Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Executive Summary Over the past several decades, public concern over exposure to ionizing ra- diation has increased. This concern has manifested itself in different ways de- pending on the perception of risk to different individuals and groups within the population of the United States and the circumstances of their exposure. One such group is made up of those servicemen who participated in the atmospheric testing of nuclear weapons at the Nevada Test Site or in the Pacific Proving Grounds, who were involved in cleanup activities in Hiroshima or Nagasaki in the fall of 1945 and the spring of 1946, or who were prisoners of war who may have been assigned duties in those cities at the times of the bombings or shortly thereafter. Initially, this concern focused on the veterans themselves and may have been stimulated by early reports of an excess incidence of leukemia among participants in the 1957 Nevada test shot known as SMOKY (Caldwell et al., 1980, 1983~. The Radiation-Exposed Veterans Compensation Act of 1988 (Public Law 100-321) recognized this concern and identified 13 cancers (specifically, leukemia, multiple myeloma, lymphoma except Hodgkin's disease, and cancers of the thyroid, breast, pharynx, esophagus, stomach, small intestine, pancreas, bile ducts, gall bladder, and liver) that were deemed to be presump- tively service connected, and thus compensable. In 1994, this act was amended to include two additional sites of cancer, namely, the salivary gland and the uri- nary tract (Public Law 102-5781. Now the concern of some people extends beyond the health risk to the vet- eran and involves health issues related to their children, grandchildren, and spouses. As a result of these concerns the U.S. Congress, in Public Law 103 1
2 ADVERSE REPRODUCTIVE OUTCOMES 446, Section 508, directed the Secretary of Veterans Affairs (MA) to enter into an agreement with the Medical Follow-up Agency (MFUA) of the Institute of Medicine (IOM) to convene a panel of appropriate individuals to carry out the following tasks: An evaluation of the feasibility of a study to determine the nature and extent, if any, of the relationship between the exposure of veterans to ionizing radiation and the occurrence of (1) genetic defects and illness in their children and grandchildren, (2) adverse reproductive outcomes experienced by their wives, and (3) periparturient diseases of the mother that are the direct result of such adverse reproductive outcomes. The task of the committee, as elaborated by the VA, was to address the fol- lowing three questions: I. Is it feasible to conduct an epidemiologic study to determine whether there is an increased risk of adverse reproductive outcomes in the spouses and of adverse health effects in the children and grandchildren of Atomic Veterans? 2. If such a study is feasible, approximately how much time and money would be required to organize and implement it? 3. Are there other sources of information that would yield similar results at a lower cost or in less time? CONCLUSIONS The committee has addressed each of these questions, with the conclusions summarized here. The background information and rationale that served as a basis for these findings are described in detail in the full report. 1. Is it feasible to conduct an epidemiologic study to determine whether there is an increased risk of adverse reproductive outcomes in the spouses and of adverse health effects in the children and grandchildren of Atomic Veterans? Conclusion: The committee's assessment is that there are insurmountable difficulties in finding and contacting a sufficiently large number of study subjects (offspring of Atomic Veterans), in establishing an accurate measure of dose for each veteran, in detecting the extremely small potential risk at low doses, in identifying and reliably documenting reproductive outcomes over a 50-year in- terval, and in the measuring of other factors that have been observed to cause reproductive problems, and therefore, might confound any observed relationship between radiation exposure and reproductive problems. These difficulties be- come even greater in the grandchildren of these veterans. The committee con- cluded, therefore, that, as a result of the difficulties enumerated above, the cohort
