A note on the biologic concept of race and its application in epidemiologic research

Volume 108

Number 3, Part 2 Summary of Workshop IV

be effective and well-tolerated medications, particu- black population because diuretic therapy is low in larly in blacks. The use of diuretics is associated cost, easily complied with (in once-a-day dosage), with changes in serum levels of potassium, glucose, and effective in lowering blood pressure (particular- uric acid, and lipids. It has been suggested that ly in patients with low-renin hypertension, who worsening of risk factors, induced by diuretic thera- compose the majority of black and elderly patients py, may be responsible for this apparent lack of with hypertension). benefit. This question is particularly relevant to the

A note on the biologic concept of race and its application in epidemiologic research

Use of the category of race in epidemiologic research presupposes scientific validity for a system that divides man into subspecies. Although the significance of race may be clear-cut in many practical situations, an adequate theoretical construct based on bidogic principles does not exist. Anthropologists have in large measure abandoned the biologic concept of race, and its persistent widespread use in epidemiology is a scientific anachronism. The assumption that race designates important genetic factors in a population is in most cases false. Racial definitions should be seen as primarily social in origin and should be clues to environmental-rather than genetic -causes of disease. An understanding of the social forces leading to racial differentials in health will give further direction to preventive campaigns. (AM HEART J 108:715, 1984).

Richard Cooper, M.D. Chicago, 121.

Racial differentials are a prominent feature of the epidemiology of mass disease in multiracial societies such as the United States. Although the concept of race may have a clear-cut practical meaning in many situations, its biologic significance is often ambigu- ous. Analytic thinking in epidemiologic research generally proceeds on the assumption that at least some important proportion of the racial differential may be explained by population genetics. From that perspective it is disconcerting to recognize that anthropologists, in whose discipline the concept of race most naturally belongs, have arrived at the conclusion that human races in fact do not exist.‘-lo If it is true that “it is an error to believe that races are things . . . whose separate evolutionary development may be traced,“’ then to what end is it justifiable to use the category of race in epidemiolog-

From the Division of Cardiology, Cook County Hospital.

Reprint requests: Richard Cooper, M.D., Division of Cardiology, Depart- ment of Medicine, Cook County Hospital, Chicago, IL 60612.

ic research? In an effort to address that question, this paper has three purposes: (1) to argue against the biologic concept of race, (2) to examine the shortcomings of the race concept in public health, and (3) to emphasize the social origins of racial differences.

RACE AND BIOLOGY: A HISTORICAL VIEW

The concept of race as applied to man was first introduced into the literature of biology by Buffon in 1749.l It was explicitly regarded as an arbitrary classification, serving only as a convenient label, not a definable scientific entity. From the point of view of the biologist, the concept of race is an attempt to extend the taxonomic classification below the level of specie& It shares the inherent weaknesses of the Linnaean system while introducing unique problems of its own.

Taxonomy is useful for two separate but interre- lated reasons: it helps us understand evolution and provides a framework for examining present varia-

715

716 Cooper September, 1984

American Heart Journal

tion in plants and animals. Yet no zoologic classification is abso1ute.‘-g As Watt7 points out: “From an evolutionary point of view taxonomy is arbitrary and static. Forcing the results of the dynamic and multi- plex process of evolution into simple two-dimension- al sets of pigeonholes cannot be done without losing information.” For most biologists, use of race results in a net loss of information.

If consensus is any measure, defining the races of man has never been successfully accomplished.‘, lo-l3 It is important to emphasize that what is at stake here is not the possibility of designating groups of people that differ in some way. Human variation is self-evident; the existence of definable groups, or races, is not. The problem is twofold: Can we adequately describe human variability through the concept of race? Does this process help us understand the meaning of diversity within our species? Watt7 described this twofold requirement clearly:

Human populations can be classified in ways that are broadly consistent with genetic affinities, just as they can be characterized by lists of “typical” traits or genes. However, they cannot be understood in those terms. Races as genetic groups can only really be understood in terms of the process by which they originated.

The scientific purpose of the concept of race is therefore to give meaning to human variation, not merely to point out one or another aspect of that variation.14 It goes without saying that race must at the same time delineate important and consistent genetic differences; it implies that a “package” of different genes, all of which ideally should be known, exists between groups. On both counts, historical efforts to use the race concept scientifical- ly have ended in failure. Why?

