The future of the Follow-up Agency remained unresolved. During a May 1953 meeting of the Executive Committee of the National Research Council's (NRC's) Division of Medical Sciences, Dr. Milton Winternitz reiterated that the follow-up program was under "critical review." 55 At the division's annual meeting, Dr. Wilbert Davison noted that a careful evaluation of the follow-up studies had been under way for a year and that the number of applications received by his committee had decreased. He added that the Veterans Administration (VA) continued to place greater emphasis on intramural research.56 In a private June meeting with Dr. R. Keith Cannan, soon to become the new chairman of the Division of Medical Sciences, Gilbert Beebe encountered a "neutral, show me" attitude. Cannan asked Beebe to make a case for the agency's continuation at the October meeting of the Executive Committee.57
At this meeting, members of the Executive Committee learned that the interest in follow-up studies was diminishing. Unless the VA showed greater interest or other funding agencies came forward, it would be necessary to consider "a progressive scaling-down and ultimate termination of the operation." Furthermore, the program had a "chequered history" because it attempted too many projects at once.58 Meanwhile, the Veterans Administration's 1954 research budget had been cut from a requested $6.5 million to $5.5 million and the number of the VA's own research laboratories had tripled in three years. In such a situation, any external research program that depended on VA funds would face difficulties. The follow-up program had received nearly $2.5 million from the VA over seven years, and had spent the bulk of it on clinical research done in 40 universities. The peripheral nerve injury study alone had absorbed about a quarter of the
funds. Listening to Beebe and Michael DeBakey defend the basic concept of follow-up studies, the Executive Committee agreed that "the medical experience of the Armed Forces and of the veteran population provide[d] a unique opportunity for ... studies of importance to clinical medicine and to the Armed Forces and the Veterans Administration." It decided that "steps should be taken to reestablish, as a broad inter-agency program, a significant program of medical follow-up studies."59
The Follow-up Agency had won a reprieve and a chance for survival if it could raise funds from sources other than the VA. If, as Keith Cannan and other NRC officials fully expected, a follow-up program of "carefully selected studies" could be justified, then the NRC would make an "intensive effort" to underwrite the program.60 Encouraged, the Committee on Veterans Medical Problems recommended that the VA continue its financial support of the Follow-up Agency ''to the fullest extent of its ability even though this mean[t] that other parts of the extramural contractual program [might] have to be curtailed.''
In the meantime, the committee urged the NRC to "continue efforts to acquire multiple agency support."61 With the Korean War ending and interest in civilian medical research booming, it now appeared logical to deemphasize studies of military concern and highlight "general medical interests in the natural history of disease."62
With this new mandate, the Follow-up Agency set out to win contracts from the many federal and private agencies that funded medical research. Although initial discussions about funding individual projects went well, money to maintain the central staff proved more difficult to obtain. The NRC estimated that a reserve of $50,000 to $70,000 in unrestricted funds would be necessary for this purpose.63 The NRC looked to the Department of Defense, the National Institutes of Health (NIH), and the American Cancer Society as possible benefactors. Cannan summarized the situation in March 1954: "The efforts of the Follow-up Agency to enlist the cooperation of a number of agencies in a program of cogent investigations are meeting with encouraging success." The agency expected to gather a nucleus of direct support from the Army for the study of epidemic hemorrhagic fever, and from the National Institutes of Health for the follow-up of approximately 1,000 patients with Buerger's disease and 100,000 naval recruits with tuberculosis (see Box 3). The possibility of contractual support from the VA existed as well.64
By November 1954, the Follow-up Agency had assisted in the publication of reports from its initial round of studies and had initiated three new projects. Papers on such subjects as psychoneurosis, schizophrenia, viral hepatitis, fractures of the carpal scaphoid, wounds of the hand, testicular tumors, rheumatic fever, sarcoidosis, and the aftereffects of million-volt x-ray all reflected research that had been done under the auspices of the Committee on Veterans Medical Problems. As for new projects, the National Heart Institute supported Michael DeBakey and Bernard Cohen's study of Buerger's disease, which consisted of a
BOX 3 Studies of Tuberculosis
In 1955, a study was published by the Medical Follow-up Agency (MFUA) and its collaborators comparing induction and separation x-ray films for roughly 3,000 veterans who were discharged from military service because of tuberculosis (TB) and 3,000 control veterans who were not discharged for this reason. The researchers found that about half of the men discharged from military service because of TB had had the disease in x-ray-detectable form at induction. The films themselves did not always provide an evident reason why lesions on the induction x-ray appear to have been overlooked, but it was noted that failures of detection were greater at stations known to have inferior medical service. The rate of apparently new cases of tuberculosis developed in service was found to be nearly twice as high among nonwhite as among white men, and tall, thin men developed tuberculosis more frequently. In addition, former prisoners of war had an excessively high rate of tuberculosis. In the fall of 1957, a study of tuberculosis morbidity in Navy recruits was published. The data came from a multiple skin-testing program carried out between 1949 and 1951 in San Diego, California. Data were available for nearly 70,000 young men. One of the two goals of the study was to provide nationwide data on the prevalence of sensitivity to tuberculin antigen; the other was to determine the number of now tuberculosis cases following testing. As part of the second goal, the incidence of TS among those who had a positive reaction to the skin test was to be compared to its incidence among nonreactors. Investigators found an average annual incidence rate of TB of 40 per 100,000, with a rate of 29 per 100,000 in the nonreactors and 157 per 100,000 in the reactors (a positive reaction was defined as one in which the induration measured 10 mm or more across). Height or weight did not appear to be associated with tuberculin reactivity, although lighter men did have a significantly higher rate of TB than heavier men. The study further determined that the value of x-ray screening for tuberculosis among those with an unequivocably negative tuberculin reactivity test was questionable.
Long, E.R., Jablon, S. Tuberculosis in the Army of the United States in World War II: An Epidemiological Study with an Evaluation of X-Ray Screening. VA Medical Monograph. Washington, D.C.: U.S. Government Printing Office, 1955.
Palmer, A.E., Jablon, S., Edwards, P.Q. Tuberculosis morbidity of young men in relation to tuberculin sensitivity and body build. American Review of Tuberculosis 76:517–539, 1957.
record follow-up of a large number of Army hospital admissions for this disease between 1942 and 1948. The U.S. Public Health Service funded Seymour Jablon's study of tuberculosis in naval recruits. The agency also participated in a clinical experiment, funded by the VA and conducted in ten VA hospitals, studying treatment of multiple sclerosis with the drug isoniazid. This was the first multi-
centered, randomized, double-blind treatment trial carried out in the study of multiple sclerosis.65 In addition to these studies, four others, including one by Cohen on the incidence of cancer in veterans, remained active possibilities.
Each of these new studies demonstrated that agency staff was taking a more active role in writing proposals and conducting research. No longer did the agency wait for proposals to reach the Committee on Veterans Medical Problems (CVMP). As a consequence, members of the agency staff began to develop expertise in particular areas and to publish more papers. In 1954, for example, a paper that appeared in the Journal of the American Medical Association on the sequelae of rheumatic fever carried only the name of Ephraim Engleman, the principal investigator who had received a grant from the committee, and his collaborators. In contrast, a monograph on the follow-up of World War II prisoners of war went to press under the names of agency staffer Bernard Cohen and an investigator from the VA.66
The new emphasis on entrepreneurial activity made Gilbert Beebe uncomfortable. He realized that the competence of the group he had assembled lay in methodology. The staff functioned best when someone brought them a problem to solve. In the changed circumstances of the mid-1950s, the members of the CVMP were no longer supplying the staff with projects. This required staff members to go out and sell their services, as Beebe put it, "to those who might have appropriate problems to solve."