EXECUTI HE SUMMER Y of Atomic Veterans does not provide a practical opportunity for a scientifically adequate and epidemiologically valid study. 2. If such a study is feasible, approximately how much time and money would be required to organize and implement it? Conclusion: Since the committee does not believe that an epidemiologic study is feasible, it did not consider in detail the time and money that would be required. However, on the basis of past and current studies of radiation-exposed cohorts, the committee estimates that such a study would cost tens of millions of dollars and would last at least a decade. 3. Are there other sources of information that would yield similar re- sults at a lower cost or in less time? While experimental animal studies could address some of the scientific is- sues discussed in this report, the committee has interpreted this charge to pertain to alternative epidemiologic studies that could yield similar results at a lower cost or in less time. . Conclusion: The committee suggests some studies that might be informa- tive, but notes that these too will have limitations. Commonly, these limitations are related to sample size, population composition, uncertainty of dose, the pres- ence of concurrent disease, and other confounding factors. Although studies of -these groups may have their own merits, the committee concludes that they may not adequately address the immediate concerns of the Atomic Veterans. ORGANIZATION AND FRAMEWORK OF THE REPORT no evaluate the feasibility of conducting an epidemiologic study of the families of Atomic Veterans, the committee felt it necessary to begin with a re- view of the fundamental principles of epidemiology, radiation biology, and ge- netics. This review is then followed by discussions of current information on the risk of genetic mutations due to environmental exposure, definitions and possible causes of adverse reproductive outcomes, the factors to be considered when de- termining the feasibility of a study, and finally, a review of possible alternative approaches for evaluating the health effects of exposure to low levels of ionizing radiation. ADVERSE REPRODUCTIVE OUTCOMES Adverse reproductive outcomes include such endpoints as the inability to conceive (sterility or infertility), the premature spontaneous termination of a
4 AD VERSE REPRODUCTIVE OUTCOMES pregnancy (spontaneous abortion), the birth of an infant with a congenital mal- formation or with mental and physical retardation, and premature death (stillbirth, neonatal or infant death). These events are not rare in the general population. For example, the committee estimates that 15,000 children with major birth defects would be expected among the 500,000 or so offspring of the 210,000 Atomic Veterans in the absence of any radiation effects. Many host and environmental factors contribute to the origin of such outcomes. These out- comes may arise through maternally or paternally derived inherited defects, through exposure to noxious environmental agents, including ionizing radiation, smoking, or the consumption of alcohol, through preexisting maternal illness (such as diabetes) or illness during pregnancy, and through poor nutrition. The presence of so many causes of adverse reproductive outcomes makes it difficult to design an epidemiologically valid study and to know the cause of the health outcome in any particular person. The likelihood that genetic effects may be seen following exposure of the human male to ionizing radiation is an important factor in assessing the feasibil- ity of studying adverse reproductive outcomes among Atomic Veterans and their families. The human data on which an estimate of this likelihood can be based are limited, and much rests on animal studies. The role of paternal exposure to radiation or chemicals has not been investigated extensively in humans (Olshan and Faustman, 1993~. However, some effect of chemicals or radiation on the sperm, chromosome, and fertility has been demonstrated (Wyrobek et al., 1983; Martin et al., 1986; Geneseca et al., 1990~. Thus, male-mediated environmental exposures may affect pregnancy outcome. Nonetheless, it should be emphasized that most of the findings are based on a small number of studies, and the paucity of information and definitive mechanistic models make interpretations of the available data arguable. ESTIMATION OF GENETIC RISK Three observations need to be understood when estimating genetic risk: 1. The results of the mutagenic effects of ionizing radiation are indistin- guishable phenotypically from those caused by other agents or those that arise spontaneously. 2. If a study shows statistically that an increase in an adverse health out- come in a population may be due to additional radiation exposure, this fact can- not be taken as proof that the illness of any particular individual is a result of exposure to radiation. 3. Relatively high doses of radiation (greater than 2,000 mSv [200 rem]) would add only a small number of additional cases of genetic disorders to the large number that are expected to occur as a result of spontaneous mutations, most of which have existed in the population for many generations.