First, geographic variation in gene frequency is mainly quantitative, or clinical, in nature. From Europe eastward across the Asian land mass, populations meld into one another. Circular phenomena also occur. That is, gene frequencies moving in one direction gradually change until they arrive back at the point of origin, where a sharp break may be observed. The problem of indistinct borders for racial groups might be solved by accepting several large races, such as two to five. However, the number of nonclassifiable groups will be great. On every continent there are isolated subpopulations that cannot be accommodated, such as the pygmies, Kalahari bushmen, Basques, and Lapps. Mass migration, old as well as recent, further blurs dis- tinctions. The other extreme would be to reduce each population, or “ethnic group,” to a separate category. This is essentially the process of naming,

not classifying, and is the approach taken by Dobz- hansky et all2

Second, human variation is primarily discordant, rather than concordant. That is, although two groups may be similar in skin color, they differ in other important features, such as height, blood type, or facial features. To classify on the basis of skin color arbitrarily assigns primary importance to that characteristic and forces all others to be ignored. Is there any reason to believe that variations in skin color subsume all, or any, biologically important human variation? The Masai, pygmies of the Afri- can rain forest, inhabitants of southern India, Aus- tralian aborigines, and natives of the Amazon delta are all dark skinned: Are they members of the same race? Skin color in these groups probably reflects a common history of exposure to ultraviolet radiation, not a common racial origin. It is characteristic of this problem of race that although skin color is the most important single trait used in classification, it is the one least well suited to the purpose.

The solution to both these problems-clinal and discordant variation-lies in the use of multivariate analysis. No discrete “package” of gene differences has ever been described between two races-only relative frequencies of one or another trait. Simulta- neous treatment of many continuous variables, not dichotomous categories, is the only appropriate method for studying human variation.14 Only in this manner can we begin to manage the data that represent the entire range of variation across the species. This understanding is damaged, not assisted by, creation of arbitrary cutoff points.

Yet one might ask, is it not common practice in medicine and biology to divide a continuous distribution into discrete categories? Hemoglobin, blood pressure, and body mass index are all skewed normal distributions; yet we accept the validity of the concepts of anemia, hypertension, and obesity. For pathologic conditions (such as hypertension), we have objective criteria (association with morbid sequelae), whereas for racial traits no such external reference system exists. The unusual case, such as sickle cell anemia, can easily be handled on its own merit without reference to a malfitting racial category. Just as single-gene abnormalities are recognized to be different from most diseases that require complicated gene-environment interaction, consistent difference in one gene between populations is rare and cannot express anything more than a minuscule proportion of the total potential variation.

A single trait, such as skin color, is not an adequate basis for characterizing human diversity.

Volume 103

Number 3, Part 2 Biologic concept of race and epidemiologic research 717

Variation in a single gene, such as that for sickle cell anemia, does not imply that populations which vary in that gene will vary in important ways for any other health condition. Transferring racial susceptibility from sickle cell anemia to cancer, for example, is an unjustified assumption. Discordant variation implies that unless two conditions can be shown to be linked, they should be assumed to be independent. Most public health analysis based on racial categories adopts this false analogy-that a “package” of genetic differences exists-and is thus fatal- ly flawed.

Even if one could classify groups of people useful- ly by race, little would have been accomplished from the point of view of the anthropologist. As noted earlier, “Races as genetic groups can only really be understood in terms of the process by which they originated.“7 This point is central to my argument and deserves to be emphasized. It has been stated in a similar way by authors contributing to Montague’s work’:

Race has been considered as a concept of the Linnaen system of classification within which it is applied to groups of populations within a species. To apply a concept of the Linnaen system to a group of populations implies something about their evolutionary history.

Race is a useful concept only if one is concerned with the kind of anatomical, genetic, and structural differences which were in time past important in the origin of races., . . We may look backward to the explanation on the differences between people-then the concept of race is useful, but it is useful under no other circumstances, as far as I can see.

Dobzhansky et all2 also state:

Race differences are then adaptations to diverse environments in different parts of the distribution area of the species. , . . Numerous racial differences between populations have originated and are being maintained by natural selection.

Human variation is of interest to study primarily because adaptive value is assumed for the observed traits. Thus Coon13 argued that the five races of man represent separate origins of the human species, with independent evolutionary histories. This position has been widely discredited.’ The search for “pure” races is a false hope.lpg We do not have a history of man’s wandering, nor will one ever be forthcoming. “It is an error to believe that races are things . . . whose separate evolutionary development may be traced.“*

Although this limitation may be fatal for the anthropologist, it could still be argued that the epidemiologist needs a system that gives broad

categories of gene distribution7 The epidemiologist presumably is not very much interested in the origin of genes. He or she wants to know only their present distribution and does not need to know anything about the “process by which they originated” to understand “races as genetic groups.” Again, false analogy applies. Race is potentially useful only because it might tell us something about evolution. Man’s health problems today are not in any major way a result of evolutionary history in the physical sense; it is cultural history that has produced the human disease burden. The genetic and evolutionary baggage purportedly contained in the concept of race is therefore of no interest to the epidemiologist either, except in the rarest of cases. The rare cases- for example, sickle cell anemia and skin cancer in white Australians-in no way justify the dominant role given to genetic explanations of between-race disease variation.

