publisher colophon
M. G. Marmot - Understanding Social Inequalities in Health - Perspectives in Biology and Medicine 46.3 Supplement Perspectives in Biology and Medicine 46.3 Supplement (2003) S9-S23

Understanding Social Inequalities in Health

Michael G. Marmot

[Tables]

Abstract: A prominent feature of health in all industrialized countries is the social gradient in health and disease. Many observers believe that this gradient is simply a matter of poor health for the disadvantaged and good health for everyone else, but this is an inadequate analysis. The Whitehall Study documented a social gradient in mortality rates, even among people who are not poor, and this pattern has been confirmed by data from the United States and elsewhere. The social gradient in health is influenced by such factors as social position; relative versus absolute deprivation; and control and social participation. To understand causality and generate policies to improve health, we must consider the relationship between social environment and health and especially the importance of early life experiences.

 

IT IS A PLEASURE TO HONOR the contributions of Alvin Tarlov. It may seem surprising that a man who was Chairman of the Department of Medicine at a top medical school should be leading the charge on the social determinants of health. It would indeed be surprising in someone who did not have Al's breadth of knowledge and insight. Al recognized very early that a full understanding of how health and disease are distributed in society requires an investigation of how social processes affect biology. He has fostered efforts to improve knowledge of the links between society and health, and has nurtured those few of us who have labored in this particular vineyard. [End Page S9]

He and Fraser Mustard latched on to an early finding from the Whitehall Study of British Civil Servants (Marmot et al. 1978). They encouraged me to pursue it scientifically and stimulated many of us to think about its implications for health policy and for society. It is that finding, and its implications, that I propose to explore.

Demonstrating a Social Gradient

One of the dominant features of the health situation of all industrialized countries is the social gradient in health and disease. The Whitehall Study of British Civil Servants showed that, even among people who are not poor, there is a social gradient in mortality that runs from the bottom to the top of society (Marmot, Shipley, and Rose 1984). Twenty years after the original Whitehall study, the Whitehall II study documented a similar gradient in morbidity (Marmot et al. 1991).

It came as no surprise to most observers that British society would be highly stratified. But the social gradient in health is not a unique product of the British class system. To quote just one American example, the Panel Study of Income Dynamics classified people according to household income and demonstrated a continuous gradient in mortality (McDonough et al. 1997). The poorest people had the highest mortality rates, while the mortality rates of people in the middle income range were intermediate between those at the bottom and those at the top. Once again, the differences were not limited to poor health among the poor and good health for everyone else.

Them and Us?

The demonstration of the social gradient in health requires that we go beyond binary thinking and appreciate that we are not dealing simply—as if it were simple—with the problem of absolute deprivation and health. We need to understand inequalities. Among the many possible reasons for health inequalities, perhaps the most uncomfortable are inequalities in society, because they suggest that there is not a straightforward technical solution to the problem. If, for example, health inequalities could be eliminated by equalizing access to high-quality medical care, there would be direct political, social, and organizational approaches to reducing these inequalities. The challenge would not be either understanding how features of social organization affect health or working out how to improve the conditions in which people live and work. These, however, are just the challenges posed by the social gradient in health.

Emphasizing the gradient should in no way divert attention from the astonishingly poor health of those suffering from social exclusion. For example, Geronimus, et al. (1996), compared the health of blacks and whites in rich and poor communities. The probability that a white male would survive from age 15 to 65 years was 77 percent. But among blacks in one poor community in New [End Page S10] York, the probability was only about 37 percent. If current mortality rates persist, nearly two-thirds of the young black men in this deprived community would die before the age of 65, compared to less than a quarter of the young white men. We need to be concerned with the problem of social exclusion as well as with the social gradient in health.

By and large, wherever investigators have had the data to analyze, they have observed the social gradient in health. There is a clear relation between position in the social hierarchy and mortality, not only in Britain and the United States, but also in the countries of Europe and Australasia (Eckersley, Dixon, and Douglas 2001; Kunst and Mackenbach 1994; Mackenbach et al. 1997; Shkolnikov and Cornia 2000). Although the social gradient is steeper for some diseases, and is absent, or even reversed, for a few, it is clearly present for most of the major causes of death (van Rossum et al. 2000).