The agency also faced an age-old dilemma encountered by nonprofit agencies: no project funds could be obtained until planning had been done, but no support was available for planning activities. As the NIH's James Shannon pointed out, the agency needed a specific program with specific objectives, but Beebe resisted this notion. His group, he believed, was not equipped to identify scientific problems or assess their importance. In the end, the agency functioned best when it facilitated the work of others or engaged in collaborative enterprises with clinical researchers.67
The decline in extramural research support from the VA affected not only the Follow-up Agency but its parent committee as well. Dr. Theodore S. Moise, who had been the NRC's chief staff officer to the Committee on Veterans Medical Problems and the only medical doctor who worked with the committee, resigned to take a position with the Veterans Administration. With his departure, the committee faced a period of reorganization, since its primary purpose, to award research contracts, no longer occupied much of its time.68 Since the Follow-up Agency received most of its funds from agencies other than the VA, it no longer made as much sense for the Committee on Veterans Medical Problems to supervise it. Beebe worked to assemble a more informal steering group, composed of representatives from the NRC, the VA, the National Cancer Institute, and the armed forces. This group, it was hoped, would provide a forum for problems faced by one or more of the agencies for which "the records resources of the cooperating agencies provide the best means of attack."69 Under this arrangement, the CVMP would continue to advise the VA on projects that required
access to VA records, but a prolonged slippage in its influence with the Veterans Administration began.70
By the end of 1955, the survival of the Follow-up Agency appeared to be ensured. The budget for fiscal year 1956 totaled $120,000 after overhead. Although the VA continued to be the largest single financial sponsor, the Public Health Service and the Army each contributed almost as much. The budget guaranteed the agency's continuation, but it barely covered the more than 21 individuals whom the agency hoped to retain, including 4 professional statisticians and 17 clerical and technical personnel.71
The core staff found itself pulled in many different directions. Staff members participated actively in designing current projects, including a study of the incidence of cancer in veterans, a follow-up on fractures of the femur that had been treated in the European and Mediterranean theaters of operation, and a study of the mortality rates of World War I veterans exposed to mustard gas or post-influenzal pneumonia. They also collaborated with doctors in universities and the government on final reports for completed projects. In this role, Bernard Cohen worked with Dr. H.F. Smetana of the Armed Forces Institute of Pathology on a report dealing with the course of Hodgkin's disease from onset to death, and Jablon applied the final statistical touches to a report concerning head injuries sustained in World War II (see Box 4). Gilbert Beebe completed a collaborative report on "variations in psychological tolerance to ground combat." Cohen, Jablon, and Beebe were also in the process of reviewing manuscripts on the subjects of prisoners of war and tuberculosis in the Army. Even as they undertook these activities, they also explored the possibility of starting new projects such as creating a roster of twins, which later emerged as a central activity for the agency, and studying the natural history of hypertension.72
As the core staff did more, the Committee on Veterans Medical Problems did less. It still met several times a year and discussed each of the major studies being considered by the Follow-up Agency. The CVMP officially recommended whether Veterans Administration records should be used for a particular study. With a study proposed by Bernard Cohen on coronary heart disease in young adult males, the committee faced a situation in which the project had already been reviewed favorably by the National Advisory Heart Council. This meant that the study would be funded by NIH. Still, the committee found it necessary to review the proposal with an eye toward "the propriety and desirability" of granting the investigators access to VA records.73 The committee also contemplated procedures that had proved troublesome from the very beginning of the program. In June 1956, for example, it responded to a complaint from Dr. Barnes Woodhall about the difficulty in getting patients to come to the hospital to participate in follow-up studies. He argued that the Veterans Administration should use its authority to compel people to appear, but how to do so was unclear. The resulting discussion ended as had many similar discussions. The committee reached no conclusion and took no action.
BOX 4 Studies of Head Injury
A paper published in the Journal of Neurosurgery (see Walker and Jablon, 1959) documented the work of the Medical Follow-up Agency and its collaborators on the study of head wounds in World War II; this publication was later followed (Walker and Jablon, 1961) by a book-length monograph. The material on which these reports were based came from examinations of 739 World War II veterans with head wounds who appeared at one of four clinical centers in Baltimore, New York, Boston, or Long Beach. The most common wound was a penetrating wound of the brain (i.e., a compound, comminuted fracture). Mental disturbances, such as impaired judgment and altered personality, were found to be related to the extent of injury, but not to its location, whereas aphasia (impaired production or comprehension of speech) was present when right-handed patients had been wounded in the left hemisphere of the brain. Headache, present in 82 percent of the men, was not correlated with the severity of injury. Whereas 28 percent of the men examined suffered from epilepsy, only 23 percent were found to have had multiple attacks. Wounds of increased diameter and depth were found to be associated with higher rates of epilepsy, particularly with multiple focal attacks. Similarly, a longer duration of unconsciousness after receiving a wound was related to increased risk of epilepsy. However, neurological symptoms at the time of wounding were not associated with later epilepsy. Finally, men who had their first epileptic attack sooner after receiving a wound had a more favorable prognosis than those whose first attack occurred later.
Walker, E.A., Jablon, S. A follow-up of head-injured men of World War II. Journal of Neurosurgery 16:600–610, 1959.
Walker, E.A., Jablon, S. A Follow-up Study of Head Wounds in World War II. VA Medical Monograph. Washington, D.C.: U.S. Government Printing Office, 1961.
New Projects in the Late 1950s
The June 1956 meeting featured the first extended discussion of the study of twins. The project came to the agency's attention in 1955 through Dr. William B. Wartman of Northwestern University. As early as 1953, Beebe had identified the veteran population as a valuable resource for twin studies. Together, Bernard Cohen and Seymour Jablon explored the feasibility of' creating a roster of twins. They discovered that a roster of as many as 10,000 pairs of twins could be developed from military records. It would, however, involve a great deal of effort and would not be cost-effective unless the roster could he used over a long period of time for many studies, Jablon estimated that each pair of veteran twins would cost between five and ten dollars to identify. The laborious process would involve checking with vital statistics authorities in each state to obtain a list of twins born during the period that made them likely to have served in the armed forces during World War II. The
list would then have to be checked against military records to see if a particular set of twins did, in fact, enlist. This method was cheaper than beginning with the military records and then checking to see if two servicemembers with the same last name were actually twins. To perform experiments that tried to discriminate between the effects of environment and heredity, it would also be necessary to separate the sample into monozygotic and dizygotic pairs (fraternal and identical twins). Jablon believed that, by using physical characteristics such as eye color, hair color, height, weight, and complexion, as well as fingerprints (obtained from the Federal Bureau of Investigation [FBI]), he could make accurate classifications in 80 percent of the cases.74 The committee expressed enthusiasm for the project, provided that a source of funds could be found.
The Follow-up Agency sought external advice from Dr. James Neel of the University of Michigan's Institute of Human Biology, who advised Bernard Cohen that, in an ideal situation, one would like to be able to observe the twins from birth to the grave. A sample of twins from Veterans Administration records would eliminate the first and much of the second decade of life and would not contain many people with congenital defects who, presumably, would have been screened out in the induction process. Hence, the sample would not be of much use in studying congenital malformations or the malignant diseases of childhood such as leukemia. Yet Neel believed the sample "would certainly be far better than anything that the world has ever had before."75
Beginning in 1957, Neel chaired an ad hoc committee on studies of veteran twins that gave the NRC and the Follow-up Agency still more advice on the project. It became a permanent institution in the Follow-up Agency and continued the ongoing separation of the agency and the Committee on Veterans Medical Problems by removing much of the twin study oversight from the committee. At the first meeting, after a lengthy discussion of such technical matters as determining zygosity (i.e., whether the twins were identical or fraternal), the ad hoc committee reached the same positive conclusion as the CVMP. It recommended that "steps be taken to establish a roster of twin veterans as soon as possible." As soon as the roster was established, the committee suggested that the agency undertake a pilot study to compare the relative effectiveness of various diagnostic methods for establishing zygosity.76
Although the study of twins appeared to set the Follow-up Agency off in a new direction, Gilbert Beebe hesitated to loosen the ties with the Committee on Veterans Medical Problems. His staff, he realized, focused on methodology, not specific medical subjects. Stimulus for selecting specific subjects had to come from external investigators, and the agency needed the committee's help in making connections with such researchers. The agency, Beebe noted, had ready access to records, not to investigators.
In 1957, another important development occurred when Keith Cannan asked the agency to participate in the program of the Atomic Bomb Casualty Commission (ABCC; Box 5). Seymour Jablon became the first staff member to go to
BOX 5Studies of Atomic Bomb Survivors
The world was introduced to the atomic age in 1945 when bombs devastated Hiroshima and, a few days later, Nagasaki, Japan. Japanese scientists soon arrived and began to investigate and assess the physical damage, deaths, and injuries among inhabitants and the effects on survivors. The occupation forces dispatched their own teams, which combined with the Japanese on the scene to form a joint commission to investigate the effects of the bombings. Impressed with the obvious magnitude of this task and the time that would be required to complete it, President Truman, in a November 1946 executive order, asked the National Academy of Sciences to undertake a program, funded by the Atomic Energy Commission (now the Department of Energy), to study the long-range biological and medical effects of the atomic bomb on man. Thus was born the Atomic Bomb Casualty Commission (ABCC). Guided by existing information from studies of mice about the health and genetic effects of radiation, James V. Neel inaugurated large-scale genetic studies of the offspring of survivors. Other investigators sent by the Academy did not, however, recognize that a strong, long-term, epidemiologic effort would be needed; hence, no comprehensive plan was implemented. The major early emphasis was on pathology and clinical medicine. Problems soon became apparent, and in 1955, Dr. R. Keith Cannan, chairman of the NRC Division of Medical Sciences, formed a team led by Dr. Thomas Francis, professor of epidemiology at the University of Michigan, to produce an objective, scientific appraisal of the ABCC program. This appraisal resulted in a plan for a long-term epidemiologic study of the survivors. One of the needs identified by the Francis group was collaboration by strong U.S. departments that would assist in recruiting scientists, provide scientific direction, and lend prestige to the program. The Yale and University of California at Los Angeles medical schools agreed to sponsor medical and pathology efforts, respectively and the Medical Follow-up Agency (MFUA) assumed responsibility for statistics (there was not yet a department of epidemiology at ABCC). For many years, MFUA personnel rotated between Washington and Japan, and helped to recruit statistical staff. Later, an arrangement with the University of Washington greatly strengthened the program. In 1975, the ABCC was succeeded by the Radiation Effects Research Foundation (RERF). Under the auspices of this organization, epidemiologic research concerning the long-term health outcomes of survivors of the atomic bombings of Hiroshima and Nagasaki, and their descendants, continues today. The association of the MFUA with the studies of the ABCC and subsequently the RERF, although in some ways a distraction from the studies of veterans, was on balance a very positive development. The research in Japan gave MFUA staff a degree of expertise and competence in studying the effects of ionizing radiation. These skills found expression in many other projects, such as studies of veterans who had participated in tests of nuclear weapons.
