in studies of health effects of water-supply contamination. They have ranged from measures of exposure defined by geographic region or job classification (group-level or ecologic exposure) to more sophisticated measures that yield individual exposure estimates. Selecting an optimal approach for a given study is dictated in part by the epidemiologic-study design, the size and geographic extent of the affected population, and the quantity and quality of available exposure-related data. The approaches that have been used in epidemiologic studies of water-supply contamination are more fully described in Chapter 6. The following sections provide information on the water-supply contamination and exposure scenarios at Camp Lejeune.
In the early 1940s, the U.S. Marine Corps constructed a water-distribution piping system at Camp Lejeune. The source of water in the system was, and continues to be, groundwater wells. The water-treatment processes, distribution systems, and contributing wells have been modified to accommodate the additional demand due to population growth and to improve water quantity and quality. Four water systems—Hadnot Point, Tarawa Terrace, Marine Corp Air Station, and Holcomb Boulevard—have supplied water to most of the residences and workplaces (see Figure 2-1). Other water-distribution systems on the base are Onslow Beach, Courthouse Bay, Rifle Range, and Camp Johnson.
In late 1984 and early 1985, Marine Corps authorities removed a number of supply wells from service in the Tarawa Terrace and Hadnot Point systems after concluding that they were contaminated with solvents (GAO 2007). The sources of contamination of the two systems were different. Investigation into the source of perchloroethylene (PCE) contamination of the Tarawa Terrace water system concluded