So Human an Animal

Nonhuman primates exhibit a similar social gradient in the risk of cardiovascular disease and the degree of atherosclerosis—the pathological process that underlies clinical heart disease (Sapolsky 1998). Carol Shively's (2000) studies of cynomolgus macaque monkeys are illuminating. Not surprisingly, male macaques have more atherosclerosis than females, but removing the ovaries of a female renders her as liable to atherosclerosis as a male. The effect of social organization can be as dramatic as that of ovariectomy. These monkeys live in troops, and within the troop there are clear social hierarchies. A subordinate female is as liable to atherosclerosis as one whose ovaries have been removed.

One explanation for this could be that whatever causes a monkey to be subordinate is also responsible for increased liability to atherosclerosis, but this explanation is contradicted by experiments in which animals change status. Females' social ranks are fairly stable. The way to change ranks, therefore, is to take high-status monkeys from different troops and put them together. In the new troop, a new hierarchy is formed. As a result, some previously dominant monkeys will now have a subordinate place. Similarly, low-status monkeys can be put into new troops and they, too, form new hierarchies. In general, the level of atherosclerosis is related to rank in the new hierarchy.

The situation for males is slightly different. Under stable conditions, subordinate male macaques have more atherosclerosis than do dominant animals. Under unstable conditions, in which the monkeys from different troops are mixed together, the dominant monkeys have more atherosclerosis. This reversal of the protective effect of being dominant is prevented by pre-treatment of the animal with a beta-adrenergic antagonist (Kaplan et al. 1987). These data are consistent with an effect of stress on the sympathetic nervous system that increases the risk of atherosclerosis. [End Page S11]

The Poor Are Always With Us

Social gradients in health in British civil servants, Americans, Canadians, Scandinavians, Dutch, Italians, Australians, and New Zealanders—in human primates and non-human primates! There is a line of argument that runs as follows. Perhaps a social gradient in health is a feature of all societies: we can no more do anything about it than ask that when a race is held, everyone should come in first. If inequalities in health are related to inequalities in society, isn't a social gradient in health a feature of all societies? Not only would a society where everyone was equal be gruesome, it would be unachievable. This implies a certain inevitability to health inequalities.

My response to this is twofold: health status for everyone can improve, and the slope of the gradient can change (Marmot 2002).

Health Status for Everyone Can Improve

A dramatic illustration of overall improvement can be made by comparing inequalities in mortality one hundred years ago with present figures. In 1901, Benjamin Seebohm Rowntree reported on infant mortality rates from the northern English city ofYork. He classified the "working classes" into social levels based on his assessment of the quality of neighborhoods in which they lived. Infant mortality per 1,000 births was 173 in the highest; 184 in the middle working class; and 247 in the poorest; compared with 94 in the servant-keeping class. The poorest thus had infant mortality rates 2.6 times higher than the richest. In the year 2000, infant mortality rates for England and Wales range from 3.7 per 1,000 in social class I (the highest, equivalent with the servant-keeping class described in Rowntree) to 8.1 in social class V (the lowest). (For single mothers it was 7.6 per 1,000.) The poorest people (social class V) in 2000 had infant mortality rates less than one-tenth those of the richest people 100 years previously.

Of course, the fact that a century ago, everyone was subjected to the same dirty water and unhygienic practices could account for the improvement. But could there still be an improvement in health across all social groups when poor hygiene was no longer the dominant cause of ill health? More recent data suggest there could. Table 1 shows life expectancy for social classes in England and Wales in the 1970s and the 1990s (Hattersley 1999). (These figures come from the ONS [Office for National Statistics] Longitudinal Study, which started in 1971 and follows a 1 percent sample of the Census. The analysis is based on the Registrar-General's social classes, the standard social classification system in Britain, I being higher professional and V unskilled manual laborer.) There are two important points to take from this table. First, these are national figures, not confined to British civil servants, but the gradient in life expectancy is clear in both time periods: the lower the status, the shorter the life expectancy. Second, everyone improved, so much so that life expectancy for men in the second-lowest social class, IV, in the 1990s was greater than for the highest social class, I, in the 1970s. [End Page S12]

Changes in the Slope of the Gradient in Health

Whether or not it is true that the poor are always with us, there is no reason to believe that today's lowest social classes should not, in the future, have the good health enjoyed by today's highest classes. This could, in principle, happen regardless of whether today's more favored classes continue to improve. In other words, improvement could occur whether or not the magnitude of the social gradient in mortality remains the same.