Putnam, F. Hiroshima and Nagasaki revisited: The Atomic Bomb Casualty Commission and the Radiation Effects Research Foundation. Perspectives in Biology and Medicine 37(4):515–545, 1994.
Japan to assist in the commission's work.77 Cannan decided that it would be best for the statistical program of the commission if senior statisticians from the Follow-up Agency took successive tours of duty in Japan. He hoped that Beebe would make the first formal tour, beginning in July 1958. When Beebe went to Japan, data from the commission would be sent to Washington, enabling some of the commission's work to be done by the Follow-up Agency staff there. The assignment strained the agency's staff. Perhaps for this reason, the agency, for the very first time, had no formal proposal before the Committee on Veterans Medical Problems when it met in December 1957.78
Reflecting at the end of 1957, Beebe noted that study findings often defied expectations. He pointed to the study of psychoneurosis that had been done by Norman Brill, using a sample of 1,500 Army and Navy veterans. Only a few of the men who had broken down could have been identified on induction. Their breakdowns reflected the stresses of war and could not easily be foreseen. Many men with an apparent predisposition to psychoneurosis served with distinction during the war. Upon follow-up, only 8 percent seemed to be severely disabled, and suicide was comparatively rare. These findings could not have been predicted in advance: it took a follow-up study to reveal them. Similarly, the studies of hepatitis and of scrub typhus addressed the question of whether an acute infectious disease set the stage for a chronic, degenerative process. Here too, only empirical research of the sort the agency performed could uncover the answers. Beebe noted also the extent to which the agency relied on statistical epidemiology, rather than clinical follow-up, and the fact that some of the agency's work involved using records from World War I.79
As Beebe recognized, the agency's studies followed no orderly pattern. In fiscal year 1957, for example, the agency contemplated studies on rheumatoid arthritis, multiple sclerosis, lesions of the intervertebral disk, infectious hepatitis, and coronary artery disease. In addition, the agency received requests to collaborate with VA hospitals on two clinical trials, one of "chemical adjuvants to surgery in the treatment of cancer" and the other of "anticoagulants and diet in the treatment of cerebrovascular disease." The former series of clinical trials became an ongoing effort for more than two decades. In the same year, the staff assisted in producing an article on a treatment for multiple sclerosis and helped to complete a monograph for the VA on peripheral nerve regeneration.80
Transitions in Personnel
The late 1950s marked an important transition in the agency's history. For the first time, the title of the agency in some official reports became the Medical Follow-up Agency (MFUA), although references to the Follow-up Agency were also common. Whatever the agency's name, the staff, with the aid of its clinical investigators, managed to complete five studies, including the study of Buerger's disease, in fiscal year 1958.
Changes in personnel also occurred at this time. As expected, Gilbert Beebe took a leave of absence to serve as chief of biostatistics for the Atomic Bomb Casualty Commission. In February 1959, Bernard Cohen suffered a heart attack, and he did not return to work until September 1960. To compensate for this loss, the agency hired two new professional staff members, the most significant additions since Jablon and Cohen had joined the agency a decade before. Dean Nefzger, who had a Ph.D. in psychology and statistics, came to the agency from the University of Buffalo. Robert Keehn left his job as director of the Bureau of Vital Statistics for the State of Connecticut to work for the MFUA.81
This transition in personnel slowed the agency's progress. When Gilbert Beebe went to Japan, he had to put aside three manuscripts on subjects related to acute viral hepatitis, cold injuries, and arterial injuries. When Bernard Cohen got sick, it stopped his work on monographs related to Hodgkin's disease and Buerger's disease. Michael DeBakey, Cohen's collaborator on the second study, was overwhelmed with other work and unable to attend to the Medical Follow-up Agency's projects.82 Then, once Beebe returned to Washington, Jablon left for Japan, a pattern that would be repeated in the years ahead as staff members shuttled back and forth.
In Beebe's absence, the staff concentrated on five major projects during fiscal year 1959. The first involved a follow-up of about 2,200 cases of coronary heart disease that had been diagnosed in the Army during World War II. The second study, which had a similar design, centered on the natural progression of multiple sclerosis in 850 soldiers diagnosed during the war. The third study consisted of the clinical trial of chemotherapy as an adjuvant to surgery in the treatment of cancer, and the fourth was a study of men admitted to Army hospitals during World War II with ulcerative colitis or regional ileitis.83
The fifth study, which proved to be the most time-consuming, concerned twins. The agency began the the cross-country search of birth records to obtain the names of live, white, male twins born between 1917 and 1927. During the first year of the project, the agency enlisted the participation of 44 of the 48 (soon to be 50) states and obtained information from 29 states, yielding 45,000 pairs of twin births and 170 sets of all-male triplets. Using the VA Master Index, project staff identified 6,000 pairs of twins who had both joined the military. Simultaneously, work began at the FBI to find the induction fingerprints of eligible twins. Even at this preliminary stage the project went over budget, since there were more twins than investigators had anticipated and the state offices charged more for their services than the investigators had expected.84 The VA, which financed the project along with the National Institutes of Health, agreed to make up the difference. Additionally, the project began to encounter delays, particularly in the FBI phase of the work. This in turn slowed progress in the next phase of the project—testing the effectiveness of using fingerprints to predict zygosity by comparing fingerprints with clinical observations. 85
Even as work on the twins and other major projects continued, and despite
the absence of key personnel, the agency managed to launch four new studies in fiscal year 1960. One focused on the morbidity and VA compensation experiences of soldiers who developed Japanese encephalitis during service in the Far East. Another looked at the long-term effects of acute infectious hepatitis on soldiers who served in Korea in 1950 and 1951. During treatment of this disease, doctors enforced a vigorous rehabilitation regimen: bed rest was discouraged and rehabilitation measures were started early in the recovery. The study sought to discover whether this course of managing the disease increased long-term complications. A third study followed up on a clinical trial of an influenza vaccine that had been conducted by the Army between 1951 and 1953 to see whether an oil adjuvant vaccine produced excess morbidity or mortality. The fourth study emphasized the agency's statistical expertise and involved an examination of the effectiveness of direct ultraviolet irradiation during surgery in preventing postoperative wound infections. During the year, agency staff published seven articles or monographs in publications such as the Journal of Neurosurgery. 86
Decline of the Committee on Veterans Medical Problems
The agency undertook all of this work without much guidance from the Committee on Veterans Medical Problems. William S. Middleton, the VA's chief medical officer, contemplated altering the committee's structure to reflect the fact that the VA no longer had much contract research and sponsored only a few projects conducted by the Medical Follow-up Agency. Keith Cannan of the National Academy of Sciences, for his part, noted the need to support the MFUA with competent advice but questioned the need for a committee such as the CVMP to provide this service.87
Gilbert Beebe, who was back from Japan, realized that very little of the agency's work was being done in VA hospitals and that the agency had not influenced the VA's research program. He argued that the "yield of the program might be significantly increased if the committee were able and willing to function more actively in the area of program planning." Still, Beebe found much to applaud. He cited the great variety of projects completed, the warm response the MFUA had received from federal agencies, and the unexpected opportunities for epidemiological investigation. The twin studies, in particular, revealed "an unexpected depth in the veteran material."88 Above all else, the program had produced a considerable volume of new scientific information.