Despite these criticisms, the concept of race is still used routinely today by many biologists and anthropologists.1’-13 Is there any defensible application of the concept? Among the most prominent supporters of the race concept has been Theodosius Dobzhan- sky, who wrote in 1962 that “anthropologists are confronted with a diversity of human be- ings. . . . Race is the subject of scientific study and analysis simply because it is a fact of nature.“” While acknowledging the views of critics, he maintained this position in the most recent edition of his influential textbook on evolution.12 It is argued that “races, species, genera, and other categories are needed above all for the pragmatic purpose of communication.“*2 It must be questioned, however, whether the concept of race that is being put forward bears much in common with its historical use. While concluding that “data gathered by anthropologists studying racial variation in man- kind should not be underrated,“12 Dobzhansky et al, offer a very limited rationale for continued use of the category. Race is returned to the status proposed by Buffon, a label of pure convenience. The number of races is to vary widely depending on the question being studied. Thus there may be “European” and “African” races on one occasion and, on another, “Nordics, Alpines, Mediterraneans, Armenoids, etc.“12 If one accepts this definition, it is hard to disagree with an earlier conclusion by these same authors: “Saying that a population is of such-and- such geographic origin is good enough.“12 An ever- changing label of convenience, based on “some criteria” and used in an ad hoc fashion to refer to different groups of traits, is what is proposed.

To name various population groups for the sake of

718 Cooper September, 1984


comparing gene frequencies is not the same as creating a classification of Homo sapiens based on race. The “naming” process in effect denies the assumption that races have a structural relationship to each other, because of evolution, and that each category has systematic differences in a block, or package, of genes. In my view the use of the racial category implies that the groups differ in characteristics other than the primary one being compared. Otherwise, why not specify precisely the population and the trait? As we have argued, it is this assumption of consistent difference in many traits that is the essential failure of the race concept.

One final question must be addressed: How important is the potential racial variation? Glass15 has suggested that the number of gene differences between a West African and a native of Denmark, for example, may be very small indeed. He estimates that the “white race” differs characteristically-that is, on a population basis-by no more than six pairs of genes from the black.15 Although others consider this estimate low, it serves to place this question in the proper context.s Typical variation within a population is orders of magnitude larger. Given the overall magnitude of potential variation, two per- sons of the same race could vary by thousands of genes. Lewontin l6 has estimated that diversity between individuals in a population accounts for 85% of the total species variation, diversity within race accounts for 8.3%) and between-race diversity contributes only 6.3 % . Although consistent differences, such as in skin color, may tell us more about adaptive change than random, within-population variation, they will be overwhelmed by the weight of similarity when man is examined as a species. More recent consideration of “Haldane’s dilemma” sug- gests that systematic variation between populations may be primarily due to genetic drift, rather than adaptation, thereby eliminating even the possible significance of racial differences from an evolutionary point of view. l2 At any rate there is no evidence, as noted earlier, that racial variation in disease is, in any but the most trivial way, genetic in origin. Different human populations exposed to similar environments are much more alike than different in their disease rates.

RACE CONCEPT AND PUBLIC HEALTH

The race concept is prominent in public health and medicine. The classic anthropologic concept of race-in which the race label identifies a set of important genetic differences, however ill- defined-is used almost exclusively. The most obscure phenotypic variations, based on observa-

tions from small numbers of individuals with totally unknown genetic resemblance, are ascribed to “racial differences.“17 True genetic markers are rarely used, and the relationship between genotype and phenotype is almost universally ignored.