It need not remain the same. In Britain and the United States, the magnitude of health inequalities according to social group has not been fixed. In the period 1972 to 1976, the gap in life expectancy between bottom and top social classes in Britain was 5.5 years. Twenty years later, this had increased to 9.5 years. The improvement in life expectancy over this period was 5.7 years for social class I and 1.7 years for social class V. Not only did women have longer life expectancy than men, but the women of lower socioeconomic position had greater improvement than their male counterparts. The widening of the life expectancy gap was, as a result, less in women than in men. When people started to concentrate on the problems posed by figures such as these, the various problems of interpretation caused vigorous debate (Illsley 1987; Wilkinson 1986). I will not go through all those arguments again. The point I wish to draw out of these figures is that if the life expectancy gap can increase, it can, in principle, decrease. If we think this is a problem worth tackling, the challenge is to understand the reasons for the social gradient in order to do something about it.

Group Differences and Individual Differences

The explanation of why one individual becomes ill and another does not may differ from the explanation for differences in illness between groups. For example, [End Page S13] if one village was served by polluted water and a neighboring village had clean water, and one was seeking to understand the high rate of diarrheal disease in the first village, one would not start by confining attention to that village and looking at the differences among individual children. The solution to the problem lies in the group differences between the villages.

It is not that individual differences are unimportant. First, they might tell us why some individuals are more susceptible to an exposure than others, for genetic or other reasons. Second, they may give us a guide to an exposure. In the 19th century, before the germ theory of disease had been elaborated, John Snow hypothesized that polluted water was the source of cholera because of what the water contained. Part of the evidence that clinched it for him was that on the same London street, some households were affected by cholera and others not. It turned out that there were two different water companies plying their trade in the same district. They took their water from different sources. Houses provided by one company had a greater risk of cholera than did houses provided by the other company. In this case, individual differences were a guide to the environmental exposure because there was variation in that exposure.

When we ask about differences in health between socioeconomic groups, differences between groups may be a guide to one set of social causes, while differences among individuals within groups will be a guide to a different set of causes.

Explanations for Health Inequalities

Many observers assume that social inequalities in health result from inequalities in health care. There is an enormous literature on the efficacy and effectiveness of health care, but surprisingly little on the contribution of health care to inequalities in health. In the United Kingdom, the government set up an Independent Inquiry into Inequalities in Health—the Acheson Inquiry (Acheson 1998)—conducted by a Scientific Advisory Group of which I was a member. Our conclusions were similar to those of Black 20 years earlier (Townsend, Davidson, and Whitehead 1992). The Black committee had been set up to answer the question of why, after 30 years of the National Health Service, there were persistent inequalities in health. Black's conclusion was that it was the wrong question to ask: inequalities in health were the result of inequalities in society.The contribution of medical care was primarily to treat illness when it occurred, not to prevent its occurrence.The Acheson committee noted evidence of inequalities in the provision of medical care within the National Health Service; these need to be addressed, but they are not the explanation of differences in incidence of disease.

Another response to the finding of a social gradient in health is that the causal direction is from health to social circumstances: sick people may be less likely to find themselves in higher social positions. In other words, it is illness that causes [End Page S14] low social position, not low social position that causes illness. This has some plausibility, but it turns out not to be the major factor accounting for the social gradients in health. Data from birth cohorts show that there is an effect of health in childhood on social position in adulthood, but this cannot account for the relation between social position and health in adulthood (Power, Manor, and Matthews 1999;Wadsworth 1996).

The Whitehall II study followed British civil servants over a 10-year period and considered two possibilities: changes in health were responsible for changes in social position; or changes in social position were responsible for changes in health. Using data carefully collected over the period, we were able to show that, overwhelmingly, it is the effect of social position on health that dominates (Chandola et al., in press).

An alternate model is that some antecedent factor (e.g., robustness) might determine both social position and health. There may indeed be such a factor. But if we take a broader perspective, it is hard to see how this external factor could account for the dramatic improvement in health in all social classes over the last century or the changing social class gap in life expectancy over the last 30 years. Whatever robustness factor there may be, we still need to ask how the social environment affects inequalities in health within and between populations.

Using Measures of Social Position to Understand Causes

By tradition, in the United Kingdom, social position has been measured by Registrar-General's social classes. In Britain, anything that refers to "class" seems to fit with the ethos. On closer inspection, however, it is not entirely clear what the Registrar-General's social class is measuring. It was conceived as a measure of general social standing and is based on assigning occupations to social classes (Stevenson 1928). Pragmatically, it has proved remarkably useful as a predictor of mortality, but what the relationship between social class and mortality tells us about causal processes is obscure. In the United States, it has been more common to use education, income, or some occupational measure, singly or in combination, as a measure of socioeconomic position. As with the Registrar-General's social classes, it is not entirely clear whether these classifications convey information about important causal pathways.

