Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 41
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues 4 Conclusions and Recommendations In Chapter 2, the committee detailed the critical elements necessary to conduct a well-designed epidemiologic study to identify health outcomes in military and veteran populations related to exposure to depleted uranium (DU). In Chapter 3, the committee reviewed available datasets on health outcomes in DU-exposed military personnel and veterans and datasets that do not specifically assess DU exposure but might be useful for future study. This chapter draws on the information presented in Chapters 2 and 3 to identify options for further study of potential health outcomes in DU-exposed military personnel and veterans. For several reasons, studies of military and veteran populations (the Depleted Uranium Medical Management Program, including the Depleted Uranium Follow-Up Program at the Baltimore Veterans Affairs Medical Center; the Millennium Cohort Study; and the Department of Defense Birth and Infant Health Registry) have not been able to determine comprehensively whether exposure to DU is associated with health outcomes (see Chapter 3). Although those studies are well designed, they lack either adequate sample size or accurate exposure information (for example, biomarkers of exposure and work-assignment locations) to assess fully whether DU exposure is associated with health outcomes. For example, the Depleted Uranium Follow-Up Program includes extensive assessment of DU dose and health outcomes, but it is limited by the small study population (fewer than 100 veterans) and therefore lacks statistical power to detect many long-term health outcomes of interest. The Millennium Cohort Study has a large study population (more than 100,000 military personnel and veterans will eventually be enrolled) but is limited because assessment of exposure to DU is based on self-reported data and there is no followup through more detailed questionnaires or biomonitoring.
OCR for page 42
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues The remainder of this chapter contains the committee’s conclusions and recommendations. The next section describes four approaches to assessment of health outcomes in DU-exposed military personnel and veterans and the committee’s conclusion regarding the approach most likely to obtain useful information about DU-related health outcomes in this population. The chapter concludes with recommendations aimed at improving future epidemiologic studies and identifying current active-duty military personnel and veterans with potential DU exposure. PROPOSED APPROACHES TO THE STUDY OF HEALTH OUTCOMES OF EXPOSURE TO DEPLETED URANIUM As summarized in Table 4-1, the committee examined four general approaches to study health outcomes related to DU exposure in military and veteran populations. Two of the study designs use existing data, and two require new data collection. The first approach is a case-control study that uses existing data. The study population would be drawn from Department of Defense (DOD) and Department of Veterans Affairs (VA) health records and registries. Cases would be selected on the basis of a health outcome identified as having high priority for further study by the committee in its report Gulf War and Health: Updated Literature Review of Depleted Uranium (IOM, 2008). That report contains an extensive review of the scientific literature on long-term health outcomes of exposure to natural uranium and DU in several populations, including uranium-processing workers, people whose drinking water contains high concentrations of uranium, and Gulf War veterans. Although the committee’s overall conclusion is that the data are inadequate and insufficient to determine whether an association between exposure to uranium and a number of long-term health outcomes exists, it judged that several health outcomes should be given high priority for further study: lung cancer, lymphoma, renal disease, respiratory disease, neurologic outcomes (including neurocognitive outcomes), and adverse reproductive and developmental outcomes. Prior DU-exposure history would be assessed through a variety of means, including urinalysis, existing questionnaires, and interviews. Several limitations are associated with this approach, including a potentially small number of cases because of the low rate of disease occurrence and the low prevalence of exposure to DU. In addition, the followup time may be too short to capture some health outcomes because of their latency. The second approach, similarly, uses a case-control study design but adds collection of new data by including future cases of disease and appropriate controls. Thus, the study population would be expanded by prospectively adding new cases with health outcomes of interest and controls from military personnel in war theaters where DU is or will be in use. Information about exposure to DU before the development of diseases and outcome data would be collected
OCR for page 43
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues TABLE 4-1 Proposed Epidemiologic Study Designs to Assess Health Outcomes of Exposure to DU Study Type Existing Data New Data Retrospective Analysis Case-Control Case-Control Study population Drawn from existing DOD and VA databases (such as military health records and registries) to identify people with relevant diseases (see below) and appropriate controls without those diseases who served in a war theater where there was opportunity for exposure to DU Same as under “Existing Data” except that study population would be expanded by prospectively adding people as they enter military service in war theaters where DU exposure could occur Cases: Representative set of all persons in study population identified as having one of the followinga Lung cancer Lymphoma Renal disease Respiratory disease Neurologic outcomes Adverse reproductive and developmental outcomes Controls: Representative set of persons in study population who are free of those conditions Exposure assessment Exposure history before becoming diseased (case) or before analogous time for controls, including information from existing questionnaires, interviews, work assignments and tasks, urinalysis Exposure history before becoming diseased (case) or before analogous timepoint for controls, including information from existing or prospectively administered questionnaires, interviews, urinalysis; prospective collection of exposure information could be tailored and intensified beyond what is possible from historical questionnaire, interview data Outcome assessment (used to determine disease [case], control status) Use existing databases, such as military health records and registries, to identify people with relevant diseases Existing health conditions would be ascertained from historical records, prospective onset of conditions ascertained from similar sources; prospective ascertainment of health conditions could be intensified beyond what is possible with existing health records