The Committee on Veterans Medical Problems, for its part, believed that Beebe needed to do more to publicize the agency's program. The Committee suggested that he write an article for a national journal. Keith Cannan said that other NRC committees might be able to help more if they were "reacquainted with the potentialities of the program." DeBakey thought that the agency might broaden its approach to work with nonveteran groups. The follow-up technique
was receiving more emphasis in clinical research, and the agency staff might offer its services more widely. Although the agency did not immediately act on DeBakey's suggestion, the question of the breadth of the study population would recur.89
Toward the end of 1961, it seemed increasingly likely that the Committee on Veterans Medical Problems would not be reappointed in 1962. Beebe worked to formulate a new committee to advise the follow-up program. He approached a representative of the American Psychiatric Association and told him of the need for a psychiatrist to serve as a member of the new committee. He also talked with his professional associates in the field of epidemiology.90
In January 1962, Keith Cannan held an important discussion with the NRC about the future of the Committee on Veterans Medical Problems and the Medical Follow-up Agency. He found the response of the Veterans Administration to the suggestions of the NRC unsatisfactory. The group agreed to let the committee remain on the roster until June 1962, after which time it would be abolished. At the same time, the group decided that the agency should remain with the NRC and that an effort would be made to publicize its value and purpose.91
Beebe, Jablon, and Cohen all felt more comfortable with the role of scholar than the role of advocate. During this period, Seymour Jablon published an essay on the characteristics of a clinical trial and was the second author of a piece on acute epidemic hemorrhagic fever. Gilbert Beebe's long report on the possible relation of lung cancer to mustard gas injury (see Box 6) and his study of the 1918 influenza epidemic among World War I veterans appeared in the Journal of the National Cancer Institute. While in Japan, Jablon and Dr. A. Earl Walker reviewed their draft monograph on head wounds. Bernard Cohen continued to work with Michael DeBakey on their monograph about Buerger's disease.92
Much as he would have preferred to do his own statistical research, Beebe spent a considerable amount of time in 1961 and 1962 trying to present the agency's mission to a larger audience. He wrote an informal history of the program in which he highlighted some of the agency's major accomplishments. Beebe noted that as the program developed, it produced findings of interest to a wide range of doctors. Each study, he suggested, told a scientifically important story.93
As the end of the Committee on Veterans Medical Problems drew near, Beebe intensified his search to find people to serve on a new advisory committee. With help from outsiders, Beebe drew up a tentative list of 18 people that included, among others, Michael DeBakey; psychiatrist Alexander Leighton; internist and geneticist (and twin) Victor McKusick; Brian MacMahon, a Harvard epidemiologist interested in chronic disease; and Abraham Lilienfeld, a Johns Hopkins epidemiologist interested in cancer studies.94
In July 1962, Keith Cannan let the members of the Committee on Veterans Medical Problems know that the committee would not be reappointed. His first letter was to Michael DeBakey. ''It is hard to put into words a measure of our
BOX 6 Studies of Lung Cancer Mortality and Exposure to Mustard Gas
In July 1917, in a field outside of Ypres, Belgium, mustard gas was used for the first time in World War I. In total, mustard gas would cause nearly 400,000 casualties during the war, more than any other chemical agent. One of the long-term effects of exposure to mustard gas is chronic bronchitis, but it was unknown whether the long-term risk of lung cancer would be elevated in men who had been exposed. Three samples of World War I veterans totaling 7,151 U.S. white males were followed up for mortality through 1956 and again through 1965 in order to learn whether a single exposure to mustard gas with respiratory injury was associated with increased risk of lung cancer in later life. Men born between 1889 and 1893 who were either (1) exposed to mustard gas as documented by skin burns, or (2) hospitalized with pneumonia in 1918, or (3) hospitalized with wounds of the extremities were traced via the Veterans Administration's death records. Results of both follow-up studies failed to establish a definite carcinogenic effect for mustard gas, although the studies did find a relative risk of 1.3 for lung cancer mortality in exposed as compared to unexposed men. Observed deaths from lung cancer represented 2.5 percent of the mustard gas group, 1.8 percent of the pneumonia group, and 1.9 percent of unexposed men for the period 1919–1965. Even though these results were equivocal, presumably because men in the study received a relatively light dose, the second follow-up did find evidence suggesting a latency period for a carcinogenic effect ranging from 22 to 37 years. Although mustard gas was not used in World War II, both sides produced and stockpiled such weapons and prepared for their possible use. As part of this preparation, roughly 4,000 servicemen participated as experimental subjects in tests of the short-term effects of mustard gas. However, there has been no follow-up assessment of the long-term health effects of mustard gas exposure among these subjects.
Beebe, G.W. Lung cancer in World War I veterans: Possible relation to mustard-gas injury and 1918 influenza epidemic. Journal of the National Cancer Institute 25:1231–1251, 1960.
Norman, J.E. Lung cancer mortality in World War I veterans with mustard-gas injury: 1919–1965. Journal of the National Cancer Institute 54:311–317, 1975.
indebtedness to you over so many years and in such a diversity of problems,'' he wrote. Writing to William Stone, then chair of the committee, Cannan noted, "We plan to continue the program of medical follow-up studies and will need the active guidance of a group of medical investigators." He added, however, that he had not decided on the composition or the exact mission of such a group. Esmond Long, a long-time committee member, said that the fact that the follow-up pro-
gram would continue was "reassuring." Don Mainland, the New York University medical statistician who had evaluated the committee, agreed that the follow-up program still had what he called "potentialities, but I would think that its future would not be easy to plan."95 Ultimately, Cannan delayed the appointment of the successor committee for several years, allowing the agency to operate on the advice of ad hoc groups as it worked out various problems related to its future.
There were two problems as the CVMP passed into memory. First, the Veterans Administration had funded the Committee on Veterans Medical Problems and no similar means of support existed for a committee that would advise the Medical Follow-up Agency. As Cannan put it, the agency had to be supported on a "fluid basis," and the NRC was actively pursuing this possibility with the National Institutes of Health. Second, the MFUA was a "research tool, not a research institution." It therefore needed to develop "a better relationship with clinical investigation.''96
The death of Bernard Cohen in 1963 exacerbated this second problem. For 15 years he had played a key role in the program's development. At the time of his death, his monograph with Michael DeBakey on Buerger's disease was in press, and he had completed major papers on coronary heart disease and Hodgkin's disease. He also played a key role in the twin project. Adding to the interruption caused by Cohen's death, the agency continued its relationship with the Atomic Bomb Casualty Commission. Dean Nefzger left in 1962 for a two-year tour of duty, and Robert Keehn planned to go to Japan in 1963.
Simply put, the agency faced a desperate need for new staff to deal with a growing workload. For the twin study, it hired Glenn Atkinson, a graduate student in statistics at Cornell with considerable background in biology. He spent his first few months with the agency at the University of Michigan, studying human genetics with Dr. Neel. After a brief tenure with the agency, however, he left in 1965. In December 1962, Zdenek Hrubec joined the agency after having spent three years working for the ABCC in Nagasaki. He held a doctorate in public health statistics from the School of Public Health at Pittsburgh. Unlike Atkinson, Hrubec remained with the agency for a lengthy period of time and ultimately became the lead staff officer for the twin study.
The new staff members went to work on projects with subjects as disparate as multiple sclerosis (see Box 7) and herniated lumbar disks. Only one of the studies, on the effects of light mineral oil used as an adjuvant in experimental trials of influenza vaccines, was completed by fiscal year 1964. Even so, the agency initiated two new studies: one a long-term follow-up of x-ray technicians trained by the Army in World War II and the other an examination of the association between late-appearing cataracts and exposure to military radar. As these studies proceeded, Beebe and his associates also tried to shore up the statistical resources available to the agency. They persuaded the VA to retain a representative sample (2 percent) of the World War II National Service Life Insurance record cards, an invaluable source for developing control groups.97
BOX 7 Studies of Multiple Sclerosis
One of the most prolific subjects of study for the Medical Follow-up Agency has been the study of multiple sclerosis (MS). Beginning in 1963, MFUA staff and collaborators began a series of publications on the natural history of MS based on a set of cases identified from World War II Army hospitalization records. Papers were published on the progression from optic neuropathy to multiple sclerosis, the onset bout and its clinical features, long-term survival, clinical and laboratory findings at first diagnosis, and correlates of clinical change. Partly as a result of these investigations, it was decided to attempt to assemble a much larger series of MS cases and controls from military sources so that topics such as geographic variation in MS risk could be studied. With the assistance of collaborators, MFUA assembled a set of 5,300 MS cases and matched controls who served in World War II and the Korean War, and embarked on a long series of reports based on this material. The report on race, sex, and geography, for example, showed that white females had the highest risk of multiple sclerosis, followed by white males and then black males. The risk of multiple sclerosis was shown to increase substantially by geographic location, with veterans who were born or entered military service at more northern locations having a higher risk of developing MS. This increase in MS risk with latitude was shown not to be due to climatic factors, and migrants who moved south (e.g., those who were born in the North and entered service in the South) showed a substantially reduced risk of multiple sclerosis, whereas those who moved north from the middle tier of the United States had a substantially increased risk of MS. Subsequent research showed that age at onset of MS is younger as one moves northward. Using aggregate data from the U.S. Census, it was shown that MS risk was significantly higher in states whose residents reported higher rates of Swedish or Scandinavian ancestry; English and Dutch ancestry were associated with lower risk of multiple sclerosis. However, when ethnicity was determined for each individual by using the surname, there was no association of ethnicity with MS risk, except possibly for southern European ancestry. An analysis of other risk factors
New Core Funds
In fiscal year 1964, the agency's efforts at outreach and modernization appeared to be producing results. The National Institutes of Health joined the VA and the Department of Defense in providing general support, as distinguished from specific project support. In addition to the "research resources" contract, the NIH also awarded the agency a contract for its twin study. The new core funds, beginning in fiscal year 1965, made it possible for the agency to engage in more planning, undertake more pilot studies, and maintain the data bases on which its work depended. At the same time, the agency received new computers for its work, a move that made its data processing easier and enabled it to work more effectively with the Veterans Administration.98
With more funds and better hardware in place, the agency and the NRC turned once again to the task of appointing an advisory committee. In construct-
in white males showed that in addition to northern latitude, higher levels of education, an urban versus rural address, poor uncorrected visual acuity at entry into service, and a high proportion of persons with Swedish ancestry in the state from which a subject entered the military all increased the risk of MS. As the MFUA begins its second half-century, research on the topic of multiple sclerosis continues. Mortality follow-up studies of the two case series described above are under way, as is the assembly of a post-Korean War series of matched MS cases and controls.
