Sociology assignments

American Anthropological Association and Wiley are collaborating with JSTOR to digitize, preserve and extend access to Medical Anthropology Quarterly.

http://www.jstor.org

American Anthropological Association Wiley

Health in the African American Community: Accounting for Health Inequalities Author(s): William W. Dressler Source: Medical Anthropology Quarterly, New Series, Vol. 7, No. 4, Racism, Gender, Class, and

Health (Dec., 1993), pp. 325-345 Published by: on behalf of the Wiley American Anthropological Association Stable URL: http://www.jstor.org/stable/649213 Accessed: 27-10-2015 19:54 UTC

REFERENCES Linked references are available on JSTOR for this article:

http://www.jstor.org/stable/649213?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

ARTICLES

WILLIAM W. DRESSLER Department of Behavioral and Community Medicine University of Alabama School of Medicine

Health in the African American Community: Accounting for Health Inequalities

African Americans are at a higher risk of having a variety of health problems and have less access to health care than white Americans. This article explores these health inequalities and their explanations. Three conventional models of health inequalities-a racial-genetic model, a health behavior or lifestyle model, and a socioeconomic status model- are examined andfound to be insufficient to accountfor observed dispari- ties. A fourth alternative, termed a "social structural model," is proposed. In this model, it is argued that the primary index of ethnic status, namely skin color, serves as a criterion of social class in color-conscious societies such as that of the United States. This alters social mobility processes and creates health inequalities for African Americans. [health inequalities, ethnicity, African American]

frican Americans suffer by comparison with white Americans on essen- tially every indicator of morbidity, mortality, and access to health care. The issue, from the standpoint of social epidemiology or of health policy, is to

understand why this is so. An understanding of this health inequality can provide insight into fundamental social and pathophysiologic processes; additionally, an understanding of this health inequality can provide an empirically verified basis for intervention strategies to reduce this disparity at the community, family, and individual levels.

My contention here is that social epidemiologic research has been unable to adequately address this inequality because of poorly specified social scientific models of ethnic group differences in moder industrial society. Only by examining our fundamental assumptions about the nature of ethnic groups will we be in a position to understand the ways in which beliefs, behaviors, and social structural

Medical Anthropology Quarterly 7(4):325-345. Copyright ? 1993, American Anthropological Asso- ciation.

325

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

variables, not to mention whatever biological factors could prove relevant, contrib- ute to health inequalities.

The difficulty in understanding these issues does not stem from a dearth of research. Over the past 70 years, 64 percent of all epidemiologic research articles dealing with human subjects published in the American Journal of Epidemiology had some reference to "race," while 34 percent explicitly included some "nonwhite" population (Jones et al. 1991). Were this survey of the inclusion of group differences in health research to be expanded to all scientific publications, the number of published articles would be truly prodigious; however, the understanding of the basis for differences in health by ethnic group remains untheorized and poorly analyzed. The authors of the survey of epidemiologic articles on "race" and health conclude by urging "epidemiologists to more actively engage the issues of 'race', rather than avoiding them and allowing 'race'-associated differences to persist well documented but unexplained" (Jones et al. 1991:1083).

It is my contention here that the problem with addressing these issues stems largely from the failure of social scientists in general, and anthropologists in particular, to formulate empirically useful (i.e., operational) theories of ethnicity, theories that can be used to evaluate the observations of ethnic group differences in health in ways that reflect the biocultural complexity of the phenomena. It can only be the development of such models, and the careful empirical evaluation of those models, that will contribute to a rethinking of issues of race and ethnicity in the less socioculturally (and more biomedically) inclined disciplines.

Clearly, the critique of oversimplified notions about group differences in biology and behavior is well developed in anthropology, as evidenced by the consistent (if not complete) rejection of the concept of race in biological anthro- pology. Some years ago, Littlefield et al. (1982) documented the decline in the use of the term race in basic texts in biological anthropology. Although the term has not disappeared from use in biological anthropology, the tendency has been to recognize explicitly the social, as opposed to the biological, basis of its application. When the evolutionary biology of populations is addressed, "cline" or some other geographic referent of a smaller scale than "race" is deemed to be more appropriate (Harrison et al. 1988:322-332). This is, of course, partly because of the recognition over the years of the extreme heterogeneity within so-called racial groups in essentially any characteristic considered. Following others (Montagu 1945), I will use the term "ethnic group" here, but also following the recommendation of Crews and Bindon (1991), I will try to make explicit just what is meant by that term as the argument progresses.