A common sequence of progress in understanding the cause of disease consists of replacing a genetic theory with an environmental explanation. Perhaps the most dramatic example is the story of a disorder known as kuru.7 In 1959 a chronic progressive neu- rologic disorder was described among the Fore people, a remote tribe in New Guinea. It occurred most commonly among women and young children, accounting for 96% of deaths among adult females in some villages. There was at first no evidence of contagiousness, since visitors were not affected and some Fore developed the disease years after they had left the endemic area. It clearly ran in families and appeared to be sex linked. On that basis it was assumed to be genetically determined, and a single- gene hypothesis was proposed.7 Laws were considered to sterilize affected families. Further examina- tion of local customs, however, demonstrated an alternative explanation. Burial rites for male rela- tives consisted of close contact with the corpse, with all tissues being cooked and eaten by women and young children. Subsequently of course, a filterable viruslike particle was identified and a “slow virus” disease described. Thus a disease that was thought to be caused by a single gene and to be restricted to one racial group was finally recognized to be the result of an infection. Although both racial and sex linked, the determinants were purely social, not genetic. Kuru has now been eliminated from the Fore.

Racial differentials in virtually all diseases have been attributed to genetic factors. Common infec- tious diseases are no exception.18-21 Classic studies on rheumatic fever in the last century from England clearly demonstrated the risk associated with pover- ty and crowding.ls On the basis of a careful review of these data, NewsholmezO concluded in 1895 that “the influence of heredity has been exaggerated” and “as to race there is no evidence.” Nonetheless, in the United States rheumatic fever was a disease to which the Irish were considered racially susceptible, and it was linked to genes for red hair. As the Irish became more prosperous, the disease disappeared, only to recur in black migrants from the rural South to urban centers.2o The incidence has now markedly decreased in that racial group as well. Today rheumatic fever in the United States is seen primarily among Latin American immigrants.

What is the central meaning of the race concept as

Volume 108


it applies to public health? In public health and medical work, race is virtually never defined; that is, specific gene frequencies are rarely measured and the criteria for assignment to one or another race are not stated. One is even hard-pressed to find a classic or typical formulation of the race concept in medicine.

Recently a group of anthropologists addressed the race concept in public health directly, and their text serves as a useful reference point in this discussion. Rothschild: the editor states the rationale for the book in the preface:

Comparative ethnogenetics, a fresh field in which ques- tions may be cultivated anew and unexpected insights harvested, may delineate some of the factors of evolutionary change. More importantly, knowledge of disease distribution . . . may. . . help to distinguish between the roles of exogenous and endogenous factors in causes and patho- genesis of these diseases.

In the first chapter, entitled “The Biological Race Concept and Diseases of Modern Man,” Watt’ provides the theoretical framework for “comparative ethnogenetics.” After reviewing the history of race, she takes the majority view of anthropologists: “Racial classifications by themselves have little explanatory value. . . . Physical anthropologists have abandoned race as a subject for research. . . . Race is no longer either a starting point for research on human evolution or an end.” (One wonders whether the editor read that last statement before he wrote that “comparative ethnogenetics . . . may delineate some of the factors of evolutionary change.“) Although an intellectual corpse in anthropology, however, race is said by Watt7 to be vital for epidemiology: “When diseases are being related to genetic factors, . . . biological race is an easily applied general indicator of the genetic makeup of populations and individuals.” More explicitly, the paradigm, says Watt,7 goes like this: “The race concept is a useful device where one may compare disease rates of genetically similar populations living in different environments and genetically different populations living in similar environments.”

Is it true that biologic race is helpful in explaining disease variation? What are these genetic differences and similarities that are being identified? We almost never know. Are they important for the disease in question? Specific estimates are not available. In fact, infinitely more is to be learned by taking precisely the opposite approach from the one suggested by the ethnogeneticists. All humans, in terms of their susceptibility to all but the rarest of diseases, are genetically similar. Systematic varia-

tion in susceptibility has not been shown to fall along racial lines for any common disease. (Hyper- tension might be considered the most important single exception to this proposition and will be discussed subsequently.) Racial differences reflect different social environments, not different genes, even where two groups live side by side, as do blacks and whites in the United States. Race does not mark in any important way for genetic traits; rather, it demonstrates beyond question the paramount role of the social causes. We have much more to learn from that paradigm, rather than the one offered by ethnogenetics.

Blacks in the United States have age- and sex- specific mortality ratios that are 25% to 300 % higher than those of whites. These ratios are likewise highly mobile over time. How can they be genetic? Age-adjusted death rates for blacks were 37 % higher than for whites in the United States in 1977. The most common fatal illness for which we have a clear-cut racial-genetic explanation is sickle cell disease. In 1977 there were 30,000 excess deaths among blacks compared with whites; 277 deaths among blacks were coded to hemoglobinopathies, or 0.3% of the total excess. We must look for the explanation of the remaining excess mortality primarily in social causes.

