Share this article on:

A Conversation With Frank Speizer

Dockery, Douglas W.

doi: 10.1097/EDE.0b013e3182117ec2


From the Harvard School of Public Health and the Channing Laboratory, Brigham and Women's Hospital, Harvard Medical School, Boston, MA.

Editors' note: VOICES is a project of the journal to provide personal, historical, and scientific perspectives on the field of epidemiology, as seen through the eyes of the field's most senior and accomplished practitioners. Readers are welcome to nominate candidates for the editors' consideration.

This interview was conducted at the Harvard School of Public Health on 19 September 2007. Frank Speizer has approved this transcript for publication.

Frank Speizer's curriculum vitae is available with the online version of this article at

Correspondence: Douglas W. Dockery, Harvard School of Public Health, 665 Huntington Ave, Boston, MA 02115. E-mail:

Frank Erwin Speizer was born in San Francisco on 8 June 1935. He received his MD from Stanford University in 1960, and did his internal medicine residency at Boston City Hospital and Stanford-Palo Alto Hospital. His epidemiologic career was shaped by fellowships at the Harvard School of Public Health and the British Medical Research Council Statistical Research Unit. Returning to the Channing Laboratory at Harvard Medical School, he started a research program in the epidemiology of chronic diseases. This included 2 seminal prospective cohort studies, the Harvard Six Cities Study of the health effects of air pollution and the Nurses' Health Study of the causes of cancer, heart disease, and other diseases among women. He also initiated a training program in the clinical epidemiology of lung and heart disease that has produced many of the leaders in the field. He has been active in the translation of scientific evidence into policy, eg, as a member of the Clean Air Scientific Advisory Committee of the US Environmental Protection Agency. He is a Member of the Institute of Medicine of the National Academy of Sciences. His awards include the John Goldsmith Award for Outstanding Contributions to Environmental Epidemiology (International Society for Environmental Epidemiology), the William Silen Lifetime Achievement in Mentoring Award (Harvard Medical School), the World Lung Health Award (American Thoracic Society), the Charles S. Mott Prize (General Motors Cancer Research Foundation), and the Excellence in Women's Health Award (Jacobs Institute of Women's Health). He currently is the Edward H. Kass Distinguished Professor of Medicine at the Harvard Medical School and Professor of Environmental Science at the Harvard School of Public Health.

Back to Top | Article Outline


DD: Few people start out with the intent to be an epidemiologist. How did you get on this pathway?

FS: Early on in medical school, it was clear I wasn't a typical student. I couldn't memorize as well as my fellow students, and I was much more interested in the mathematical basis of things. I struggled to find a career that combined medicine and mathematics. At the end of my first year, for personal and social reasons, I needed a job in Los Angeles. John Goldsmith, who had just come from the Harvard School of Public Health to the California State Health Department, was doing air pollution epidemiology in Los Angeles. He offered me a summer job in air pollution health research. I wasn't particularly interested, but I discovered there was lots of mathematics in what was then the field of environmental epidemiology (not named as such) that could be combined with medicine. I got some experience in respiratory disease using pulmonary function tests in men at a veterans' hospital in Los Angeles. I was turned on by that.

DD: So how has the field changed since then? What do you see?

FS: We measured air pollution at the Wadsworth VA in Los Angeles by whether we could see the mountain. This probably was a good measure of small particles in the air. From the standpoint of monitoring exposures, there has been dramatic change over the last 40 to 50 years—in terms of technology as well as understanding what's in the air. There have been very dramatic changes in terms of instrumentation and increased collaborations between exposure scientists and epidemiologists. We've developed a wide variety of standardized approaches for getting historical information, by questionnaires as well as physiologic tests.

DD: From those air pollution studies, you went on to the large, prospective cohort studies.

FS: My main interest was not just air pollution. I was interested in trying to understand the natural history of chronic respiratory diseases. I also recognized that one person can't study the natural history of chronic respiratory disease—we don't live long enough. We developed the idea that with overlapping, relatively short-term (in terms of life span) prospective studies, say 10–15 years, in different age groups, we could piece together the natural history of chronic respiratory disease. That became one of the main focuses of my work.

DD: So how did you transition to that?

