Sever, Lowell E.
Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
William Jackson Schull was born in Louisiana, Missouri, on 17 March 1922. After receiving his PhD in genetics in 1949, he began an influential partnership with James Neel on the study of effects of radiation exposure among atomic bomb survivors in Japan. Much of what is known today regarding the health effects of ionizing radiation is the result of the studies that were initiated by Schull and Neel and then guided by them for more than 40 years. In 1954, Dr. Schull coauthored one of the first textbooks of human genetics, laying the foundations and directions of the field. He was a founding member of the human genetics department at the University of Michigan, the first in the United States. In 1972, he founded the Center for Demographic and Population Genetics (now known as the Human Genetic Center) at The University of Texas Health Science Center in Houston, where population geneticists, epidemiologists, and demographers are brought together to understand the impact of genetic and epidemiologic variation on individuals, families, and populations. In the late 1970s, Dr. Schull recognized the need to investigate the health of Mexican-Americans in Texas. He set up a series of continuing studies that have contributed much to our understanding of the genetics and epidemiology of cancer, noninsulin-dependent diabetes, gallbladder disease, and cardiovascular risk among Mexican-Americans. He is a Fellow in the American Academy of Arts and Sciences and a member of the Order of the Sacred Treasure, Third Class, Emperor of Japan.
LS: Few people start out to be epidemiologists. How did your path lead you into this discipline?
JS: I was trained in genetics and in human population genetics, in particular, but those techniques are similar in many respects to those used in epidemiology. The focus on populations, and on the mechanisms that lead to change in those populations, provides common ground with epidemiology. When I began graduate studies, there were few places in the United States where an individual interested in human genetics could be trained. Perhaps the best recognized was Ohio State University (in Columbus, OH), so I went to graduate school there. At Columbus, I met people conducting research in cancer epidemiology. Working with and near them was a great experience, and led me to see the similarities between the population genetic and epidemiologic approaches.
When I graduated in 1949, human genetics did not enjoy the cachet it does now, and job opportunities were limited. My options were to take an academic position at McGill or to go to Japan to work with Jim Neel on the genetic studies of the children of the atomic bomb survivors. I chose the latter. In so many unforeseen ways, that choice has determined much of what has happened to me since.
My first formal tie to epidemiology occurred in 1972, when I accepted a position at the University of Texas Health Science Center in Houston and was simultaneously appointed to a position in epidemiology in the School of Public Health. At the University of Michigan, where I had spent most of the previous 20 years, epidemiology was in the School of Public Health, whereas the population geneticists were in the Medical School. Although we had collaborative relationships, there was not a joint program of research.
LS: Looking at epidemiology as a discipline over time, what would you say have been the biggest changes in the field?
JS: I would say the subject matter. When I was becoming acquainted with epidemiology, the focus was on infectious diseases. Not much attention was paid to chronic diseases, although interest was growing in diabetes and cardiovascular disease. It was the 1960s or 1970s before the study of chronic disease took its proper place in epidemiologic research.
LS: In terms of looking at your career, who would you say most strongly influenced you?
JS: Without question, it would be Jim Neel. He and I worked together at Michigan for over 20 years, and then more distantly after I came to Texas. However, in so many ways, events that changed my life were initiated by Jim. He was a consummate investigator, a person of enormous vitality and interesting ideas.
LS: Collaboration has been a key part of your career. What do you think are the most important things are that support professional collaboration?
JS: Openness, fairness, and sensitivity. One’s relationship to one’s collaborators must be open and nurturing. There must also be a sense of fairness in sharing whatever credit might come from the research. No one enjoys being used, and in international research there must be cultural awareness. If your foreign collaborators do not perceive you as trying to understand the cultural differences, you will have difficulty.
LS: Much of your work has been in Japan. Can you say more about the Atomic Bomb Casualty Commission, or the Radiation Effects Research Foundation as it is now known?
JS: This is an interesting story in many respects. We learned soon after the genetic studies began in Japan that they offered a marvelous opportunity to study the consequences of inbreeding in human populations. We proposed a study involving the examination of some 6300 5- to 11-year-old children. It required bringing these children from their homes to examination centers, getting the school authorities to allow them to miss school, and the like. The negotiations to initiate this study seemed interminable. I think you could have floated a battleship in the green tea we drank. However, we were amply rewarded in the end. Once a contract was agreed on, whether verbal or written, the Japanese would adhere to the letter of that contract.
We learned, too, that although the groups with whom we negotiated did not necessarily expect their position to prevail, it was important to them that their position be heard. When working with school-age children, we found that we had to speak to the PTAs, to the major educational authorities in the study communities, and to the people at the individual schools. This took time, but the negotiations were never confrontational. As a result of our effort to be culturally understanding, we never had a problem. I am not aware of any other study of a group of largely healthy children that has had the level of cooperation we had in our study of inbreeding. It was fantastic. In Nagasaki, only one half of 1% of eligible participants refused to participate. The situation was not quite as good in Hiroshima, but the refusal rate was still less than 2%.