ORIGINS OF RACE

If, as we have argued, the concept of race has no legitimate scientific claim to existence, why has it remained such a vital part of our cultural experience? Unlike outmoded concepts of evolution or the solar system, the concept of race is firmly held by the vast majority of people in our society, including, no doubt, most scientists; by the same token, the race concept retains a prominent role in the explanation of disease variation. Why is it so resilient? Why is it constantly being reincarnated in “new disciplines,” such as comparative ethnogenetics and sociobiology? Clearly, the history of the race concept transcends any strictly scientific analysis. It must fulfill an important social need.

What are the historical origins of racism and the concept of race? The cultural anthropologists have responded to this question which so compellingly emerges from a critique of the biologic literature. A very brief summary of their explanation is presented here. Boyd argues that although early societies had many forms of class and caste, “the concept of race, as an immutable physical entity as we now know it, probably did not exist.“g Montag@ has been one of the more prominent exponents of the idea that the emergence of the African slave trade, the key to

720 Cooper

Table I. Ratio of nonwhite to white median income, would simply not have been possible without forced United States, 19451977 servitude.

Year

Nonwhite

families

Black

families

Nonwhite

Males Females

1945 0.56

1946 0.59

1947 0.51

1948 0.53 1949 0.51

1950 0.54 1951 0.53 1952 0.57

1953 0.56

1954 0.56 1955 0.55

1956 0.53

1957 0.54

1958 0.51

1959 0.52

1960 0.55

1961 0.53

1962 0.53 1963 0.53

1964 0.56

1965 0.55 1966 0.60

1967 0.62

1968 0.63 1969 0.63

1970 0.64

1971 0.63 1972 0.62

1973 0.60

1974 0.64 1975 0.65 1976 0.63

1977 0.61

0.54 0.54

0.58

0.59 0.60

0.61

0.61

0.60 0.59

0.58

0.60 0.61

0.59

0.57

n.a.

0.61

0.54 0.54

0.49

0.54 0.55

0.55

0.55 0.50

0.53

0.52

0.53

0.50 0.47

0.53

0.52 0.49

0.52

0.57 0.54

0.55

0.59 0.61

0.59

0.60

0.61 0.62

0.63

0.63

0.63 0.63

0.61

n.a.

n.a.

::i

0.51

0.49 0.46

i.z 0.55

0.54

0.58

0.58 0.59

0.62

0.70

0.67 0.67

0.67

0.70 0.73

0.76

0.78 0.79

0.85

0.92

0.90 0.95

0.93

0.92 0.92

0.95

0.88

~.a. = Not available.

economic development in the seventeeth and eigh- teenth centuries, led to the invention of the modern concept of race. Blacks had to be branded as inferior in order to justify this barbaric practice. Although the slave trade was crucial in the origin of racism, a second development of this historical era had a more enduring effect. It was in America that the impact of slavery was felt most directly. The long-term needs of a growing multiracial society demanded the invention of institutionalized racism.

The first Africans in Virginia, most historians agree, were classified as indentured servants. The status of “slave” simply did not yet exist. The low condition to which indentured servants were subjected signalled a form of labor organization unique to the New World. It was not possible to observe the European forms of labor organization and relations in a context where land was widely available and opportunities generally open. The large fortunes to be made from the cultivation of tobacco

September, 1994


It was into this world that Africans were brought in 1619. At first they were treated little differently from indentured servants, except perhaps in the length of their service. Details are obscure, but it appears that while some Africans served their entire lives, others earned freedom, at least in the period before 1660. Only after 1650 did laws and social practices distinguishing black workers from whites begin to appear in Virginia.*

With the arrival of greater numbers of white workers and the survival of many beyond the period of servitude, class conflict began to develop in the young colonies. In 1675, Nathaniel Bacon led a rebellion of poor and restless whites aimed against the Indians, in direct violation of the orders of the colonial governor. 23 The height of the rebellion was a march on Jamestown.23 In the words of one histori- an, “For those with eyes to see, there was an obvious lesson in the rebellion. Resentment of an alien race might be more powerful than resentment of an upper c1ass.“24 Slavery and its identification exclusively with people of color thus served two purposes: it provided both a tractable work force for the plantation economy and a unifying impetus to white racism.

Racial divisions between blacks and poor whites were carefully cultivated .25 Black-white marriages were outlawed and other social interactions were strictly limited. It is of interest that the most severe punishments for crossing the color line were reserved for whites, both free and indentured.23 As Jordan has pointed out, if Englishmen held blacks in such low respect, why was it necessary to proscribe social interaction?26 In fact, these laws were passed to prevent further racial mixing, not just to formal- ize existing social practices.