EXECUTI HE SUMMER Y 5 Although much of what is known about the mutagenic effects of exposure to ionizing radiation rests on animal studies, notably those involving the mouse, two sets of human data have played important roles in estimating the risks of ionizing radiation to human populations: (1) data pertaining to the estimation of the spontaneous burden of genetic disease and disability (the number of cases normally present in the population) and (2) data on the Hiroshima-Nagasaki atomic bomb survivors and their children conceived subsequent to parental ex- posure that have been used to estimate the doubling dose. Overall, the studies of health outcomes (listed in the report) in the Hi- roshima-Nagasaki population have revealed a small but statistically nonsignifi- cant increase in negative health outcomes, including congenital defects, still- births, non-cancer mortality and some chromosomal abnormalities, among the children of the atomic bomb survivors conceived after the bombing compared with the outcomes among the children of nonexposed parents. This small in- crease is consistent with what would be predicted from studies of ionizing radia- tion in experimental animals and provides the best currently available basis for estimating the doubling dose in human beings (Neel and Schull, 19911. FEASIBILITY OF AN EPIDEMIOLOGIC STUDY The feasibility of a study of adverse reproductive outcomes among the families of veterans exposed to ionizing radiation hinges largely on the answers to four questions: (1) How is a suitable sample or cohort of exposed persons affected among the total at risk (numerators and denominators) to be defined, and can this be done without inadvertently introducing selection biases that could obscure a true effect or produce a spurious one? (2) Will that sample or cohort be large enough to reveal effects of the magnitude anticipated on the basis of present knowledge? (3) What is the probable dose distribution among the members of that sample or cohort, and how reliable are the individual dose esti- mates? (4) What approaches are available for identifying adverse reproductive outcomes accurately and completely? Each of these questions is considered separately. Question 1. Definition of the Study Sample Anecdotal information can be valuable in establishing the need for an epi- demiologic study, but self-volunteered information is unlikely to provide a basis for reliable estimates of risk since experience shows that persons with a personal interest in an exposure to some hazard are more likely to respond Accordingly, a scientifically defensible and valid study of the effects of ionizing radiation on adverse reproductive outcomes depends on the availability of a representative sample of exposed veterans and their families, and the means to establish these outcomes without reference to whether they are normal or abnormal. The Nu
6 AD VERSE REPRODUCTIVE OUTCOMES clear Test Personnel (NTPR) program of the Defense Nuclear Agency (DNA) has identified some 210,000 veterans who participated in one or more atmos- pheric tests involving the detonation of a nuclear weapon. These individuals or a suitably large and representative sample might provide the basis for a study co- hort, and it seems probable that deaths among these veterans could be deter- mined through the records of the VA or other sources. However, it is far more difficult to trace an unbiased sample of living persons, given the lack of identify- ing information in the original records. Furthermore, the available records do not contain information on the reproductive histories of the veterans (that is, their children), estimated for the purposes of this report to be about 500,000 in number. For reasons described elsewhere in this report, difficulties in locating offspring and obtaining such information reliably and accurately at this late date appear to be insurmountable. Thus, the committee concludes that, whereas a study of the life status and health problems of the Atomic Veterans themselves is feasible (and is in fact being done), the means do not exist to obtain information on adverse reproductive outcomes among their children and grandchildren in a manner complete enough for an unbiased estimate of the risk, if any, stemming from exposure to ionizing radiation. Question 2. Sample Size The committee has approached the issue of sample size by posing two re- lated but separate questions. First, the committee asked, If current estimates of the probable risk of adverse health effects among the children of the exposed fathers are correct, how large a sample would be needed to demonstrate that risk? Second, the committee asked, If sample size is fixed, how large a risk would have to exist to be statistically demonstrable? In the first instance, to cal- culate the sample size required, the committee estimated what the maximum relative risk would be, given the data on the effects of ionizing radiation pro- vided by animal and human studies. The maximum relative risk, for the purposes of determining sample size, was estimated to be 1.002 (0.2 percent increase in adverse health effects in the exposed population compared with the adverse health effects in an unexposed one). By using the maximum relative risk of 1.002 (equivalent to an excess relative risk of 0.002) and given that the expected frequency of major congenital defects in the general population observable at birth is about 3 percent, a sample size of 212,000,000 children would be neces- sary to demonstrate a statistically significant increased risk among the children of exposed veterans compared with that among the children of nonexposed fa- thers. In the second instance, assuming a sample size of 500,000 children and a frequency of major congenital defects observable at birth of 3 percent, the mini- mum detectable relative risk among the children of the most heavily exposed veterans is estimated to be 1.3, which is equivalent to an excess relative risk of 0.300. This excess risk, it will be noted, is 150 times larger than the maximum estimated excess relative risk based on current evidence, that is, 0.002. The