Beebe, G.W., Kurtzke, J.F., Nagler, B., Nefzger, M.D., Auth, T.L., Kurland, L.T. Studies on the natural history of multiple sclerosis. V. Long-term survival in young men. Archives of Neurology 22:215–225, 1970.
Kurtzke, J.F., Beebe, G.W., Nagler, B., Auth, T.L., Kurland, L.T., Nefzger, M.D. Studies on the natural history of multiple sclerosis. III. Epidemiologic analysis of the Army experience of World War II. Neurology 17:1–17, 1967.
Kurtzke, J.F., Beebe, G.W., Norman, J.E., Jr. Epidemiology of multiple sclerosis in U.S. veterans. I: Race, sex, and geographic distribution. Neurology 29(Sept.): 1228–1235, 1979.
Kurtzke, J.F., Beebe, G.W., Norman, J.E., Jr. Epidemiology of multiple sclerosis in U.S. veterans. III: Migration and the risk of MS. Neurology 35(5):672–678, 1985.
Kurtzke, J.F., Page, W.F. Epidemiology of multiple sclerosis in US veterans. VII. Risk factors for MS. Neurology 48(1):204–213, 1997.
Norman, J.E., Jr., Kurtzke, J.F., Beebe, G.W. Epidemiology of multiple sclerosis in U.S. veterans. II. Latitude, climate, and the risk of multiple sclerosis. Journal of Chronic Disease 36:551–559, 565–567, 1983.
Page, W.F., Mack, T.M., Kurtzke, J.F., Murphy, F.M., Norman, J.E., Jr. Epidemiology of multiple sclerosis in U.S. veterans. VI. Population ancestry and surname ethnicity as risk factors for multiple sclerosis. Neuroepidemiology 14:286–296, 1995.
ing this committee, Beebe thought it important to find people who had a demonstrated interest in the natural history of disease; a knowledge of the medical bureaucracies of the armed forces or the VA; broad medical interests; and specific competence in internal medicine, general surgery, epidemiology, psychiatry, pathology, medical statistics, or medical genetics. He also wanted to appoint people who had at least some knowledge of the follow-up program. He divided his list of possible nominees among epidemiologists, public health statisticians, and other medical specialists. He favored an epidemiologist, such as Thomas Francis or Abraham Lilienfeld, as chair. Francis had the added benefit of already serving on the Advisory Committee for the Atomic Bomb Casualty Commission. Beebe also hoped that the NRC would appoint a small subcommittee to supervise the twin study, perhaps chaired by James Neel.99 On December 16, 1964, Cannan wrote Thomas Francis at the University of Michigan and asked him to accept the
chairmanship of the advisory committee. He noted that the NRC had delayed appointing the group until it learned about the application to the NIH for a research facility support grant. With the award of the grant, "we now have much more opportunity to develop a balanced program in a more orderly way than in the days when support came, in the main, from project grants."100
After Francis declined the committee chairmanship, Cannan turned to Dr. Brian MacMahon, professor in Harvard's Department of Epidemiology in the School of Public Health. Cannan urged MacMahon to form "a group of active individuals, largely investigators, who would be interested in directing the program into the most promising areas of research in the natural history of disease."101 MacMahon, whose experience with the MFUA had left him with "a deep respect for its professional staff, and also with considerable enthusiasm about the potential of its research program," agreed to head the new oversight body. This began a relationship that lasted for more than three decades. 102
Committee on Epidemiology and Veterans Follow-up Studies
The very name of the new Committee on Epidemiology and Veterans Follow-up Studies (CEVFUS) reflected "the increased evidence of potential for epidemiologic as well as other studies on the natural history of disease."103 Charging the committee to "develop the scientific content of the program and to oversee the work of the Follow-up Agency and collaborating investigators," Cannan wrote to one committee member that he expected the committee "to be involved in a constant reassessment and reformulation of program in response to the progress of medical research generally.''104 The committee contained some carryovers from the former CVMP, including DeBakey, and included a rich mix of epidemiologists, public health professionals, and statisticians from academia and the public sector. Public health figures outnumbered clinicians. Other than DeBakey, the clinicians included Dr. William S. Jordan, professor of internal medicine at the University of Virginia, and Dr. Louis J. Zeldis, professor of pathology at the University of California School of Medicine. Among others, Dr. Frederick H. Epstein, professor of epidemiology in the University of Michigan's School of Public Health, brought expertise in epidemiology to the committee.105
Taking stock of its mission as the new committee got organized, the Medical Follow-up Agency could note that the "medical research environment" was "much more favorable for a program that is population-based and in which statistical methods play a primary role" than had been the case at the agency's founding in 1946. In the interim, the MFUA had "succeeded in forging a practical tool of wide applicability and demonstrated effectiveness," despite the limitations inherent in data sets specific to the military veteran population. With funding and organizational improvements in place, studies flourished and the MFUA entered a new era. 106
If statistics were to be the agency's primary focus, then attention needed to be paid to the tools of statistical analysis. In this regard, Seymour Jablon made efforts to integrate the National Academy of Science's (NAS's) IBM 1440 mainframe computer into the MFUA's procedures (see photo), work that necessitated hiring new personnel and training ongoing staff, transferring various card files to computer tape, and developing new agency procedures. Jablon emphasized the development of "utility" programs that could be applied easily to specific problems as they arose. This would reduce the need for time-consuming programming. In all, the new computer would allow the agency to bring to "more refined, more elaborate, and deeper forms of analysis" to bear on the data sets at the heart of the agency's procedures.107
The New Committee's Mission
The Committee on Epidemiology and Veterans Follow-up Studies met for the first time on November 26, 1965. Dr. MacMahon described the committee's
mission as a combination of advisory, stimulative, and protective functions, to "bridge the gap between the program and the investigator that derived from the composition of the Follow-up Agency and its location" in the NAS. Following a discussion of the methods and resources of the agency, the committee heard a summary, of the studies in progress. Research efforts were being directed at heart disease, multiple sclerosis, and the consequences of occupational exposure to ionizing radiation.