FS: I had the good fortune to be able to stand on the shoulders of a number of giants. I worked with people who provided me with the opportunities to get a basic understanding of how you do this work, the stimulus for areas in which the work ought to be done, and how to bring this to a larger scale—what today is called a “big science” approach to the problems. I spent a couple years at the Harvard School of Public Health, working in Jim Whittenberger's department. I learned basic physiology working with exposures to sulfur dioxide, in dogs and humans, with some very competent physiologists.1 At the same time, I worked with Ben Ferris in what might be called the early field studies of chronic respiratory disease.2 I had the opportunity in England to work with Sir Richard Doll, the grandmaster of long-term cohort studies, and with Charles Fletcher and Richard Peto. I came back with an understanding of how to study the natural history of chronic respiratory disease. Jim Whittenberger, Ben Ferris, and I started the Harvard Six-City Study, which I saw as an opportunity to study the natural history of disease from childhood and adult life, along with a study of air pollution effects.

DD: Soon after, you started the Nurses' Health Study. What was the motivation for that?

FS: The idea for the Nurses' Health Study started about 10 years before the actual study. When I was in England in the 1960s, working in Richard Doll's unit with Martin Vessey, we began seeing reports of healthy young women with pulmonary emboli and deep vein thrombosis in their legs. The association of these conditions with oral contraceptives was pretty apparent—it didn't take long to figure that out. What also was very clear was that hundreds of millions of healthy women were being put on a potent drug, and no one was considering the potential long-term consequences. There I was, sitting in the basement of Richard Doll's research unit, with shelves of forms used in the doctors' study of smoking and lung cancer, and it just seemed natural to consider setting up a cohort of women to assess long-term effects of oral contraceptives.

We started out with a pilot study of doctors' wives in England because doctors were registered and it was easy to find them. It was a way of getting a population that could be followed prospectively. We had a very effective pilot study with about a 95% response. Doll signed the letter that went out, which helped. We were about to launch a larger study when we realized there were only about 22,000 women in age groups we needed. This was not a large enough population to do the study.

I came back to the United States and did a very similar pilot study here, again using doctors' wives. It was effective enough—we got about 70% response rate in the initial pilot study, so we submitted a grant. Just before we got the funding, we thought we better recontact the doctors' wives. This time we tried 3 approaches. One was to send the questionnaire to the doctor as before, and ask him to give it to his wife. The second was to send it directly to the wife, addressed as “Ms.” The third approach was to send the questionnaire to the wife addressed as “Mrs.” The worst response rate was from the “Ms.” group. In 1973–1974, married women aged 30–55 apparently did not like to be called “Ms.” The most important group was the “Mrs.” group—many of whom said they had never seen our first questionnaire! Their doctor-husbands had filled it out for them. There was no way we could do a study in which the doctors were filling out questionnaires for their wives—which is why we switched to nurses, probably the reason the study has been so successful in following a large population of women.

DD: And that study evolved into a series of prospective cohort studies?

FS: In 1980, we added a large dietary component through Walt Willett's efforts. In 1986, Walt led the effort to start the health professionals' follow-up study, a similar population-based study of men. In 1989, we added a group of younger nurses to get at the question of the use of oral contraceptives before their first pregnancy (Nurses II). We could not answer that question with women in the first cohort because most of them didn't have the opportunity to use oral contraceptives before their first pregnancies. In 1993, Graham Colditz started a study of offspring of the Nurses II cohort, so we could follow a group of children.

DD: How has epidemiology changed from those early studies to what we're doing now?

FS: There have been dramatic changes. It's not clear that large-scale mail-based long-term cohorts like we established are any longer feasible or affordable. New methods, more electronically based, will be necessary.

DD: Web-based?

FS: Yes, Web-based approaches rather than paper-and-pencil approaches. These are being piloted by us, as well as others. It's also not clear how many more of these cohorts of middle-aged people are needed. There are a host of questions related to the origins of disease that drive us to think more about younger populations. It's more difficult to follow younger populations. Once kids get out of school, it's hard to continue to follow them. Our approaches will have to change. When we started, people thought you couldn't get mail responses as high as we got. I'm not sure we'd get them now. Funding for such cohorts must be secure. Once one starts a cohort, it becomes a long-term commitment as the results become more and more powerful as time goes by.

The other aspect that's changed dramatically is the biochemical and genetic side of it. This was why we collected blood from the NHS women in 1989. I think we've answered the simple questions. The more complicated ones require a combination of information from questionnaires, phenotypes, physiology, and biochemistry, as well as genetics.

DD: You've mentioned that you've stood on the shoulders of giants. Who have been the most influential people in your career?