LS: In looking at professionals in the field of epidemiology in your lifetime, who would you consider to be the most influential?
JS: Many individuals have made truly significant contributions to epidemiology, and to identify only a few does a disservice to many. The 3 who come to mind are quite different. The first 1 would be Thomas Francis, the Chairman of the Department of Epidemiology at Michigan through part of the 1950s and 1960s. His scientific reputation was established by his work on influenza, and he would be assured a prominent position in science for that work alone. However, he made 3 less well-known, although no less seminal, contributions. First, he was the spark that drove the design of the study that evaluated the efficacy of the Salk vaccine. This was a huge undertaking, done at a time when large multiinstitutional collaborations were still in their infancy.
Second, he initiated one of the first continuous public-health surveillances of a community (the Tecumseh, Michigan study). Finally, he chaired the small group, known as the Francis Committee, which was responsible for establishing new purpose and direction for the studies of the atomic bomb survivors. These studies had lost focus and were drifting. Francis and his committee recommended a series of changes that have served the institution well for 50 years, a truly remarkable achievement.
The second person would be Alexander Langmuir. His perspective on epidemiology attracted me; he believed strongly in the need for an epidemiologist to be involved in the problem he was studying. You went, you saw, you consulted, you interviewed, and you actively participated. He called this “shoe leather epidemiology.” This was the model he imposed on the Epidemic Intelligence Service while he was at the Centers for Disease Control. Establishment of the EIS was a formidable contribution in itself.
The third person would be Reuel (Stony) Stallones. Stony’s importance to epidemiology is based on his vision of the education of public health professionals, including epidemiologists. He held that public health education could be seen as a “matrix” in which the columns and rows were equated to problem areas and scientific disciplines. He contended that this view offered students flexibility and a deeper insight into the role of the various disciplines in public health. Although his notion was not as widely accepted as I think it warrants, it certainly stimulated an enormous amount of self-inspection at other schools of public health.
In addition, I admired Stony’s candor and the often-acerbic quality of his comments on the way epidemiologists behave or behaved. An example of this was his assertion that arithmetic computations are merely an answer to an equation; they are not the solution to a problem. We sometimes forget this fact and assign too much importance to the role of statistics in epidemiology.
LS: What would you say has been your most influential paper, and which of your papers do you think has made a greater contribution than has been appreciated?
JS: I am not really sure. One publication that probably has not received the recognition it warrants is the report on the findings on pregnancy terminations in Hiroshima and Nagasaki published by the National Academy of Sciences in 1956. In retrospect, the Academy was not the best choice of a publisher; its books are not well publicized. Although 2000 copies of this book were set aside for free distribution, they were gone in a short time. Thus, the book was not available to give to others who came into this area of research later and who could have had a greater interest in the findings. Presumably, they could have found the book in a library, but, as you know, monographic reports are often very difficult to find. No one seems to know how to catalog them. The book was republished in 1991 by the Academy in the Children of the Atomic Bomb Survivors.1 However, this probably did not solve the problem; it just continued it in a new cover.
There were 3 things we did in this monograph that were novel for the time. One was the extensive use of multivariate analysis. Clearly, the theoretical foundations of multivariate analysis were laid much earlier by Wishart, Hotelling, and others. However, in 1956, the application of this theory was limited by the fact that calculations had to be done on a mechanical calculator, a tedious and intimidating task. A second novelty was the use of variance component analysis to estimate the role of radiation exposure in determining birth weight. Although variance component analysis had been used in other areas of biology, it had nit been used in the human situation. The third novelty was the use of power calculations, decades before these calculations were de rigeur. We did these calculations in a somewhat different way, retrospectively rather than prospectively.
Although these 3 things were novel in 1956, by the time the monograph was republished in 1991, all had become a part of the standard tools of every epidemiologist and statistician. This was largely the result of the availability of powerful computers and software that allows one to do so much without knowing what one is actually doing.
In the past, when one knew that many hours of labor would be invested in the analysis of a particular set of data, one thought long and hard about what to do and why one wanted to do it. What is now seen as “data mining” we disparagingly saw as “fishing” expeditions. Analyses are not driven by a prior line of reasoning, but because the data and the programs are there and the analysis can be done quickly.
LS: That leads to the matter of how one chooses a research question. Is there a guiding principle you have used to decide what to look at?
JS: To a considerable extent, my research reflects my association with Jim Neel. Much of Jim’s research focused on mutagenesis. He was interested in the occurrence of mutations, spontaneously or as a result of exposure to mutagenic agents such as ionizing radiation. These issues were of considerable moment to me, too, and this naturally led to an effort to define and assess those factors that affect the loss and the spread of mutant genes in populations.
LS: In looking at your career, what contributions to the field do you take the most pride in?