Ongoing efforts in fiscal year 1965. included the study of long-term survival in Hodgkin's disease, the long-term follow-up of testicular tumors, the natural history of multiple sclerosis, and the follow-up of more than a thousand World War II veterans diagnosed with herniated lumbar disks (the agency's only clinical project then in operation). Agency staff noted that "two or three projects usually terminate each year," leaving room for "two or three" new projects each fiscal year, figures that would remain fairly consistent for the next decade. In 1965, planning began on at least five new studies: the possible influence of blood type on the incidence of salivary cancers, the link between early incidence of heat stroke and heat exhaustion and later-in-life heart and circulatory disease, the causes of chronic kidney diseases and disorders, a pilot study of amyotrophic lateral sclerosis (also known as Lou Gehrig's disease), and the treatment of thyroid nodules.108
At its second meeting, in early 1966, the committee began the task of agency oversight. MFUA staff described an important trend in the program, the shift from clinical follow-up studies to epidemiologic studies. This trend accentuated the emphasis on record studies at the expense of clinical examinations. Similarly, studies of mortality (as opposed to morbidity) became more frequent as the World War II cohort continued to age. Even in studies of morbidity, the distance from military service made it more difficult to locate some of the men.109
The agency's program development had suffered during its financial uncertainty in the early 1960s. In the meantime, agency staff had focused on ongoing efforts, such as the study of the epidemiology of multiple sclerosis. This study proved especially fruitful in 1966 and produced a paper given at the American Academy of Neurology, as well as publications in both domestic and international neurological journals. MIFUA researchers found geographic variation in preservice environment, as well as more important differences associated with urbanization, socioeconomic status, and race, to be significant variables in identifying veterans with multiple sclerosis. Other ongoing projects included studies of the epidemiology of coronary heart disease, the Hodgkin's disease project, and the study of occupational exposure to x-rays.110
The Twin and Prisoner of War Studies Mature
One of the most important ongoing projects in 1966 was the study of prisoners of war (POWs) (see Box 8), an effort to follow the experience of World War
II and Korean War POWs with respect to both mortality and morbidity. A resurvey of the former POWs began in March 1965. This study provided investigators with an unwelcome reminder of the ephemerality of paper records when, in 1965, an entire carton of records was discarded as trash by the night janitorial staff of the MFUA's downtown Washington office. After more than two days of excavation efforts at the Fairfax Sanitary Land Fill in Virginia, Beebe's staff gave up the search. Soldiering on, the agency continued the study throughout the mid-1960s, sending a mail questionnaire to its cohort of 2,500 POWs. The questions in the mailing, targeted to determine mental adjustment as well as physical well-being, were compared with those of an equal number of combat veterans who had not been captured.111
In a 1967 letter from staff statistician M. Dean Nefzger to Gilbert Beebe (in Japan on his rotation for the Atomic Bomb Casualty Commission), Nefzger noted that work on the POW study was "proceeding apace." The data had been collected, the index work completed, the mortality information coded, and the questionnaires analyzed. However, response rates for the questionnaires remained relatively low, around 70 percent, in part because of questions of the confidentiality of key information that was required from the Internal Revenue Service and the Social Security Administration. Problems with these agencies were endemic for the rest of the decade, suggesting the limits of the low-profile agency's ability to cut through Great Society-era red tape. Nefzger and those who followed him on the project turned increasingly to private organizations for help with the crucial matter of locating veterans who had lost contact with the Veterans Administration or who refused to respond to questionnaires.112
By offering the use of their mailing lists to the agency for name and address comparison and by frequent mention of its activities in their regular newsletters, veterans organizations proved of great assistance in helping the MFUA locate people in the sample of whom the VA had lost track. However, these organizations also gave the POW study an emotionally charged atmosphere, perhaps not well suited to the scientific method. For example, the American Defenders of Bataan and Corregidor wrote angrily to CEVFUS Chairman MacMahon to inquire why so few members of its organization had been targeted by the MFUA's series of surveys. Seymour Jablon responded for Dr. MacMahon, explaining the concept of "targeted surveys," which would approach only a statistically significant percentage of the entire World War II POW cohort. Still, the MFUA's relationship with this and other veterans organizations remained problematic. In general, learning how to deal with the emotional and political responses to its studies became increasingly important for the agency. By 1969, Gilbert Beebe was noting in correspondence that the POW study was "controversial at the level of federal legislation and VA policy, and any data that we generate are very likely to be used, even abused." 113
The MFUA's study of 16,000 veteran twins, another long-term study, gained momentum in the mid-1960s. At first it suffered from the death of study director
BOX 8 Morbidity Studies of Former Prisoners of War
For nearly its entire history, the Medical Follow-up Agency has been conducting mortality (see Box 2) and morbidity studies of former prisoners of war (POWs). Although the original POW study dealt in part with morbidity, the third study in the series, published in 1975, was the first major study devoted solely to POW morbidity. It is still arguably the best single study of the health of American former prisoners of war. In this extensive article, it was reported that the most persistent sequelae of military captivity were psychiatric, manifested in higher hospitalization rates and VA disability awards. These patterns were seen in all groups of former POWs, although they were more severe among those from the Pacific theater in World War II and from the Korean War. Excess morbidity was also found to correlate well with retrospective accounts of weight loss, nutritional deficiencies, and symptoms during captivity. Data collection for the fifth study in the POW series took place during 1984–1986. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression scale (CES-D). The study found a three-to fivefold excess in depressive symptoms among former POWs nearly 40 years after repatriation and also found that greater depression could be linked to self-reported weight loss and harshness of treatment during captivity. A number of reports have been prepared that link current depressive symptomatology with various factors. For example, it was shown that being younger and having less education at the time of capture, experiencing more medical symptoms during captivity, and receiving less social support after release were generally predictive of increased long-term psychiatric maladjustment.
Dr. Bernard Cohen and the fact that Seymour Jablon found verifying the zygosity of the twins more difficult than expected. His rotation in Japan with the ABCC for three years at a key moment in the planning process made finding computer time to run the necessary trials difficult and further slowed the process. The FBI continued to obstruct release of the remainder of the requested fingerprints for the twins cohort to the MFUA. When Jablon returned to the states and finished his computer trials, his findings were not as conclusive as he had hoped. Nonetheless, he and Professor Neel, the principal investigator on the project from the University of Michigan's Department of Human Genetics, determined that a combination of fingerprint analysis, serological testing, and, most especially, questionnaire analysis was sufficient to make the necessary determinations for formation of the NAS-NRC Twins Registry. They co-authored a paper that announced the Twins Registry to the medical world and interested a large number of investigators.114
Having anticipated this interest, Jablon and Neel brought the matter to CEVFUS in its March 1966 meeting. The committee agreed that the twins had to
In 1996, a book-length report of the sixth study in the series was published. This study involved the first attempted clinical follow-up examination survey of former prisoners of war. For this study, former POWs and comparison individuals were invited to be examined at a nearby VA hospital by local clinicians. Unfortunately, relatively few POWs and even fewer comparisons chose to be examined, which limited the value of the data collected. Nevertheless, analysis of the examination data yielded a number of medical findings of potential interest and again confirmed the high rate of psychiatric illness among POWs. Posttraumatic stress disorder (PTSD) was studied for the first time (the diagnostic category did not exist officially at the time the studies began) and was found to have occurred at especially high rates in former POWs. Subsequent reports have also been prepared on PTSD and on stroke among former POWs.
Beebe, G.W. Follow-up studies of World War II and Korean War prisoners. II. Morbidity, disability, and maladjustments. American Journal of Epidemiology 101:400–422, 1975.
Page, W.F. The Health of Former Prisoners of War: Results for the Medical Examination Survey of Former POWs of World War II and the Korean Conflict. Washington, D.C.: National Academy Press, 1992.
Page, W.F., Ostfeld, A.M. Malnutrition and subsequent ischemic heart disease in former prisoners of war of World War II and the Korean conflict. Journal of Clinical Epidemiology 47:1437–1441, 1994.
Page, W.F., Engdahl, B.E., Eberly, R.B. Prevalence and correlates of depressive symptoms among former prisoners of war. Journal of Nervous and Mental Disease 179(11):670–677, 1991.
be guarded from overzealous investigators who might abuse their good will, that the MFUA should regularly update the mortality and morbidity of the panel, and that further efforts should be made to seek fingerprint information from the FBI in order to ascertain zygosity definitively. The latter effort, however, never reached fruition, even with new leadership at the FBI. In addition, efforts began in 1966 to make the previously ad hoc subcommittee on twins a permanent fixture of the CEVFUS-MFUA structure.115
From its earliest days, CEVFUS functioned as far more than a rubber stamp; instead it took a hard look at the methods and appropriateness of all studies submitted. The committee considered the first three proposals to use the NAS-NRC Twins Registry in 1967. One, which became an ongoing study with follow-up into the 1980s, compared smoking patterns and other potential risk factors with the incidence of lung cancer and aimed to demonstrate the disparate roles that heredity and environmental factors played in these diseases. The other two proposals, however, lacked sufficient justification for tying them into the Twins Registry per se, and so the committee rejected them. Dr. Neel estimated that
although ''12 to 15 really good studies" might come out of the registry, making "the whole effort of creating" it worthwhile, a much larger number of applications would have to be sifted through in order to find these good applications. Throughout the late 1960s, Dr. Neel continued to push for a "hard-nosed'' approach in order to weed out those studies that did not have a "real appreciation of the nuances" or "limitations" of the twins approach.116
Early studies of twins were evenly divided between those that required clinical examination of the veteran twins and epidemiological studies that were used to judge the heritability of a particular disease. Over time, the clinical follow-up studies became less common, in part because of the cost and management issues involved in getting men to submit to examination and in ensuring consistency in examination procedures among doctors. Typically, twin studies measured concordance for a disease between both members of a twin pair and then compared these results for monozygotic and dizygotic twins. Theoretically, this method allowed a researcher to determine the degree to which the disease in question was a product of genetic instead of environmental factors. Neel, Hrubec, and the other twin study managers recognized the limitations of such research, however. Twin brothers would likely have shared environmental exposures, for example, and the twin studies were inherently observational rather than experimental. Moreover, since the presence of a particular disease in individual members of the cohort would change over time, the file necessarily had to be followed up over time, incurring additional future costs. Despite these limitations, however, the MFUA twin cohort (see Box 9) became a key resource in epidemiologic and heritability studies for several decades.117
From the beginning, CEVFUS members, especially James Neel, expressed concern that the Twin Registry would attract researchers with poorly designed studies that ran the risk of violating the confidentiality of the twins or exhausting their patience (e.g., by following up too often or overusing a regional cohort that had been used recently for another study). The committee feared in particular that a single region-specific study might exhaust the region's twin population for the foreseeable future. A possible solution was for researchers to combine forces in a collaborative effort, but this would be impossible in the case of clinical follow-up where a number of the examinations had already been conducted.