Bartley (1999) has argued that, in studying inequalities in health, we should use measures of social classification that are theoretically informed and likely to tell us something about the pathways involved in linking socioeconomic position to ill-health. I want to consider here three ways socioeconomic position could be linked to health: money, status, and power.

The Panel Study of Income Dynamics, quoted above, showed a continuous relation between income and mortality. Does this suggest that income is the most important factor influencing health inequalities? Not quite. Take the problem in two stages. First, when the investigators used both income and education as predictors of mortality, the relationship between income and mortality was [End Page S15] sharply reduced (McDonough et al. 1997). We need to interpret this with some care. Simply putting two correlated variables into a multivariate equation and asking which is left standing at the end may tell us more about precision of measurement than about which is most important. At the very least, if putting education into the model reduces the apparent association between income and mortality, it tells us that we need a different causal model. Education may be important because people with higher education have better life chances. This brings us to the second point. Income may be a predictor of ill health not because dollars in the pocket bring better health, but because income is a marker for position in society. In this interpretation, it is social position, not income, that is the important predictor.

This is illustrated by a comparison of life expectancy among black American men compared with men in Costa Rica. The GNP in Costa Rica is around $2,000 per person, and life expectancy for men is 74 years (World Bank 2000). Among black men in the United States, mean income is around $26,000 and life expectancy is 66 (Williams 1999). Adjusting for the fact that a dollar in Costa Rica buys more than a dollar in the United States would increase the GNP in Costa Rica to an amount closer to $6,600 than $2,800 per person. Nonetheless, despite having four times the income (purchasing power) of Costa Ricans, U.S. blacks still have eight years shorter life expectancy. I am not using this comparison to argue that poverty is not a problem for black men in the United States, but to point out that, as important as money might be, we need to go beyond absolute measures of income to understand the relationship between social position and health. This is especially the case when, as in the Whitehall studies, we are dealing with populations that are above the poverty line of absolute deprivation.

Relative versus Absolute Deprivation

Status is a relative concept: it implies a ranking. We can, however, move from a ranking system to one that involves quantification. If A is higher in the social hierarchy than B, and there were a quantitative measure of status, one could measure by how much B was lower than A. Pursuit of this quantitative approach raises the question of which is more important for health, relative or absolute social position. Does it matter by how much one person is better off than another, or is it the mere presence of status differentials that matter? A society without any status differentials is hard to find. Erdal and Whiten (1996) have suggested that if hunter-gatherers had little in the way of status distinctions, perhaps we evolved to be egalitarian. It would appear, however, that even in hunter-gatherer societies that share what little they have, there are status differences (Wright 2001). If, therefore, it is simply the presence of a hierarchy that matters for health, all societies would have the same social gradient in health. They do not. This suggests that it is relative position that is important, and we have to ask what it is about relative position that could translate into differential health risks and why the effect of status varies. [End Page S16]

Amartya Sen (1992) has developed the idea of measuring inequality in different spaces. This can help resolve the question of which is more important: absolute or relative differences. Sen suggests that relative deprivation in the space of incomes can yield absolute deprivation in the space of capabilities. This may sound like a linguistic maneuver, but I think it helps take us forward. The question is not how much material resource someone has, but what that level of material resources allows them to do. This may depend on the resources I have relative to those above me—a relative measure. But what I can do, my capability, is an absolute measure. This can be illustrated with the example I used above: U.S. blacks have high income on a world scale, but compared to U.S. whites they are, relatively, poorly off. What is it about being poor relative to the favored majority within the country that might be bad for health? The answer lies, presumably, in what their relative disadvantage in terms of income allows them to do. Relative inequality in income may correspond to absolute discrimination and social exclusion.

Social Position and Health:
Control and Social Participation

This brings us to consideration of what the relevant capabilities might be. I suggest there are two key ones: control and social participation. These have to be set in a context of how social position might be related to health.

Social position is related to the circumstances in which people live and work. This brings us to the third of the three ways that we classify social position: power, or how much control people have over their lives. One (but not the only) sphere in which this is important is that of work. In fact, the British government has introduced a new social classification that, like the old Registrar-General's, is based on occupation. In the new classification, the ONS Socio-Economic Classification, occupations are assigned to groups on the basis of something rather close to how much control people have at work.