The importance of clarifying terminology, and especially of developing useful theory of biocultural processes and ethnicity, can be exemplified by reference to an article by Rushton and Bogaert (1989) on population differences in acquired immunodeficiency syndrome (AIDS). These authors adopt an orientation from evolutionary biology to "predict" that "Mongoloids" will have a lower prevalence of AIDS than "Caucasoids," who in turn will have a lower prevalence than "Negroids." Even though this article was published in a multidisciplinary journal in which anthropology is well represented, the argument of Rushton and Bogaert rested on antiquated racial characterizations of large populations differing in general (and operationally nonspecific) attributes such as "brain size and intelli- gence (cranial capacity, brain weight, test scores), maturation rate (age to hold head

326

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

erect, age to walk alone, age of death), social organization (marital stability, mental disorder, law-abidingness), and temperament (activity level, anxiety, sociability)" (1989:1215). According to Rushton and Bogaert, these racial differences in behav- ior are a product of evolutionary adaptation and, in turn, influence the risk of AIDS (1989:1217).

Leslie (1990), in a blistering critique of this article, demonstrates that, shorn of its borrowed (and often distorted) terminology from evolutionary biology, Rushton and Bogaert's argument can be shown to be consistent with a North American racist ideology of the most vulgar sort, and Leslie suggests that the article never should have been published. But the commentary by McEwan, the editor of the journal, is instructive. In a carefully reasoned and carefully worded statement, he argues that, regardless of the extent to which an analysis of ethnic differences can be shown to be consistent with commonsense, perhaps racist, notions the fact remains that ethnic differences exist.

To reinforce criticism alternative explanations for apparently legitimate variations should be provided.... [T]he argument [challenging racial explanations] needs to be stated and defended, not assumed as a holy tenet which it is scientific heresy to question. Alternative explanations will then be advanced to account for what- ever variations may remain. As matters stand, there is a danger that certain branches of science, by fearing to enter the arena, fail to meet their obligation. [1990:911].

That obligation is, of course, the development of theory, from which operationally specified models that define alternatives to race-based notions can be derived and tested.

In the remainder of this article, I focus on a single ethnic contrast in health, that between African Americans (who are also referred to as black Americans) and European Americans (also referred to as white Americans). First, the differences (or inequalities) in health between ethnic groups are reviewed, although not exhaustively (extensive reviews are found in Nickens [1986] and Braithwaite and Taylor [1992]). My aim is simply to illustrate the extent of differences. Second, I briefly review the major types of explanations proposed to explain these health inequalities, primarily to note the inadequacies of each to account empirically for the general phenomenon. Third, I propose and discuss an alternative model.

Health Inequalities: African Americans and White Americans

Black Americans die at higher rates from all major causes of mortality in the United States than do white Americans. The statistics in Table 1 show, for a select set of causes, the total number by which deaths per year in the African American community exceed the number that would be anticipated were the mortality rates among blacks equivalent to the rates among whites. While blacks die at higher rates than whites, it is worth noting that this health inequality is not distributed equally across all causes of mortality. The greatest disparity occurs in the category-car- diovascular disease-that is also the major cause of death among white Americans. Major disparities also occur in categories-infant mortality and homicide-that are not major contributors to total mortality.

327

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

TABLE 1 Excess deaths* in the African American community by sex, 1979-81.

Cause of Mortality Males Females Cardiovascular disease 8,469 9,712 Cancer 5,782 2,269 Cirrhosis 1,362 782 Infant mortality 3,317 2,861 Diabetes 646 1,203 Injuries 1,113 134 Homicide 6,708 1,381

Source: Nickens 1986. *Excess deaths are defined as the difference between the annual number of deaths

observed in the African American population and the number that would be expected if the age- and sex-specific death rates in the black population were the same as those for the majority population.

The trend over time has not been in the direction of reducing health inequali- ties, at least as far as the major cause of mortality for both ethnic groups is concerned. Sempos et al. (1988) examined the well-known decline in coronary heart disease mortality rates in the United States separately for ethnic groups and gender. Between 1968 and 1975, the average annual rate of decline in heart disease mortality was similar for black males, white males, and black females, as shown in Figure 1. Between 1976 and 1985, however, the rate of decline of heart disease mortality diminished significantly for black males, black females, and white females, but not for white males. That is, white males continue to enjoy an ever-increasing likelihood of avoiding or surviving heart disease, while the likeli- hood that other ethnic and gender groups will do so is decreasing.