Then there was the problem of keeping the twins interested. CEVFUS urged the agency to feed the data back to the twins "to help maintain their interest." Unfortunately, despite sporadic efforts, this feedback plan proved too unwieldy, particularly in the largest studies, leaving the twins unaware of the uses to which their responses had been put. Occasionally, they would write angry letters to the MFUA staff, demanding to know what the agency was doing with their information.118
BOX 9 The NAS-NRC Twin Registry
Since the time of Sir Francis Galton, studies of human twins have provided material with which to study the relative effects of genetics and environment. When identical twins are more similar with respect to some characteristic than fraternal twins, this is taken as evidence of a genetic influence on a particular characteristic. In 1958, the Medical Follow-up Agency began a project to identify white male twins who had jointly entered military service during World War II, an effort funded by the National Institutes of Health and the Veterans Administation. Beginning with birth records provided by 42 vital statistics offices, 108,000 searches were made by hand against VA files to determine veteran status. In the end, nearly 16,000 twin pairs were identified in which both members had served in the military. Certain baseline data were abstracted from VA and military records, an initial questionnaire was mailed to the twin pairs, and anthropometric and fingerprint data were used to determine zygosity (i.e., to differentiate identical from fraternal twins). Interestingly, the single best question for determining zygosity by self-report—As children, were you and your twin as alike as two peas in a pod?—is essentially a translation of a similar item used in the Danish and Swedish twin registries and accurately predicts zygosity roughly 95 percent of the time. Subsequent follow-up data have come primarily from computerized VA records and mail surveys. A subset of roughly 500 twin pairs, the National Heart, Lung, and Blood Institute twins, have been examined by investigators four different times at clinical centers around the country. More than 200 journal articles have made use of the Twin Registry. Published articles cover a variety of topics. Among these are articles on smoking and respiratory function, schizophrenia, heart disease, bone mass, eye disease, type A behavior, blood chemistry profiles, headache, blood pressure, personality traits, financial earnings, dietary intake, fingerprint patterns, weight change and body fat distribution, alcoholism, cancer, diabetes, antisocial behavior, Alzheimer's disease, alcohol and tobacco consumption, suicide, declines in cognitive functioning, and prostate disease.
Hrubec, Z., Neel, J.V. The National Academy of Sciences-National Research Council Twin Registry: Ten years of operation. in W.E. Nance, ed., Twin Research. Proceedings of the Second International Congress on Twin Studies. Part B: Biology and Epidemiology, Progress in Clinical and Biological Research, Volume 24B. New York: Alan R. Liss, Inc., 1978. Pp. 153–172
Jablon, S., Neel, J.V., Gershowitz, H., Atkinson, G.F. The NAS-NRC Twin Panel: Methods of construction of the panel, zygosity diagnosis, and proposed uses. American Journal of Human Genetics 19(March):133–161, 1967.
Page, W.F. Annotation: The National Academy of Sciences-National Research Council Twin Registry. American Journal of Public Health 85:617–618, 1995.
The Agency in the Age of the Great Society and its Aftermath
In the late 1960s, MFUA staff began to raise concerns about the longevity of the raw material for their studies, particularly the hospital admission punch cards from World War II. As the basis for most of the agency's studies, these punch cards had to be protected against deterioration or, worse, loss, as had been the case with similar Navy admission cards a decade earlier. MFUA estimated the cost of transferring the fragile cards to more permanent magnetic tape in the tens of thousands of dollars, much more than the agency had at its disposal. Forced to look for outside funding, the agency secured money from the National Cancer Institute in fiscal year 1969 and stored the entire series on tape as a basic resource for future studies. 119
Even as its materials aged, the Medical Follow-up Agency, now staffed by 31 persons (5 professionals, 26 support staff) and with an annual budget of more than a half million dollars, reached important milestones in a number of its ongoing studies in 1967. The VA-funded series of controlled therapeutic trials of chemotherapy in conjunction with cancer surgery reached its tenth year in 1967. In addition to providing authoritative evaluations of the various adjuvant cancer agents tested to improve cancer-free survival after surgery, these trials provided important follow-up data on cancer survival as related to patient characteristics and treatment. Also in the same year, the agency finished a mortality study of patients who had undergone "curative" lung resection for bronchial cancer. The clinical examinations of men with herniated lumbar disks were completed in this year as well, and the agency began an epidemiologic study on the same condition to determine the significance of such variables as tissue degeneration, metabolic disorders, chronic stress, and occupational factors in mortality outcomes. Other ongoing studies during 1967 included an examination of viral and other factors in the etiology of cancer, a cooperative study of lower-limb amputations that aimed to elucidate factors of importance in the rehabilitation process for amputees, and a study of the possible relation of cholesterol levels in the blood to cerebrovascular disease.120
Despite this evidence of scientific progress, funding again became an issue for the Medical Follow-up Agency in the late 1960s. Its core support came under attack from within the National Institutes of Health. In part, this attack reflected changing circumstances: Keith Cannan, who had been so instrumental in obtaining funding from the National Institutes of Health, no longer headed the agency's parent Division of Medical Sciences, and NIH staffers with short institutional memories no longer remembered just why they were funding the MFUA. Moreover, NIH faced its own internal budgetary problems. In this, NIH was not alone; the postwar prosperity that had funded the massive expansion of government at all levels was beginning to falter. First the Vietnam War, and later the economic
difficulties of the 1970s, meant shrinkage rather than growth for federal agencies concerned with domestic policy.
As a result, the MFUA's application for renewal of the core funding grant from NIH was initially denied after a site visit in October 1967. Following protests, including those of CEVFUS Chairman MacMahon that withdrawal of support would have "tragic" consequences for "the most important single epidemiologic research program in the United States at the present time," NIH agreed to re-fund the agency's programs (including the Twin Registry), but at a lower level than requested. Furthermore, NIH guaranteed funds for only one year. Therefore, although the NIH decision solved the immediate problem, it placed MFUA on shaky financial ground. In the future, NIH would consider requests from the agency only on a year-by-year basis and only on the basis of grants, not core support. Without the assurance of core support for its program, the MFUA faced periodic difficulties throughout the 1970s and especially the 1980s.121
In the context of lost core support, individual projects took on a greater importance for the survival of the MFUA. In fiscal year 1968, several projects reached maturity. One was a study of survival rates following "curative" surgery for carcinoma of the colon, as part of the ongoing adjuvant cancer trials. Although none of the drugs tested in these trials was proven to increase survival rates, the trials gave the MFUA significant data for specific studies, one of which found that almost half of the patients studied were still living five years after their colon surgery. Another study considered the possibility of a link between smoking and lower risk of Parkinson's disease, the results of which clearly required further study. Hearing of these results, Dr. Chester S. Keefer of the American College of Physicians called the MFUA's projects "among the most valuable and the most exciting in medicine, [and] among the most important studies that have been made by any group in the NRC ... very important to the doctor and very important to the family."122
In this period the MFUA engaged in an almost bewildering array of studies. Ongoing projects as of late fall 1969 included the study of air pollution and smoking in relation to cardiovascular and respiratory symptoms, as well as a study of schizophrenia that contained a substudy investigating drug-induced parkinsonism in twins with schizophrenia. Other ongoing projects featured an investigation of changes in ocular pressure following application of steroids and a region-based study involving cardiovascular examinations of twins in New England.
In 1969, the Subcommittee on Twin Studies was formally created as an official subunit of the larger committee, with subcommittee membership drawn from CEVFUS as well as external investigators in genetics. In its first meeting, the subcommittee discussed a small number of new proposals and referred the most promising ones to the parent committee, a method of operation that continues to the present day. Beebe privately commended the Twin Registry for offer-
ing the MFUA the opportunity to return to the psychiatric studies that had been an early emphasis but had "faded out in the 50's."123
In this period as well, the agency continued such projects as the long-term follow-up of men with lumbar disk lesions in World War II, the therapeutic trials of adjuvants to surgical resection in the treatment of cancer, and the study of the natural history of multiple sclerosis. Corroborating a detailed clinical study from 1947 to 1951, an analysis of the possible association between viral hepatitis and subsequent cirrhosis of the liver failed to find a correlation in a mortality follow-up of more than 5,000 Army veterans with viral hepatitis in World War II. Preliminary results from a broad examination of blood type in relation to occlusive and hemorrhagic disease showed that a sample of 816 World War II veterans with records of myocardial infarction had blood type O less frequently than was statistically predicted.124
Despite these varied activities, the agency's financial solvency was far from assured. The threat of impending financial hardship led some committee members to suggest that the Atomic Energy Commission (AEC) "should participate in the general institutional support of the MFUA as compensation for its continuing contribution to the ABCC statistical program." Although the AEC arrangement included salary for staff and expenses for their rotations in Japan, committee members argued that the staff exchange was placing a financial burden on the agency that required compensation. Ultimately, nothing came of this matter, in part because Beebe publicly denied that the arrangement in any way weakened the MFUA's program. Still, the very suggestion demonstrated the committee's understanding of both MFUA's financial stringency and the strains that the ABCC rotations placed on the continuity of personnel in the follow-up program. By 1970, the Medical Follow-up Agency staff had dedicated 15 "man-years" to the ABCC rotation. 125
Although few new studies were added in 1970, significant progress was made in ongoing studies. However, Beebe faced a difficult situation with the resignation of Dean Nefzger in February to take a position with the Johns Hopkins School of Hygiene and Public Health. Efforts to find a qualified replacement took a surprisingly long time, and some projects suffered from the delay of transition, although Nefzger offered to manage several of them temporarily from his new workplace.