In addition to its political role, an economic function has been ascribed to racial divisions. Econ- omists have been interested primarily in the “problem” of racial discrimination in the labor market. Conservative economists, such as Friedman2’ tend to dismiss the issue. Other economists have argued that lower wages for blacks reflect economic loss taken by employers who must hire them. Supposed- ly lower productivity, which results from cultural inferiority, as well as dislike by both the employers and white co-workers for close contact with blacks, make it less desirable to hire blacks.% The conven- tional economic approach argues that racism is counterproductive in the economic sphere. Employ- ers are said to lose profits, whereas white workers may gain by preferential access to better jobs.

*From The roots of racism. Progressive Labor Magazine 15(3):1-16, 1982 (Brooklyn, NY).

Volume 108


Conservative economists further argue that the free market will eventually eliminate racial discrimination and hire workers based only on qual- ifications.

Another view has recently emerged.23y 2g-31 Racial discrimination is seen as a means of reducing labor costs by forcing minorities to take otherwise unac- ceptable jobs and by weakening the solidarity of the labor movement. Recent work by ReichN and others provides an important test of this theory. Reich bases his analysis on economic data from 46 metro- politan areas in the United States in 1960 and 1970. A measure of inequality, the ratio of black to white median income, is taken as an estimate of racism. After adjustment for a number of social factors (region, educational level, percent employment in industry, etc.), Reich demonstrates that as black income goes down 10%) white income at the median falls by 1% . Likewise, the existence of racial differentials is not dysfunctional for the system; rather, it boosts profits. An additional important finding of this study, substantiated by several other economists, is the persistence over time of racial differences in income: “The median income of black families in 1978 remained at only 57 percent of that of white families. That is, the median income of blacks was at approximately the same relative level in the early 1950s. 3o A modest improvement in relative income occurred over the 1960s; beginning at 0.54 in 1964, the black/white ratio rose to a peak of 0.61 in 1969 and fell to 0.57 by 1977.30 His- torical trends in relative income are presented in Table I.

In 1981 the median income of black families was 0.56; at the same time the education gap had narrowed to half a year.32 As noted in a recent economic report, “The income gap between blacks and whites is less related to education that to job opportunities open to blacks.“33 Approximately 5 billion dollars more would have been paid to black wage earners over the previous year if no white- black differential existed.” Clearly, racial divisions have economic value; this must in part explain why they have persisted.

SOCIAL CONTEXT OF RACE

If race is a category more defined by social than genetic characteristics, it should be possible to describe social variables that help explain differences in health status. It might even be argued that race assumes its functional role by providing the basis for assigning members of a racial group to a particular class status. Although it is true that interaction-i.e., the effect of race on class standing-may be the major effect, the relationship can-

not be understood on that basis alone, for at least two reasons. First, although socioeconomic variables may provide a partial explanation of mechanisms, they do not resolve the issue of causality. Second, race often makes an important impact on social status after equating income and education to the majority group.

Explaining racial differentials by education, income, etc., could in a casual sense be considered “overcontrol”: race is not confounded by the other variables; it is antecedent to them. It is race that influences class standing. Furthermore, the “unad- justed” difference is the disease burden actually borne by blacks, and it is the “raw” facts that must be explained-and changed. An even more simplis- tic analytic error is sometimes made: biologic variables, such as obesity, blood pressure (BP), smoking rates, etc., are used to demonstrate why blacks have a health disadvantage. Clearly these are attributes, not causes. Ascribing inferior health status to exposure to a more dangerous social environment accepts as given precisely the thing to be explained.

An interesting exercise in examining the significance of socioeconomic status (SES) in racial differentials was recently undertaken in relation to homicide.35 Homicide rates for blacks in the United States are seven times higher for males and five times higher for females. The social origins of this cause of death need not be elaborated. Based on the experience in Atlanta, Georgia, in 1971, the question was addressed, “Do blacks have a specific ‘culture of violence’?” That is, for a similar set of social conditions, are rates higher among blacks than whites? After control for education, a sizable difference remained.35 It was noted, however, that with identical education, blacks receive lower wages, and the effect of unemployment will also be hidden. Educa- tion and other measures of SES are not comparable across races. Substituting income further narrowed the gap, although it was not closed. Because of segregation, black purchasing power is also well below that of whites; for example, equal housing in a black neighborhood costs up to 30% more. The ability to buy uncrowded housing was then used to account for part of this social inequality. Using occupants per room as a direct estimate of purchasing power, the racial difference in domestic homicide rates disappeared altogether. In this analysis, of course, there was no absolute standard for comparing SES, and any variable chosen was necessarily arbitrary to some extent. Stepwise elimination of racial differences by more sensitive control of class standing, however, is an important theoretical consideration in research into potential environmental sources of health differentials.