JS: Actually the studies in Japan, and not just the genetic ones, but the teratogenic and carcinogenic ones as well. I suppose it is fair to say that I am one of the few persons who have contributed to all 3 of these areas of research in Japan. Jim was focused solely on the genetic effects, whereas most of the other individuals I have known were interested primarily in the carcinogenic effects. The teratogenic findings are, to my mind, some of the most interesting. Some of these observations were unanalyzed for almost 30 years because no one at the Foundation was aware they had been made. Fortunately, I had been in Japan when these measurements were carried out and I was certain that they had to be someplace findable.
This leads me to wonder whether we are missing other equally relevant observations. I doubt that anyone now associated with the Radiation Effects Research Foundation in Japan was there before 1965. The institution was already 20 years old by then. Some of the most critical developments in its history such as the establishment of the program of research recommended by the Francis committee had already taken place. Many of the people who were responsible for these changes are no longer alive, and a lot of institutional memory has been lost.
LS: What do you enjoy doing outside of your professional activities?
JS: My interest in epidemiology and population genetics has whetted an appetite for travel. This interest has also aroused an awareness of the history of science. I feel strongly that a means must be found to perpetuate the history of the Japanese studies. I would not have so felt 50 years ago. If you had asked any member of the Atomic Bomb Casualty Commission in 1950 if these studies would continue for 50 years, the answer would have been a resounding “no.” Well, 53 years later, these studies still continue and exhibit every indication of continuing for another decade or two. Had we foreseen this, we might have documented things more diligently, or organized the information that has been collected differently, so it would be more accessible to those who have followed.
LS: You and your wife have set up a foundation, right?
JS: Not quite. The Schull Institute came into existence through the efforts of a group of my colleagues and students who are interested in education in health care, with a special emphasis on vulnerable and indigent populations and on the application of new techniques such as telemedicine to distribute health care more widely. We are primarily concerned with training people who have a strong commitment to community service. Our goal is to bring better and higher-quality medical care to populations that are isolated either geographically or socially and are not in the mainstream of medical attention.
We are a small organization, only 2 fellows at the moment, and for the foreseeable future, we plan to remain that way. Smallness gives us flexibility. One of our interests is in scholarship. I see scholarship as a process, not an end in itself. I think that we have lost a lot in the last half century in terms of the kind of scholarship that characterized science in the first half of the 20th century and through most of its still-earlier history.
LS: What do you think has been the greatest contribution of epidemiology?
JS: Stony often said that the greatest contribution to public health was the flush toilet. The truth of this remark was brought home when we became interested in studies along the Rio Grande Valley. Infant mortality in this area had dropped dramatically after the valley had a closed sewage system. More has been achieved through developments like this one, through understanding the mechanisms of transmission of infectious disease, than by identifying yet another risk factor that no one seems able to incorporate into a health maintenance program, or through references to the wonders of molecular biology.
LS: If you look at the current state of epidemiology, what do you think looks healthy and what looks less than healthy?
JS: I am not pleased by what I see. I think epidemiology has fallen victim to its own hyperbole. Simple, obvious advances are now labeled “breakthroughs” and touted as if they will bring health to all. However, few of these alleged breakthroughs will be remembered 1 year later, and even fewer will significantly influence health. This emphasis on self-promotion has led or will lead to loss of respect for the field. I do not believe that we will ever see again the ivory-tower academic life that prevailed a half century ago; nevertheless, there were elements of that life that remain important in education today.
LS: That sense of scholarship you mentioned earlier.
JS: Yes. I have also been disturbed by epidemiology’s apparent loss of the capacity to recruit clinically qualified people. A half century ago, most epidemiologists were clinically trained, and they brought to the discipline a different mindset. They were more insistent on the demonstration of biologic plausibility before causation was invoked than now seems true. In my view, every student of epidemiology should know Bradford Hill’s criteria for causation. Temporal specificity, biologic plausibility, coherence; we do not pay enough attention to these criteria. I find this unfortunate.
As epidemiology has become attractive to more and more people, many with admirable social motivations but without a clinical or even a human biology background, biologic plausibility does not have the import that it should.
LS: Where do you think the field is heading? Do you have any advice for students today?
JS: Modern science oscillates between going to where the money is and its own sense of propriety and order. It is important in this context to encourage students to examine the writings and arguments of naysayers such as the late Peter Skrabanek. I think we have swung too far one way, but I also believe that the pendulum will come back. We need to encourage our students to look deeply at the things they do and why they do them.
About the Interviewer
Lowell Sever is a perinatal epidemiologist at the University of Texas at Houston. He first became acquainted with Dr. Schull’s work on the teratogenic effects of radiation exposure when he carried out a study of congenital malformations in communities around the Hanford Nuclear Site in the early 1980s. He later served on an Institute of Medicine Committee in 1995 chaired by Dr. Schull.
The editors are grateful to the Dean’s Office and Administration of the University of Texas School of Public Health for making possible the videotaping of this interview. Sara Ann Barton was especially helpful in providing coordination and general support. Joe Smith was the videographer and Michele Mocco the photographer. Judy Eshelman carried out the transcription of this interview.
© 2004 Lippincott Williams & Wilkins, Inc.