FS: I've thought about this question. I cannot come up with a single name. I've been fortunate to have worked in several different areas, and each of them has had very important people—outstanding scientists who have been willing to share their ideas and then give me the freedom to do what I want to do. Certainly Jim Whittenberger played an important role. When I told him I wanted to finish my clinical training, he made one phone call and I had a job on the Harvard Medical Services, Boston City Hospital. When I got to City Hospital, I worked with some very outstanding clinicians as well as laboratory researchers. When I was going off to England and told Ed Kass I would like to come back to Boston, Ed saved a place for me. Certainly the opportunity to work with Richard Doll was unique. Doll was an inspirational character. He heard my interests in respiratory disease, and connected me with Charles Fletcher, who was doing a prospective study of the natural history of chronic respiratory disease. I worked with him for 15 years on that study.3 Each of these people played a very important role. I shouldn't leave out John Goldsmith, who showed me how one could do epidemiology and be a practicing clinician at the same time.

In addition, a major advantage of being at Harvard is that one is surrounded by wonderfully smart, enthusiastic young people. I saw my job as describing the things I wanted to get done, and providing an infrastructure and database, so people who would also become giants could succeed and become independent leaders in their own right.

DD: What do you think makes a good collaboration?

FS: I think I learned this on the wards of Boston City Hospital. I was an assistant resident, which meant I had interns under me for whom I was responsible. What I learned from the senior people at the hospital is that there is more than one way to do something. Your job in that hierarchy is to make sure that the people you're responsible for do their job right, but not necessarily exactly the way you might do it. As long as they didn't hurt anybody, and the job was done in a competent way, it was okay. I learned to give people the freedom to do their thing as long as they didn't make mistakes. It's hard in clinical medicine to sort that out sometimes, and it's even harder on the research side. I think that's an important component of what people need in a training position. I think the main thing is to give—as was given to me—respect and the freedom to do your thing.

The other thing I have tried to do is to ask questions in ways that make people think about what the issues are. What I've learned over the years is that if I have a question about a particular issue, I'm probably not the only one in the room with that question. People tend to hold back and not ask questions, whereas I tend to ask them so that the issue can be brought to light and discussed. This is an important component of learning for everybody.

DD: How have you picked the research questions you wanted to pursue?

FS: My research questions are pretty fundamental. What's the natural history of chronic respiratory disease, and what are the components that cause the disease to get worse? Air pollution is one aspect of it. Cigarette smoking is obviously another. Genetic interactions are questions that spin off from the basic question of natural history of disease. In the Nurses' Health Study cohorts, the questions are the natural history of what happens to people as they get older, and the interaction with lifestyle and behavioral factors. Obviously, one needs to have hypotheses worth testing along the way. The fundamental idea is to look at what's happening to people. I set up the infrastructure and provided some of the basic hypotheses, but its other people's interests that have made for an ongoing productive enterprise.

DD: That infrastructure has been remarkably productive. I looked this morning at PubMed, and you have almost 500 publications.

FS: That reflects mostly other people's work at this point. Again, it's a matter of working with people who are competent to do their jobs, and being there to help them do the work.

DD: What are the accomplishments you're most proud of?

FS: Most proud of? Obviously they fall in different categories. On the respiratory disease side, 3 issues come to mind. Early in my career, I recognized that there was a rising death rate of asthma in children in England. While working with Doll we explored that, and the published results were subsequently claimed to have prevented about 3500 deaths in children with the disease.4,5 The natural history of the effects of cigarette smoking were pretty well known early on, although I think we substantiated some of those more completely, which led us to identifying the impact of passive smoking. Even though the physiologic effect on children was relatively minor, the findings brought home the importance of thinking about environmental tobacco smoke.6 From there, other people picked up the policy issues, and the importance of changing behavior of people on a large scale. Finally, I think the results of the Harvard Six-City Study had a major impact on setting air quality standards in this country.7 Along the way, there were important changes in how monitoring is done in this country, stemming from the experience in the Six Cities Study. You, Jack Spengler, and others, had important roles in changing the way things are monitored and documenting the health effects that changed the basis for setting air quality standards.8

In the Nurses' Health Study, I like to say we proved that what our grandmothers told us was true. Smoke, my grandmother said, will stunt your growth. So don't smoke; don't get fat; exercise—we put some numbers and documentation on those lifestyle things.

Oral contraceptives, which were potent drugs in the early 1970s, have lowered their estrogen content by a factor of 3 logs. Although there are quantifiable effects of these hormones, they are not as dramatic as we expected.9 Perhaps of greater importance has been quantifying the effects of postmenopausal hormones.10

DD: How do you see the influence of epidemiology now?

FS: I think the influence of epidemiology will continue to grow, with more attention to complex interactions. Journalists want simple answers; however, there are no simple answers. The field is in a healthy state. There are a lot more people doing the work. We are not going to find relative risks of 20–30 very often, but we can still try to understand what's going on, physiologically and biochemically, in people. I think the field will become more collaborative with basic scientists, physicians, and statisticians. All these fields come together in what's now being called “Big Science” to answer the big questions.