OCR for page 44
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues Study Type Existing Data New Data Limitations Small number of identifiable cases of disease because of low rate of disease occurrence Although small number of identifiable cases of disease, according to existing data, can be expanded, prospective collection of data defers timing of study Low prevalence of exposure to DU Followup time may not be adequate to capture some health outcomes because of latency Longitudinal Analysis Nonconcurrent or Historical Cohort Prospective Cohort Study population Veterans exposed or reportedly exposed to DU (such as those at Camp Doha at time of 1991 fire) Same as for existing data plus addition of military populations assembled prospectively Veterans with no history of exposure to DU Exposure assessment (used to determine exposure status) Exposure history, including information from questionnaires, interviews, urinalysis Better and more accurate measures (or indicators) of exposure information, including opportunities for corroboration by biologic data, such as individual urinalysis or more details on work assignment (location and length of time in proximity to theater) Outcome assessment Documentation of occurrence of relevant adverse health outcomes via linkages with death certificates or disease registries In addition to assessment of outcome with existing data, study population would be monitored for relevant adverse health outcomes via clinical surveillance or linkages with death certificates or disease registries Limitations Identifying study population likely to be challenging given time elapsed since 1991 Gulf War Small study population unless new persons (beyond current military rosters) are recruited Characterizing exposure may be problematic Few health outcomes, low disease rates Low disease rates Time and cost to conduct study may be considerable NOTE: DOD = US Department of Defense, DU = depleted uranium, VA = US Department of Veterans Affairs. a Health end points listed were identified as having high priority for further study by committee in its report Gulf War and Health: Updated Literature Review of Depleted Uranium (IOM, 2008).
OCR for page 45
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues prospectively. Because this approach requires prospective ascertainment of exposure and outcome data, the study would be deferred until troops are sent into relevant war theaters. The third approach is a cohort study that uses existing data. The study population would consist of veterans with exposure or reported exposure to DU (for example, soldiers at Camp Doha during the 1991 fire) and a comparable group of military personnel unexposed to DU. Exposure would be assessed through questionnaires, interviews, and urinalysis. Adverse health outcomes would be documented by linkage with registries or death certificates. This type of study has several limitations. First, given the time that has elapsed since the Gulf War, it would be problematic to identify the potentially exposed population and characterize their specific individual exposures. And selection of an appropriate unexposed comparison group that would be matched to the DU-exposed group on all known prognostic factors related to the health outcomes of interest would be challenging. The final approach is a cohort study that would expand the study population of the historical cohort study (as in the study description above) with military personnel either exposed or unexposed to DU added prospectively. This study design provides opportunities to gather extensive exposure information, including biomonitoring data and detailed work records for people enrolled prospectively. In addition to determination of health outcomes by using disease registries and death certificates, the study population would be monitored clinically. Limitations of such a study include low rates of health outcomes of interest and a small study population of DU-exposed military personnel. And the length of time and the cost to conduct a prospective cohort study could be considerable. The committee recognizes that information from the 1991 Gulf War and from Operation Iraqi Freedom concerning the extent of exposure and identification of exposed people is suboptimal and that this limits the ability to conduct epidemiologic studies of health outcomes of DU exposure. Given the limitations of the studies described above, particularly the low statistical power and the lack of adequate exposure data, the committee concludes that it would be difficult to design a study to assess health outcomes of DU exposure in military and veteran populations comprehensively. Detecting a small increased risk for a given health outcome of DU exposure in military and veteran populations is not feasible in an epidemiologic study. For example, as detailed in Chapter 2, a minimal sample of more than 1 million DU-exposed people (for a relative risk of 1.01) would be required to detect a statistically significant difference in risk of lung cancer, a common cancer, between DU-exposed and DU-unexposed populations. Fewer DU-exposed people would be needed to evaluate renal disease than lung cancer because renal disease is more prevalent than lung cancer in the population. A range of potential study sizes is provided in Chapter 2 of this report with the associated assumptions. As discussed there, the committee advises that if a study is undertaken, the sample-