With respect to health inequalities in morbidity, it is useful to examine conditions that underlie major causes of mortality among African Americans. These are essential hypertension, a precursor of coronary heart disease, and low birth- weight, a precursor of infant mortality. The rate of essential hypertension among black Americans (approximately 37 percent) is about twice that among white Americans (approximately 18 percent) (Heymsfield et al. 1977). This inequality in prevalence has been observed in numerous studies (Ford and Cooper 1991). Similarly, the risk of having low-birthweight infants among African American women (11 percent) is about twice that among white women (5 percent) (Starfield et al. 1991). Again, this disparity has been observed in numerous studies (David and Collins 1991).

Much research effort has been devoted to the explanation of these health inequalities, with little success. Conventional risk factors have not explained differences, nor has controlling for differences in conventional measures of socio- economic status. If years of education are controlled for, blacks with fewer than ten years of education are about 60 percent more likely to have hypertension than are blacks with a college education; however, at any level of education, blacks are more

328

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

- White Male -- Black Male -- White Female - Black Female

, Rate of Decline IV

9

. B U 7

6

5

4

3

2

1968-75 1976-85

Time Period

FIGURE 1 Average annual rate of decline in CHD mortality rates (in deaths/100,000 pop.) over

time by ethnicity and gender. (Source: Sempos et al. 1988.)

likely to have hypertension than are whites, as shown in Figure 2 (Ford and Cooper 1991; Heymsfield et al. 1977). Black males with a college education are over twice as likely to have hypertension as white males with a college education. The pattern is somewhat more complicated for low birthweight. In some studies, controlling for maternal education has no impact on black-white disparities in low birthweight (see Figure 3). Other studies have found that imposing controls for maternal education actually increases the black-white differences. For example, Starfield et al. (1991) found that for mothers with less than a high school education, the black-white ratio for risk of low birthweight is 1.9. For mothers with more than a high school education, the corresponding ratio is 2.8 (Starfield et al. 1991).

The complicated relationships between ethnicity, conventional measures of socioeconomic status, and morbidity are also illustrated in the work of Kessler and Neighbors (1986) on depression. A number of papers (e.g., Comstock and Helsing 1976) have shown that adjustments for sociodemographic variables eliminate differences between blacks and whites in rates of high-depressive symptomatology. Kessler and Neighbors (1986) examined the statistical interaction between indica- tors of socioeconomic status and ethnicity in relation to depression; that is, they addressed the issue of whether or not the effect of socioeconomic status was the same for blacks and whites. They found that, for example, the decline in risk of

1

329

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

% Hypertension * Blacks

E Whites

35-

30-

25-

20-

15-

10-

5-

0-

Years of Education

FIGURE 2 Prevalence of hypertension for black and white males by education. (Source:

Heymsfield et al. 1977.)

high-depressive symptoms associated with increasing income was steeper for blacks than whites. Put differently, low income placed blacks at higher risk of depression than whites; high income was equally associated with lower risk of depression for the two groups.

Before leaving the issues of morbidity and mortality, it is useful to further complicate the picture by noting the phenomenon of intra-ethnic diversity. For example, the risk of infant mortality varies by the nativity of the mother. Kleinman et al. (1991 ) found the crude rates of infant mortality to be 40 percent higher among native-born black mothers than among foreign-born black mothers; the rates were equal for white mothers regardless of nativity. Differences persisted, although reduced in magnitude, when other factors (including maternal education) were controlled. Similarly, Dressler and Badger (1985) compared the relationship be- tween sociodemographic variables and depressive symptoms in three samples of African Americans. Two were drawn from urban communities in the northern United States, and one was from a small city in the South. There was a sharp decline

330

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

% Low Birthweight 14

12 a

10

8

6 *

4

2

0

-- Black - White

<12 12 13-15

Maternal Education (yrs.)

FIGURE 3 Percent low birthweight by education and ethnicity. (Source: Starfield et al. 1991.)

in the prevalence of depression with increasing income in the two northern communities, while in the South the effect was attenuated.