722 Cooper

An important example of this problem exists in the field of cardiovascular research. Hypertension represents the key potential exception to the proposition that racial differentials in common disease are social in origin. Although it is widely assumed that blacks are genetically predisposed to hypertension,36 an environmental hypothesis is equally tenable. In the Hypertension Detection and Follow-up Program data, BP differences between black and white females disappear with control for education and obesity.37 The same is not true for males (black males being no more obese than white), and yet a clear social class gradient-as estimated by education-exists, which will sizably reduce the black- white gap for males as we11.38 Rates of hypertension for blacks with a college education were similar to those for whites who did not finish high school.3s Earning capacity of black college graduates is almost identical to that of white high school graduates32; even use of more appropriate measures of income and purchasing power might eliminate these BP differences. Other unmeasured aspects of the social significance of race could also be playing a role. A more sophisticated approach to explaining racial differences in hypertension, as was attempted in relation to homicide, would be a useful test of the environmental hypothesis. In the absence of this test, it is distressing that genetic explanations have gained such widespread acceptance.

CONCLUSIONS

An explicit discussion of coronary heart disease (CHD) among blacks has not been attempted in this paper. It is obvious that cigarette use and nutritional factors, such as hypercholesterolemia and obesity, are determined primarily, if not exclusively, by social conditions. It is not the purpose of this paper, however, simply to reemphasize the point that social factors should be given primary importance. What is in need of clarification is the meaning of the concept of race.

The argument put forward here leads to three major conclusions: (1) In the biologic sense there are no such things as races. Human variation does not occur in discrete packages. The appearance of a highly consistent pattern of differential mortality between races can be ascribed only to environmental (i.e., social), not genetic, factors. (2) The concept of race itself is a social category. Whether it be Catholic in Ulster, Jew in Germany, Tamil in Sri Lanka, or black in the United States, the definition of a population subgroup is a result of economic and historical, not evolutionary, developments. (3)

September, 1994


Health status of racial groups should be viewed within this context.

Blacks in the United States have a historically determined structural relationship to the social system. They do not represent a homogeneously dis- tributed subsegment of the population, nor is there any indication that they are becoming so; black/ white income ratios have not changed appreciably during this century. For all intents and purposes, black people in this society are imprisoned by in- stitutional racism; this is the attribute of blackness which at bottom determines their health status. Although the character of the health disadvantage may evolve, for example, as a result of migration from the rural South to the urban North, the disadvantage itself is not likely to diminish until the intensity of racial discrimination is successfully reduced. The epidemiology of CHD among blacks in the United States has likewise been determined by these social conditions. Greater cigarette use, relative exclusion from preventive campaigns, bad nutrition, excess hypertension, and obesity are allimportant attributes of the contemporary experience of black Americans. Higher rates of CHD are to be anticipated.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

a.

9.

10.

11.

12.

Montagu A, editor: The concept of race. Toronto, 1964, Collier Macmillan, Canada, Inc. Dobzhansky T: Genetic diversity and human equality. New York, 1973, Basic Books, Inc. Boyd WC: Genetics and the human race. Science 140:1057, 1963. Woodward V: Heredity and human society. Minneapolis, 1981, Burgess Publishing Co. Chase A: The legacy of Malthus: The social costs of the new scientific racism. Champaign, Ill, 1980, University of Illinois Press. Gossett TF: Race: The history of an idea in America. New York, 1965, Schocken Books, Inc. Watt ES: The biological race concept and the diseases of modern man. In Rothschild HR, editor: Biocultural aspects of disease. New York, 1981, Academic Press, Inc, pp 3-22. Ehrlich PR, Holm RW: A biological view of race. In Montagu A, editor: The concept of race, Toronto, 1964, Collier Macmil- lan, Canada, Inc, pp 153-197. Boyd WC: Genetics and the races of man. Boston, 1950, Little. Brown & Co, DD 186-209. Gould SJ: Ever since- Darwin: Reflections in natural history. New York, 1979, WW Norton Co. Dobzhansky T: Comment on Livingstone. Curr Anthropol 3:279, 1962. Dobzhansky T, Ayala FJ, Stebbins GL, Valentine JW: Evolu- tion. San Francisco, 1977, WH Freeman and Co.