EX.FCUTI HE SUMMAR Y 7 committee concludes, therefore, that the sample size requirements are too great for a valid epidemiological study of adverse health effects among the children of Atomic Veterans to be performed. Question 3. Dose A crucial issue in assessing the feasibility of a study of adverse reproductive outcomes among the families of Atomic Veterans is the matter of dose. The demonstration that a particular endpoint increases in frequency as dose increases is a compelling argument for causality. If the individual doses are, on average, low and, moreover, unreliable, the demonstration of such a trend is unlikely. Inspection of the estimated doses received by the Atomic Veterans indicates that the majority probably received less than 5 mSv (0.5 rem). Everyone is exposed to background radiation that can vary in intensity depending upon location. Roughly 90% of the Atomic Veterans have estimated doses below the extra background radiation received from living in Denver, CO, or in an area of simi- lar altitude, compared to a more typical background area in the United States. Thus, without detailed residential histories the movement of these veterans in their post-service years could be an important confounder in the estimation of their cumulative doses. Moreover, medical exposure to ionizing radiation in the United States is very common. While the typical effective dose (NCRP, 1987) from a chest x-ray is small approximately 0.06 mSv (0.006 rem) other diag- nostic procedures such as an upper gastrointestinal examination (2.45 mSv, 0.245 rem), barium enema (4.05 mSv, 0.405 rem), or computed tomography (CT scan) (1.1 mSv, 0.110 rem) result in significantly higher doses. Thus, the ma- jority of Atomic Veterans would have received exposures to some tissues very similar to those that occur in certain medical examinations, and in the absence of detailed information on diagnostic irradiation, this, too, poses a possible source of error in the estimation of their total doses. The Committee to Study the Mor- tality of Military Personnel Present at Atmospheric Tests of Nuclear Weapons (IOM, 1995) concluded that the existing dose information on Atomic Veterans is inadequate to estimate individual doses with the requisite consistency to support an epidemiologic study. Taken together, these concerns make it highly improb- able that a valid study is possible. Question 4. Identification of Adverse Reproductive Outcomes Study of reproductive outcomes among the families of Atomic Veterans re- quires being able to identify both normal and abnormal outcomes in an unbiased manner. A cohort study identifying groups of veterans who are similar with the exception of exposure and following them over time to determine if the rates
8 ADVERSE REPRODUCTIVE OUTCOMES of reproductive outcomes differ by exposure group would seem to be a logical approach. However, this is more difficult than it would first appear. These groups are likely to have completed their families at least 15 years ago, and ex- perience has shown that the hospital records necessary to identify adverse repro- ductive outcomes during a period of from 15 to 50 years ago are not likely to be available at a quantity and of a quality sufficient for the purposes of an epidemi- ologic study. In addition, when one considers health outcomes that are not read- ily diagnosed at birth, such as learning disabilities and mental retardation, the challenges of finding documentation become even greater. As a result, these latter health endpoints are difficult to study epidemiologically in defined con- temporary populations and would be extremely difficult, if not impossible, to study adequately in a historical cohort. The potential for recall bias is of particular concern when studying health problems for which individuals may not routinely receive medical attention, such as spontaneous abortions, infertility, and developmental delays Spontaneous abortions present a number of methodological problems even for studies in con- temporary populations (Sever, 1989~. Unbiased studies of spontaneous abortions in populations that were at the height of their reproductive lives more than 30 years ago would not be possible. In groups of women who have been questioned about their history of spontaneous abortion, recall seems to be relatively accurate for the period up to 20 years prior to interview; however, before that time, recall is poor on the basis of a comparison of contemporary reports with later recall (Wilcox and Homey, 19849. This is in the absence of any concern about a po- tential association with an exposure that might lead to reporting or recall bias (White et al., 19891. ALTERNATIVE APPROACHES Data on the occurrence of adverse reproductive outcomes following expo- sure to ionizing radiation could be derived from a variety of cohorts in addition to the atomic bomb survivors, such as the children of (1) people residing in areas where the background of naturally occurring radiation is substantially higher than usual, (2) individuals, other than the Atomic Veterans, exposed to fallout from atmospheric weapons testing, (3) people living near nuclear installations, (4) individuals exposed occupationally, (5) patients undergoing medical diag- nostic procedures, and (6) patients undergoing medical therapy for benign or malignant disease. Each of these cohorts has strengths and limitations in sample size, population composition, adequate dose information, the presence of con- current diseases, and the presence of confounding factors. The committee rec- ognizes the real concerns of the Atomic Veterans as expressed by their represen- tatives, but it must conclude that epidemiologic studies cannot adequately address these concerns.