A large body of literature links psychosocial work conditions to ill health, cardiovascular disease in particular (Schnall et al. 2000). One way this has been formulated is that jobs characterized by high psychological demands and low control put people at high risk for cardiovascular disease. A number of studies provide empirical support for this model; some show that low control, but not high demand, is related to risk of cardiovascular disease (Hemingway and Marmot 1999). In the Whitehall II study, low control was related to risk of heart disease, independent of other predictors (Bosma et al. 1997). Further, low control at work, together with smoking and other coronary risk factors, made a major contribution to explaining, statistically, the reasons for the higher incidence of coronary heart disease in lower employment grades (Marmot et al. 1997).

People may be deprived of control outside of work. In the Whitehall II study, we found that low control at home predicted the onset of depression. The association was particularly marked in low status women (Griffin et al. 2002). Likewise, our studies of health in Central and Eastern Europe have shown that low [End Page S17] control over life circumstances is related to increased risk of poor health (Bobak et al. 2000). Power, then, appears an important way that position in the social hierarchy is translated into greater risk of ill health.

A second way is differential degree of social participation. Let me illustrate with a quote from the father of modern economics, Adam Smith:

By necessaries I understand not only the commodities which are indispensably necessary for the support of life, but what ever the customs of the country renders it indecent for creditable people in the lowest order to be without. A linen shirt, for example, is, strictly speaking, not a necessary of life. The Greeks and Romans lived, I suppose, very comfortably though they had no linen. But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty which, it is presumed, nobody can well fall into without extreme bad conduct. Custom, in the same manner, has rendered leather shoes a necessary of life in England. The poorest creditable person of either sex would be ashamed to appear in public without them. (quoted in Sen 1999)

Smith eloquently describes how relative deprivation in one space translates into absolute deprivation in another. A linen shirt is deprivation only in a relative sense, but the want of a linen shirt when the custom of the day demands it means absolute deprivation in the space of shame. This lack translates into inability to participate fully in society. Does lack of full participation lead to ill health? One obvious way it might is through a lack of social ties. Socially isolated people have worse health than those who are more fully integrated into a pattern of social networks (Berkman 1995).

Data we collected from a study in Hungary provide a more direct test of the Adam Smith model of participation (Marmot and Bobak 2000). Participants were asked to list the consumer goods that their household contained. These were divided into basic, socially oriented, and luxury. The more goods the household contained, the better was participants' health. This was more true of luxury goods—cable television, video recorder, a personal computer, a country "dacha"—than of basic goods. The implication was that it was not just deprivation that was bad for health, but missing out on the luxuries that defined what it meant to participate fully in what society had to offer.

Upstream and Downstream:
Looking at the Social Environment

There is a tendency to treat socioeconomic position as if it were a characteristic of an individual, like age or gender. In a multivariate statistical model, these variables are entered and independent relations sought. This is too limited; we need to look both upstream and downstream. The meaning of a particular socioeconomic [End Page S18] position will depend on the society and the social environment in which an individual is located. Being a low-status clerical assistant will have a different meaning in a society where the state provides high levels of health and social services, amenities, and education than in a society where these are available patchily and to a higher level for those with the ability to pay.

I will not review here the evidence on the relationship between social environment and health (Marmot and Wilkinson 1999). However, I will point to two different lines of research on the social environment, the resolution of which may do much to increase understanding of how social processes can affect health. The first is the series of observations sparked by Wilkinson's (1996) finding of a relation between level of income inequality of a society and its overall health—high inequality predicts worse health. The second comes from Robert Putnam's work on social capital (Putnam 2000). Much of the criticism ofWilkinson's thesis revolves around whether income inequality is causal or whether it is a marker for other features of the social environment (Deaton 2002). What might these features be? Social capital is one answer. Putnam has raised the question of which comes first: income inequality or social capital. This is a fertile area for improved understanding, the results of which may be crucial for making good policy decisions.

The upstream focus looks at the social environment; a downstream focus examines the pathways through which position in the hierarchy can affect health and disease. At the bottom end of the social scale, factors associated with absolute deprivation—infections, inadequate food and shelter—are important. However, these are less likely to be crucial in explaining the social gradient in people above a threshold of absolute deprivation (Lynch et al. 2000; Marmot and Wilkinson 2001).