13. Coon CS: The origin of races. New York, 1962, Alfred A Knopf, Inc.

14. Barnicot NA: Taxonomy and variation in modern man. In Montagu A, editor: The concept of race. Toronto, 1954, Collier Macmillan, Canada, Inc. DD 180-227.

15. Glass B: Genes and the man. New York, 1943, Columbia University Press.

16. Lewontin RC: Apportionment of human diversity. Evol Biol 6:381, 1972.

Volume 108

Number 3, Part 2 Biologic concept of race and epidemiologic research

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

Dunn FG, Oigman W, Sungaard-Riise K, Messerli PH, Ventura H, Reisin E, Frohlic ED. Racial differences in cardiac adaptation to essential hypertension determined by echocardiographic indexes. J Am Co11 Cardiol 1:1348, 1983. Paul JR: The epidemiology of rheumatic fever. New York, 1957, American Heart Association. Stollerman GH: Rheumatic fever and streptococcal infection. New York, 1975, Grune & Stratton, Inc. Newsholme A: Milroy lectures: Rheumatic fever. Lancet 1:589,657, 1895. Wintrobe S, et al: Harrison’s principles of internal medicine. New York. 1970. McGraw-Hill Book Co. D 865. Montagu .& Man’s most dangerous my&: &The fallacy of race. New York, 1945, Columbia University Press. The roots of racism. Progressive Labor Magazine 15(3):1, 1982 (Brooklyn, NY). Morgan E: American slavery, American freedom. New York, 1975, WW Norton & Co, Inc. Bennet L: Before the Mayflower: A history of the Negro in America, 1619-1964. Baltimore, 1966, Penguin Books. Jordan W: White over black: American attitudes toward the Negro, 1550-1812. Baltimore, 1968, Penquin Books. Freidman M: Capitalism and freedom. Chicago, 1962, The University of Chicago Press. Becker G: The economics of discrimination. Chicago, 1957, The University of Chicago Press.

29. Cherry R: Economic theories of racism. In Gordon DM, editor: Problems in political economy: An urban perspective. Lexington, Mass, 1978, DC Heath & Co, pp 170-182.

30. Reich M: Racial inequality: A political-economic analysis. Princeton, NJ, 1957, Princeton University Press.

31. Tabb W: Capitalism, colonialism, and racism. Review of Radical Political Economy. 3:90, 1971.

32. Gap in income between the races as large as in 1960, study finds. New York Times, July 18, 1983, p 1.

33. Dream deferred: The economic status of black Americans: A working paper. Washington, DC, 1983, Center for the Study of Social Policy.

34. News release. Washington, DC, August 1, 1983, U.S. Depart- ment of Labor, Bureau of Labor Statistics.

35. Centerwall BS: Race, socioeconomic status, and domestic homicide: Atlanta, 1971-1972. Paper presented at the 1982 Society for Epidemiologic Research conference, Cincinnati, June, 1982.

36. Genetics, environment, and hypertension. Lancet 1:681, 1983.

37. Langford HG: Is blood pressure different in black people? Postgrad Med J 57:749, 1981.

38. Hypertension Detection and Follow-up Program Cooperative Group: Race, education and prevalence of hypertension. Am J Epidemiol 106:351, 1977.

Working Conference on Coronary Heart Disease in Black Populations

Elijah Saunders, M.D., chairperson Provident Hospital, Inc. Baltimore, Maryland

Katrina Johnson, Ph.D. NHLBI/DHVD/BMB Bethesda, Md.

CONVENORS Gerald Payne, M.D. NHLBI/DHVD/PCB Bethesda, Md.

Steven Clyburn, M.P.A. NHLBI/DHVD/PCB Bethesda, Md.

Lucile Adams, Ph.D. University of Pittsburgh School of Public Health Pittsburgh, Pa.

PARTICIPANTS Luther T. Clark, M.D. Chief, Cardiology Woodhull Medical & Mental Health Center Brooklyn, N.Y.

Gerald Berenson, M.D., Ph.D. Professor of Medicine Louisiana State University Medical Center New Orleans, La.

Jay Brown, M.D. Chief, Cardiology Harlem Hospital Center New York, N.Y.

Richard Cooper, M.D. Assistant Professor Department of Community Health

and Preventive Medicine Northwestern University Chicago, Ill.

Charles Curry, M.D. Director, Cardiology Disease Division

723

Documents

A note on the biologic concept of race and its application in epidemiologic research