Short-term prevention probably doesn't pay off, but it does in the long-term. The only way you can study prevention is to do it through epidemiology. I don't think waiting for people to get disease is the answer. As a society, we can't afford to wait for people to get sick to treat their diseases, even if the treatments get a lot better. They're just too expensive. We've got to work toward trying to understand risk factors to prevent diseases. That can only be done with the leadership of epidemiology.

DD: What is your advice for a young epidemiologist starting today?

FS: It's a tough environment to work in at the present time. I think we're going to be moving more and more toward larger groups, and we have to figure out a way to reward the junior people in those groups. What I tell people (and have told them for years) is that they need fundamental training in epidemiology, along with training in statistics, and the basic understanding of physiology, chemistry, and genetics. They don't necessarily need to go into the laboratory and do it, but they should learn the language. They need to find a place to work where they can apply those skills and have fun doing it. If they're not having fun they ought to find a different place to work. The message has got to be that prevention will pay off. So we've got to continue to push that message to get more money put into prevention.

DD: You've always emphasized the importance of enjoying what you do. What do you do for fun?

FS: I still come to work and ask questions. I think fun is stimulating people to think about topics, and make suggestions. That's my professional fun. I've taken up painting in my old age, but that's been more learning for me than fun, so far.

DD: And you've always emphasized family ...

FS: Institutions have neither hearts nor memories. The people who will cry at your grave are your family and friends. So you've got to keep them prominent in your life. Having had the good fortune of being married to the same person for over 50 years, and still being good friends, we enjoy together our 8 grandchildren, even though they're scattered all over North America.

Back to Top | Article Outline


We gratefully acknowledge the support of the Department of Environmental Health, and Christopher Riley of Media Services of the Information Technology Department of the Harvard School of Public Health in making the video recording of this interview.

Back to Top | Article Outline


DOUGLAS W. DOCKERY is Professor of Environmental Epidemiology and Chair of the Department of Environmental Health at the Harvard School of Public Health. He received his ScD from the Harvard School of Public Health where he was mentored by Frank Speizer, Benjamin Ferris, and John Spengler. He has coauthored almost 50 papers with Frank Speizer on lung development, chronic respiratory disease, and the health effects of air pollution.

Figure. F

Figure. F

Figure. F

Figure. F

Back to Top | Article Outline


1. Speizer FE, Frank NR. A comparison of changes in pulmonary flow resistance in healthy volunteers acutely exposed to SO2 by mouth and by nose. Br J Ind Med. 1966;23:75–79.
2. Speizer FE, Ferris BG Jr. The prevalence of chronic nonspecific respiratory disease in road tunnel employees. Am Rev Respir Dis. 1963;88:205–212.
3. Fletcher CM, Peto R, Tinker CM, Speizer FE. The Natural History of Chronic Bronchitis: An Eight Year Follow-up Study of Working Men in London. Oxford: Oxford University Press; 1976.
4. Speizer FE, Doll R, Heaf PJ. Observations on the recent increase in mortality from asthma. Br Med J. 1968;1:335–339.
5. Speizer FE, Doll R, Heaf P, Strang LB. Investigation into use of drugs preceding death from asthma. Br Med J. 1968;1:339–343.
6. Tager IB, Weiss ST, Rosner B, Speizer FE. Effect of parental cigarette smoking on the pulmonary function of children. Am J Epidemiol. 1979;110:15–26.
7. Dockery DW, Pope CA, Xu X, et al. An association between air pollution and mortality in six U.S. cities. N Engl J Med. 1993;329:1753–1759.
8. Spengler JD, Ferris BG Jr, Dockery DW, Speizer FE. Sulfur dioxide and nitrogen dioxide levels inside and outside homes and the implications on health effects research. Environ Sci Technol. 1979;13:1276–1280.
9. Hankinson SE, Colditz GA, Manson JE, et al. A prospective study of oral contraceptive use and risk of breast cancer (Nurses' Health Study, United States). Cancer Causes Control. 1997;8:65–72.
10. Buring JE, Hennekens CH, Lipnick RJ, et al. A prospective cohort study of postmenopausal hormone use and risk of breast cancer in U.S. women. Am J Epidemiol. 1987;125:939–947.
Back to Top | Article Outline

BONUS: Online Exclusive



John Hutchinson, 1811–1861: The First Respiratory Disease Epidemiologist

Frank Speizer recounts the remarkable story of a young Victorian-era physician who pioneered the epidemiologic study of lung function more than a century before respiratory disease became an epidemiologic specialty.

Supplemental Digital Content

Back to Top | Article Outline
© 2011 Lippincott Williams & Wilkins, Inc.