We have been considering two types of pathways by which social position may be translated into ill health: through health behaviors, and via psychosocial pathways that affect the neuroendocrine system more directly. There are a variety of plausible pathways by which psychosocial factors linked to social position could increase the risk of cardiovascular disease (Steptoe and Marmot 2002). McEwen (1998) has provided an organizing concept, the allostatic load, that brings these apparently diverse pathways together. The notion here is that an accumulation of stressors exert their effect. The body, in doing work to maintain equilibrium in the face of these cumulative burdens, shows biological deviations—allostatic load. Those with higher levels of allostatic load will have a breakdown in functioning and a higher risk of disease (Seeman et al. 2001).

Health behaviors are also important. In particular, smoking shows a clear socioeconomic gradient in the United Kingdom and the United States, although not in all countries (Fuhrer et al. 2002). Smoking is part of the explanation for the social gradient in mortality. In the first Whitehall study, a combination of smoking, high blood pressure, plasma cholesterol, body mass index, and sedentary lifestyle accounted for about a quarter of the social gradient in mortality (Marmot, Shipley, and Rose 1984; van Rossum et al. 2000). The important question [End Page S19] here is why there should be a social gradient in smoking. The smoking and health question is not answered only by showing the powerful link between smoking and disease. We need to understand the reasons for the persisting social gradient in smoking.

Early Origins

It is clear that we cannot simply take socioeconomic position of middle-aged people as if they had been randomly assigned to social strata as they crossed the threshold into adulthood. The experience of people from conception, through the life course, and into adulthood may be highly relevant to their health status in adulthood (Barker 1998; Kuh and Ben-Shlomo 1997; Kuh and Hardy 2002). Barker has been a leader in showing that experiences in utero and in the first year of life have an effect on risk of cardiovascular and other diseases later in life. In a recent study conducted in Finland, Barker, et al. (2001), showed an interaction between thinness at birth and social class later in life in increasing the risk of cardiovascular disease. I found a similar phenomenon—thinness at birth increased risk of cardiovascular disease in people who were of low social status in adulthood (Marmot 2001). Income was not a crucial determinant once other socioeconomic features were examined. Barker speculated that these data were consistent with an interaction between early exposures, possibly nutritional, and later psychosocial processes.

Words of caution are in order. One common measure that is used for social classification is education. People with less education often have worse health than those with more. It is tempting to believe that this is a direct causal process, and it may be, but I would offer two caveats. First, there is a close link between the level of social deprivation and children's performance in school (Blane, Brunner, and Wilkinson 1996). Education may be telling us as much, therefore, about the backgrounds from which children come as it does about their level of knowledge and skills. Second, education may be only an indirect cause of ill health (Singh-Manoux, Clarke, and Marmot 2002). Level of education is a crucial determinant of where people end up in adult life. The causal role of education may then be strongly affected by the social conditions affecting adults. If the conditions of people with low education are desperately bad, health will suffer more than if there is a more egalitarian distribution of resources, amenities, and opportunities. This is a likely explanation of why, in the United States, blacks with low education have worse health than whites with equivalent levels of education (Williams 1999).

Can Anything Be Done?

My teacher at Berkeley, Leonard Syme, fresh from a dismal experience attempting to change individual behavior, offered the opinion that it was too difficult to [End Page S20] change individuals—changing society is easier. That is all it takes to reduce inequalities in health—change society. Is he hopelessly utopian, misguidedly revolutionary? I think not. Robert Fogel (2000) points to the dramatic egalitarian achievements in U.S. society in the hundred years since about 1870. Some of the best evidence for this, suggests Fogel, are the narrowing gaps in life expectancy between rich and poor, white and black. Given that these gaps are no longer narrowing and may again be widening, there is a new challenge. In Britain, the Acheson Committee referred to above made 39 recommendations but highlighted three: all government policies should be assessed for their potential impact on health inequalities; priority should be given to women of childbearing age and families with young children; and the tax and benefit system should work to improve the living standards of those worse off. We live in the real world. Major social changes do not happen overnight. Nevertheless, we welcome the U.K. government's commitment to recognizing the problem of health inequalities and to putting policies in place to improve the situation. Avoidable health inequality is inefficient because it is a waste of valuable human resources. And it is unjust.

 



Sir Michael G. Marmot is Professor of Epidemiology and Public Health and Director of the International Centre for Health and Society at University College London. He is also Adjunct Professor of Health and Social Behavior at the Harvard School of Public Health. In addition to the Whitehall Study, Sir Michael coordinated two European Research networks and is currently the co-coordinator of the European Science Foundation network on inequalities in life expectancy.The author of over 400 scientific papers, Sir Michael was elected founding fellow of the Academy of Medical Science.