In the meantime, work continued on the agency's disparate ongoing projects. Preliminary results for a mortality follow-up of 10,000 inductees discharged from the Army in 1944 for psychoneurosis showed that mortality risk was greater for those who had been diagnosed with psychoneurosis than for the control group. Agency staff presented a paper on regional differences in mortality from cerebrovascular diseases at a conference on stroke epidemiology and began planning for a third test of the completeness of VA information on the mortality of the general population of war veterans. Planning efforts started on a study of behavior patterns in coronary heart disease; an examination of bleeding as a potential
side effect of anticoagulant therapy; and an investigation of parental radiation exposure and Down's syndrome, with particular attention to ionizing radiation and radar.126
Vietnam and the Veterans Administration
Although work had settled into a routine, the MFUA faced a changing environment. As the decade reached a close, agency staff found themselves increasingly concerned with how to utilize the Vietnam War experience in future studies, at first purely to strengthen ties with the military establishment. Early response from the military was quite favorable: one Navy Department staffer noted that "a potential gold mine of unique and valuable long-term follow-up medical data" existed regarding various studies conducted in Vietnam, including head injury, shock, and tissue graft studies, along with a variety of neuropsychiatric possibilities. In May 1969, military liaison officers participated in a wide-ranging discussion of the impact of the Vietnam War on the MFUA's future agenda. These liaison officers expressed continued interest in the MFUA and its work.
These same officers noted the limited extent of current procedures for the central indexing of potentially valuable research materials, which inhibited the potential for following up the Vietnam cohort. 127 The MFUA offered to access and preserve specialized rosters that might otherwise become lost or destroyed and to inventory research ideas and investigative interests brought to it. Beebe described the agency's potential research uses of the Vietnam experience as "essentially parasitic," in the sense that the agency would attach itself to already existing data sets and rosters. Ultimately, however, little came of this offer. Nor did the MFUA express much interest in Dr. Lyndon Lee's suggestion that the agency begin to plan to expand the ongoing POW effort to the Vietnam cohort, in part because of the departure of Dr. Dean Nefzger. Ultimately, the military itself took over the effort, in conjunction with the VA and associated outside consultants.128
The new chairman of the division of medical sciences, Charles Dunham, who took over in the fall of 1967, wrote to the surgeons general of each service asking that "medical officers be alerted to the research potential of routine medical and allied observations during wartime" and that efforts be made to preserve access in later years to observations made by medical officers in Vietnam. Later, MFUA staff became convinced that these alerts were never distributed by the surgeons general to field officers. However, a series of meetings resulted between MFUA staff and armed forces and VA medical personnel to discuss follow-up studies coming from the war in Vietnam.129
The group agreed to consider the need for follow-up studies on battle casualties from the Vietnam War as a test case for a more organized effort to identify priority problems and potential investigators. By late 1970, however, Dr. Beebe was privately concerned that the MFUA was missing an opportunity. "These efforts," he wrote of the meeting with the VA and the military, "have produced
little thusfar and raise some basic questions about the feasibility of any real forward planning" of the MFUA's Vietnam War program.130
At the third meeting of the ad hoc group on Vietnam issues, the group agreed on a list of ten likely areas of study and picked one, liver injuries, as a "sure-fire subject" around which discussions of program and method would coalesce. Beebe's concerns continued; he wrote Michael DeBakey in the spring of 1971 that "The [Vietnam] effort thusfar has been unsuccessful, but [I] am uncertain why. The meetings have been lacking in inspiration, and participants have not set one another off. The military representatives want to be cooperative, but all are busy and few ... would themselves be prepared to devote time to such studies. Perhaps it is the enthusiasm of the interested investigator that has been lacking, the compelling zeal of the man who wants to find the answer."131
In other words, the military of the 1960s contained no Michael DeBakey, no figure who combined clinical practice and organizational vision in a way that rose above the immediate concerns of the military bureaucracy. Beebe was understandably puzzled; the Medical Follow-up Agency had been founded almost three decades earlier because of "the enthusiasm of the interested investigator" that overflowed in the military medical establishment after World War II. This was a very different war—one that, instead of unifying the nation, divided it, and from which many of the "best and brightest" fled in horror and disgust. Beebe could not mobilize the surgeons general or their top advisers as DeBakey and he had done in the late 1940s. Too many doctors remained outside, evading or even assailing the war effort. Beebe worried that "an aging organization" would begin to falter without the new ideas and fresh outlook of the generation of the 1960s, noting that committee members had ''not had much to say about problems growing out of Vietnam.''132
The liver study gradually came to an end, and, despite some continuing efforts to bring the Vietnam War cohort into the MFUA program of studies, the agency returned its focus to the generation that had served in the 1940s. In 1973, Beebe would write to one CEVFUS member to express his regret that "we have not been able to get any work going on surgical problems growing out of the battle casualty experience in Vietnam, but perhaps we did not go about it in the right way." Michael DeBakey tried to organize another effort in the late 1970s, but it too fell short.133 Not until the 1990s did the MFUA successfully integrate a new cohort of veterans into its studies.
Denied a role in the medical activities of the Vietnam War, the MFUA expanded its activities in other ways. It approved new uses of the Twin Registry in 1971, including studies of risk factors in coronary heart disease, hereditary and environmental factors in dermatological conditions, and familial factors in early mortality. In addition, CEVFUS approved studies of amyotrophic lateral sclerosis and Hodgkin's disease. The committee also discussed how best to take advantage of the forthcoming automation of the VA Master Index, used for identifying veterans, linking VA and military files, and ascertaining mortality. The commit-
tee hoped to add this Master Index to the major MFUA rosters such as the twin and body-build registries. The new computer file, called BIRLS (for Beneficiary Identification and Records Locator Subsystem), ultimately became one of MFUA's main resources for the establishment and maintenance of research cohorts. Agency staff also secured funds during the year for preserving the Army diagnostic index to the Korean War on magnetic tape in order to ensure its availability for future use. Finally, the committee began a lengthy review of the Army's proposed modernization of the routine physical examination (MORPE) given to all new recruits, with hopes that the new examination would provide standardized baseline information for research in the history of disease and would be more easily retrievable by researchers.134
The early 1970s brought strained relations with the Veterans Administration. Beebe and Dunham met in early 1971 with Dr. Lyndon Lee, assistant chief medical director for research and education in medicine at the VA, to discuss an old problem. Lee, who had been the agency's scientific liaison officer with the VA for many years before being promoted to his new position, had urged the MFUA to expand its program of studies to the Vietnam generation. Praising the overall MFUA program, he nonetheless lamented the agency's low profile, particularly among armed forces staff. As the VA's assistant chief medical director, Lee "expressed a desire to have some interlocking of the Committee with the main VA advisory structure," on the theory that this organizational change would give the recommendations of CEVFUS more weight with the VA. He also suggested recommendations "better grounded in the realities of the VA situation," a proposal that could be implemented if his suggestions to have "men presently close to the VA'' added to the committee were implemented.135
Reflecting on the meeting with Dr. Lee, Beebe wrote to his superior Dr. Charles Dunham, still new enough as Keith Cannan's replacement that he had not yet become familiar with the place of the Medical Follow-up Agency within the NRC structure. Beebe reminded Dunham of the history of the earlier Committee on Veterans Medical Problems, phased out by Dunham's predecessor when its "broader" mandate proved impossible for its role as advisory body for the MFUA. Lee, he clearly implied, wanted to return to the earlier VA-NRC relationship, but Beebe cautioned against changing "the mission of the present Committee without a pretty good discussion of the reasons therefor," and without verification that it would be in the best interests of the Medical Follow-up Agency's program as well as those of the VA. If nothing else, Lee's requirements required adding several more clinical investigators to the committee, either at the expense of the existing disciplines represented there or at the cost of having to expand the committee. 136
The CEVFUS meeting to discuss these matters lasted more than six hours, one of the longest on record. At this meeting, it appeared that Dunham's ideas meshed with Lee's plans for the committee. Dunham wanted a committee that approximated the active policy-setting oversight functions common to the NRC
organization. If Lee and Dunham expected the committee to remake itself so quickly, however, they had grossly miscalculated. CEVFUS members expressed their unwillingness to take an "aggressive" stand in the development of the program, preferring instead to "encourage the kinds of studies that seem best suited to the opportunity presented by the military-veteran experience." So far removed was the committee from day-to-day operational thinking that its members declined to consider funding prospects for future proposals, on the somewhat dubious grounds that "worthwhile proposals would always find support.'' The committee effectively ceded to MFUA staff the basic program-setting responsibility that Dunham clearly had hoped the CEVFUS would seize. For his part, Lee's plan to return to the old MFUA-VA relationship was also denied. The committee expressed a desire to "recommend policy" to the VA, but only as that policy affected Medical Follow-up Agency matters.137