OCR for page 46
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues size requirements be viewed as guidelines that are likely to be underestimates given the challenges of exposure classification, anticipated attrition of subjects from study dropout, other losses to followup, and death over the 20-year followup period. Before DOD considers undertaking an epidemiologic study, it should re-examine the sample-size estimates for lung cancer and renal disease and calculate sample-size requirements for lymphoma, respiratory disease, neurologic outcomes, and adverse reproductive and developmental outcomes to determine whether there is a reasonable possibility of detecting an effect. The committee believes that of the four approaches outlined above, the approach most likely to obtain useful information about DU-related health outcomes would be a prospective cohort study if future military operations involve exposure to DU. The committee’s power analysis (that is, of minimal sample-size estimates) demonstrates that the feasibility of an epidemiologic study will depend primarily on the ability to define sufficient numbers of people exposed to DU. Beyond the size of future DU-exposed military populations, the success of any cohort study would depend on DOD’s ability to collect accurate and complete individual-exposure information on military personnel who enter war theaters in which DU munitions and armor are used. DOD should design and implement a data-collection system to assess potential exposure, obtain biologic and environmental samples, and define the cohort for long-term followup of health outcomes. RECOMMENDATIONS FOR IMPROVING FUTURE EPIDEMIOLOGIC STUDIES DOD should use the most sensitive assays when collecting biomarker data and should investigate available in vivo assay techniques other than measuring urinary uranium to determine whether they offer advantages (for example, increased sensitivity) over urinalysis. DOD should consider assessing uranium concentrations in lung, kidney, and brain tissues from military personnel who were potentially exposed to DU and died while on active duty. It is the committee’s understanding that DOD maintains a repository of autopsy material from all military personnel who die while on active duty (M. Kilpatrick, personal communication, March 22, 2007, and February 11, 2008). Analysis of uranium in autopsy tissue might provide information on concordance between renal uranium concentrations and model-based estimates, such as those presented in a Royal Society report on the health hazards associated with DU munitions (Royal Society, 2001). It also would provide information on pulmonary retention of DU, which has implications for estimating lung cancer risk, and insight into the toxicokinetics of DU. DOD should continue to link and integrate available databases so that information can be assessed. With additional data-linkage capabilities, DOD and
OCR for page 47
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues VA will be better able to use available data to assess DU exposure and related health outcomes. ADDITIONAL RECOMMENDATIONS The committee commends DOD’s Depleted Uranium Medical Management Program and recommends that DOD expand its efforts to identify all active-duty military personnel and veterans with potential DU exposure to determine their DU body burdens. Specifically, the committee recommends that DOD determine the feasibility of collecting biomarker data through the Millennium Cohort Study from the 2,823 members of Panel 1 (or potentially from members of all three panels if such information is available) who reported being exposed to DU in the preceding 3 years. That information could be used to conduct an exposure study to assess the accuracy of surrogate exposure measurements (material on questionnaires). Any study participants who test positive for DU exposure should receive health monitoring through the Depleted Uranium Follow-Up Program throughout their lifetime. Comprehensive clinical assessments of exposed people provide extensive data on health outcomes potentially related to DU exposure and might have value in stimulating further study. DOD determine the feasibility of identifying military personnel who were at Camp Doha during the time of the fire in 1991 and of collecting biomarker data on them. This work could provide exposure information on level II exposure and level III exposure during the Gulf War. Any veteran who tests positive for DU exposure should receive health monitoring (see above). The committee’s final recommendation is related to studying adverse reproductive and developmental outcomes. The committee recommends further study of the potential reproductive and developmental toxicity of DU with animal models (a review of reproductive and developmental toxicity studies of DU is included in Gulf War and Health: Updated Literature Review of Depleted Uranium [IOM, 2008]). Specific studies may address reproductive measures (for example, the number of motile sperm, which is highly predictive of intrauterine-insemination success and, thus, correlated with fertilization rate). Important observations on the effects of DU on developing animals might be derived from in vivo animal studies. For example, they could provide information on growth, reproductive capacity, cancer, neurobehavioral function, and transplacental exposure. Animal studies focusing on the health effects of DU on the progeny’s health might be useful for corroborating findings reported in epidemiologic studies, and outcomes may be correlated with DU exposure (as measured by urinary DU excretion).
OCR for page 48
Epidemiologic Studies of Veterans Exposed to Depleted Uranium: Feasibility and Design Issues REFERENCES IOM (Institute of Medicine). 2008. Gulf War and health: Updated literature review of depleted ura-nium. Washington, DC: The National Academies Press. Royal Society. 2001. The health hazards of depleted uranium munitions: Part I. London, UK: The Royal Society Working Group on the Health Hazards of Depleted Uranium Munitions.