Head, Department of Epidemiology and Public Health, University College London, 1-19 Torrington Place, London WC1E 6BT, United Kingdom.

Email: m.marmot@ucl.ac.uk.

References

Acheson, D, chair. 1998. Independent inquiry into inequalities in health. London: Stationery Office.

Barker, D. J. P. 1998. Mothers, babies and health in later life. Edinburgh: Churchill Livingstone.

Barker, D. J. P., et al. 2001. Size at birth and resilience to effects of poor living conditions in adult life: Longitudinal study. BMJ 323:1273-76.

Bartley, M., et al. 1999. Understanding social variation in cardiovascular risk factors in women and men: The advantage of theoretically based measures. Soc. Sci. Med. 49: 831-45.

Berkman, L. F. 1995. The role of social relations in health promotion. Psychosom. Med. 57: 245-54.

Blane, D., E. J. Brunner, and R. G. Wilkinson. 1996. Health and social organization.London: Routledge.

Bobak, M., et al. 2000. Socioeconomic factors, material inequalities, and perceived control in self-rated health: Cross-sectional data from seven post-communist countries. Soc. Sci. Med. 51:1343-50.

Bosma, H., et al. 1997. Low job control and risk of coronary heart disease in the Whitehall II (prospective cohort) study. BMJ 314:558-65.

Chandola, T, M., et al. In press. Health selection in the Whitehall II study. Soc. Sci. Med.

Deaton, A. 2002. Policy implications of the health and wealth gradient. Health Aff. 21(2): 13-30.

Eckersley, R., J. Dixon, and B. Douglas, eds. 2001. The social origins of health and well-being. Cambridge: Cambridge Univ. Press.

Erdal, D., and A. Whiten. 1996. Egalitarianism and Machiavellian intelligence in human evolution. In Modelling the early human mind, ed. P. Mellars and K. Gibson. Cambridge: McDonald Cambridge. [End Page S21]

Fogel, R. W. 2000. The fourth great awakening and the future of egalitarianism. Chicago: Univ. of Chicago Press.

Fuhrer, R., et al. 2002. Socioeconomic position, health and possible explanations: A tale of two cohorts. Am. J. Public Health 92(8):1290-94.

Geronimus, A. T., et al. 1996. Excess mortality among blacks and whites in the United States. N. Engl. J. Med. 335:1552-58.

Griffin, J., et al. 2002. The importance of low control at work and home on depression and anxiety: Do these effects vary by gender and social class? Soc. Sci. Med. 54(5): 738- 98.

Hattersley, L. 1999. Trends in life expectancy by social class: An update. Health Statistics Q. (2):16-24.

Hemingway, H., and M. Marmot. 1999. Psychosocial factors in the aetiology and prognosis of coronary heart disease: Systematic review of prospective cohort studies. BMJ 318(7196):1460-67.

Illsley, R. 1987. Occupational class, selection and inequalities in health: Rejoinder to Richard Wilkinson's reply. Q. J. Soc. Affairs 3:213-23.

Kaplan, J. R., et al. 1987. Inhibition of coronary atherosclerosis by propranolol in behaviorally predisposed monkeys fed an atherogenic diet. Circulation 76(6):1364-72.

Kuh, D., and Y. Ben-Shlomo. 1997. A life course approach to chronic disease epidemiology. Oxford: Oxford Univ. Press.

Kuh, D., and R. Hardy. 2002. A lifecourse approach to women's health. Oxford: Oxford Univ. Press.

Kunst, A. E., and J. P. Mackenbach. 1994. The size of mortality differences associated with educational level in nine industrialized countries. Am. J. Public Health 84:932-37.

Lynch, J. W., et al. 2000. Income inequality and mortality: Importance to health of individual income, psychosocial environment, or material conditions. BMJ 320:1200- 1204.

Mackenbach, J. P., et al. 1997. Socioeconomic inequalities in morbidity and mortality in Western Europe:The EU Working Group on Socioeconomic Inequalities in Health. Lancet 349:1655-59.

Marmot, M. [G.] 2001. Editorial. Aetiology of coronary heart disease. BMJ 323:1261-62.

Marmot, M. [G.] 2002. The influence of income on health: Views of an epidemiologist. Health Aff. 21(2):31-46.

Marmot, M. [G.], and M. Bobak. 2000. International comparators and poverty and health in Europe. BMJ 321:1124-28.

