Sir Michael Marmot Interview: Conversations with History; Institute of International Studies, UC Berkeley
|Photo by Jane Scherr|
Page 2 of 4
What does an epidemiologist do?
Lots of things. But essentially, the thing that distinguishes epidemiology is the study of disease in populations. So to contrast what I did as a clinician, in clinical medicine, [with] what I do in epidemiology, clinicians are trained to treat individual patients, that's what they do first, last, and always, they treat individual patients. What I was doing in my own primitive way in Sydney, asking, "Why are these kids getting asthma again and again? Why are they coming back? What is it in the environment in which they live that's causing these kids to come back and back with asthma?" was thinking like an epidemiologist, was thinking about causes that are out there in the population. So what epidemiologists do is study disease in populations, and they study the causes of disease in populations.
Now, my own view is to do that effectively, they need to do it in collaboration with other disciplines. So epidemiology has its own set of methods -- we tend to do large population studies, we follow large groups of people, and so on -- but ideally, to try to understand the problems that we're seeking to understand, we need to do it in concert with others.
What are the sources of the paradigms, let's use Kuhn's words, that guide you in the things that you look at? We can move into talking about the Whitehall Study, which is a body of data that you've done extensive work with. But what I want to get at is, what shall we look at given this data?
One important paradigm, one important understanding, is that the causes of individual differences in disease may be different from the causes of population rates of disease. Now, I could generalize it and say that all disease is a combination of genes and the environment. But let me illustrate.
In England and Wales last year, there were about 150,000 deaths from coronary heart disease, and the year before there were about 150,000 deaths, and this year there will be slightly fewer because the rate is going down. But there won't be 100,000 and there won't be 200,000. If it's not 149,400, it will be 149,300. Now, the people who died last year won't die again this year. So we can ask why did those individuals who died last year die, and other individuals not die? That's one sort of question. Or we can ask why is the rate such as it is? Why do men have a higher rate than women? Why is the rate in Hungary higher than in England? Why is the rate in Scotland higher than in England? And why is the rate in Italy lower than in England? And do people who migrate from Italy to England get higher rates when they migrate from Italy to England? We could know that's the case without knowing very much about any individual Englishman or Scotsman or Italian, [because] we know these general patterns.
Now, it's crucial to know something about the individual Englishman, Scotsman and Italians or Hungarians to understand the problem properly, but the determinants of population rates of disease are different, potentially, from the determinants of individual difference. Or they may be different, one should not assume that they're the same. That's really difficult for somebody who is trained in looking at the individual patient to come to grips with initially. If you tell me that somebody who migrates from Italy to England is subject to a higher risk of disease when he gets to England, doesn't it depend on which individual he is? And the answer is, of course it does. But standing back, you see this general pattern relating the characteristics of societies to those characteristic societal rates of disease; that's terribly important. So when we come to look at inequalities in health, differences between different social groups, we don't necessarily understand why differences between different social groups comes about only by studying individuals.
Now, in a way, one could say that this is a radical perspective, is it not? Because when you think about looking at an individual patient, then there is a whole panoply of organization that will provide you with drugs to treat this individual, without going back to the question of how did this group of individuals come to me.
Yes. Oh, it's quite different.
I had two very important teachers, when you were asking about early influences and I was drawing a blank, but I had two very important teachers: Geoffrey Rose in London, and before him, Len Syme at Berkeley, and they both had a big influence. And, you know, in a funny way, coming from very different backgrounds and very different perspectives and, really, very different approaches to the world, in a funny way they had a rather consistent message.
Geoffrey Rose said that when he started teaching medical students about epidemiology and public health, he used to say the medical student should ask the question, "Why did this patient get this disease at this time?" And he then later in his life decided that that was a rather restricted question. That was the question about this individual, and that's an appropriate question for a clinician to ask. But to ask, "Why am I seeing so many more patients from this part of the city than that part of the city?" is a different sort of question. "Why is the rate of disease different?" is a different question.
The way Len Syme used to put it was that there are three questions one should ask: If somebody got sick, what disease would he get? The second question is, why would one individual get sick and not another? And the third question is, why is the rate of disease higher under these circumstances and not another?
They were really very similar perspectives coming from very different angles, but asking the same sort of questions. So it is quite a radical question. It's looking at the determinants of illness, and that leads you into looking at the societal determinants of illness.
Next page: The Whitehall Studies
© Copyright 2002, Regents of the University of California