Marmot, M. G., M. J. Shipley, and G. Rose. 1984. Inequalities in death: Specific explanations of a general pattern. Lancet 1:1003-6.

Marmot, M. G., and R. G. Wilkinson, eds. 1999. Social determinants of health. Oxford: Oxford Univ. Press.

Marmot, M. [G.], and R. G. Wilkinson. 2001. Psychosocial and material pathways in the relation between income and health: A response to Lynch et al. BMJ 322(7296): 1233-36.

Marmot, M. G., et al. 1978. Employment grade and coronary heart disease in British civil servants. J. Epidemiol. Community Health 32:244-49.

Marmot, M. G., et al. 1991. Health inequalities among British civil servants:The Whitehall II study. Lancet 337:1387-93.

Marmot, M. G., et al. 1997. Contribution of job control and other risk factors to social variations in coronary heart disease. Lancet 350:235-40. [End Page S22]

McDonough, P. G., et al. 1997. Income dynamics and adult mortality in the United States, 1972 through 1989. Am. J. Public Health 87:1476-83.

McEwen, B. S. 1998. Protective and damaging effects of stress mediators. N. Engl. J. Med. 338:171-79.

Power, C., O. Manor, and S. Matthews. 1999. The duration and timing of exposure: Effects of socioeconomic environment on adult health. Am. J. Public Health 89(7): 1059-65.

Putnam, R. D. 2000. Bowling alone: The collapse and revival of American community. New York: Simon and Schuster.

Rowntree, B. S. 1901. Poverty: A study of town life. In Poverty, inequality and health in Britain, 1800-2000: A reader, ed. G. Davey Smith, D. Dorling, and M. Shaw. Bristol: Policy Press, 2001.

Sapolsky, R. M. 1998. Why zebras don't get ulcers: An updated guide to stress, stress-related diseases, and coping. New York: W. H. Freeman.

Schnall, P., et al., eds. 2000. Occupational medicine: State of the art reviews, vol. 15(1). Philadelphia: Hanley and Belfus.

Seeman, T. E., et al. 2001. Allostatic load as a marker of cumulative biological risk: MacArthur studies of successful aging. Proc. Natl. Acad. Sci. USA 98(8):4770-75.

Sen, A. 1992. Inequalities reexamined. Oxford: Oxford Univ. Press.

Sen, A. 1999. Development as freedom. New York: Knopf.

Shively, C. A. 2000. Social status, stress and health in female monkeys. In The society and population health reader: A state and community perspective, ed. A. R. Tarlov and R. F. St. Peter. New York: New Press.

Shkolnikov,V. M., and G. A. Cornia. 2000. Population crisis in transitional Russia. In The mortality crisis in transitional economies, ed. G. A. Cornia and R. Paniccia. New York: Oxford Univ. Press.

Singh-Manoux, A., P. Clarke, and M. G. Marmot. 2002. Multiple measures of socio-economic position and psychosocial health: Proximal and distal effects. Int. J. Epidemiol. 31(6):1192-99.

Steptoe, A., and M. [G.] Marmot. 2002. The role of psychobiological pathways in socio-economic inequalities in cardiovascular disease risk. Eur. Heart J. 23:13-25.

Stevenson, T. H. C. 1928. The vital statistics of wealth and poverty. BMJ 1:354.

Townsend, P., N. Davidson, and M. Whitehead. 1992. Inequalities in health. The Black Report:The health divide. London: Penguin.

van Rossum, C., et al. 2000. Employment grade differences in cause specific mortality: 25 year follow-up of civil servants from the first Whitehall study. J. Epidemiol. Community Health 54:178-84.

Wadsworth, M. 1996. Family and education as determinants of health. In Health and social organization, ed. D. Blane, E. Brunner, and R. [G.] Wilkinson. London: Routledge.

Wilkinson, R. G. 1986. Socio-economic differences in mortality: Interpreting the data on their size and trends. In Class and health, ed. R. G. Wilkinson. London: Tavistock Publications.

Wilkinson, R. G. 1996. Unhealthy societies:The afflictions of inequality. London: Routledge.

Williams, D. R. 1999. Race, socioeconomic status and health:The added effects of racism and discrimination. Ann. NY Acad. Sci. 896:173-88.

World Bank. 2000. World development report 1999/2000. New York: Oxford Univ. Press.

Wright, R. 2001. Non zero. New York: Vintage.

Additional Information

ISSN
1529-8795
Print ISSN
0031-5982
Launched on MUSE
2003-07-23
Open Access
No
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.