Turning next to the issue of access to health care, health inequalities persist. By survey measures, overall access to health care has improved over the past 30 years for both blacks and whites, and there is some convergence in some areas. For example, data reported by Jaynes and Williams (1989) show that disparities in physician visits between blacks and whites have declined. This decline is probably associated with the introduction of federal health care programs for the poor and the aged. There continue, however, to be inequalities in the proportions of blacks and whites under age 65 not covered by health insurance (22 percent versus 14 percent, respectively). Furthermore, as Neighbors and Jackson (1986) have shown, the proportion of medically uninsured can climb to alarming proportions among some categories of African Americans. For example, 32.3 percent of blacks living in rural areas have no health insurance.

While access to care has improved somewhat for black Americans, there is disturbing evidence of inequality in treatment within the health care system. This inequality is most precisely documented for the major cause of mortality in the United States for both blacks and whites, coronary heart disease. A white person admitted to a hospital for chest pain and a presumptive diagnosis of coronary artery disease is more likely to receive sophisticated diagnostic procedures such as arteriography. If the diagnosis has been confirmed, a white person is twice as likely

331

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

TABLE 2 Ratio of whites to blacks referredfor coronary artery bypass grafts in three studies.

Study White/black ratio 1. Oberman and Cutter 1984 2.34a 2. Wenneker and Epstein 1989 1.89b 3. Fordet al. 1989 2.85c

aAdjusted for blood pressure, age, congestive heart failure, occurrence of myocardial infarction, angina, smoking, and number of diseased vessels among persons referred for evaluation.

bAdjusted for age, sex, diagnosis, insurance status, income, number of diagnoses, and admission type among persons referred for evaluation.

CAmong persons who have had a myocardial infarction, but unadjusted for other factors.

to receive a coronary artery bypass graft (see Table 2). This association persists after controlling for disease severity, other risk factors, income, and insurance status (Ford et al. 1989; Oberman and Cutter 1984; Wenneker and Epstein 1989).

Finally, looking at perceived health status, blacks are more likely than whites to assess themselves as being in poor health. Controlling for income reduces this disparity considerably; however, even within the high-income group, blacks are about 50 percent more likely to assess their health as poor (Navarro 1991). With respect to the perceived adequacy of health care, 12 percent of the sample in the National Survey of Black Americans indicated that they had "unmet health care needs" (Chatters 1991).

Models to Account for Health Inequalities

As noted in the introduction to this article, the health differences between ethnic groups have been amply documented but remain untheorized. Investigators seem to be more comfortable with noting the existence of differences than with contemplating the basis for those differences, beyond near platitudes concerning the need for more research to investigate one or another aspect of the problem. Of course, researchers ought not to be judged too harshly for this timidity. Discussions of ethnic differences are fraught with dangers for the unwary, and broadside indictments of a tainted ideology on the part of the investigator are likely to come from just about anywhere on the political spectrum. Nevertheless, the question of the basis for these differences cannot be avoided. Here I will review four models that have been proposed to account for the health inequality between black and white Americans. These are: (1) a racial-genetic model; (2) a health behavior/life- style model; (3) a socioeconomic status model; and (4) a social structural model. The first three models are commonly encountered in the literature, although not necessarily under these labels. They cannot be exhaustively reviewed here; rather, I will simply note a few major problems with each, leading to a recommendation for an alternate approach.

332

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

Racial-Genetic Model A racial-genetic model posits group differences in the distribution of some set

of genes that gives rise both to phenotypic differences between groups (e.g., skin color) and to the propensity to poor health. Rarely will one see in the literature a blatant assertion that all differences in health between blacks and whites are genetic in origin, yet the cumulative effect of individual epidemiologic studies reinforces the notion. Explicitly, "race" has been demonstrated to be nearly a useless biological construct (Montagu 1962), primarily because of operational nonspecificity. In the Western industrialized societies, race primarily refers to overt, phenotypic differ- ences, usually skin color, and sometimes including other features. Rarely, if ever, is race operationalized in terms of specific frequencies of specific genes. Therefore, the consideration of race as a biological variable proceeds by implicit, rather than explicit, definition. There is a long history in the West of the reification of the categories of race and, unfortunately, health research contributes in no small measure to that reification. That is, if race is defined in a way that is implicitly-but not operationally-based in biology and it can be shown that "races" differ with respect to health, the "biologicalness" of race is reinforced. That, however, depends crucially on the initial assumption that skin color and other characteristics index biological differences that are important with respect to health and disease. This is, at best, a dubious assumption (Robins 1991); at worst, it can be shown that skin color indexes many social facts about a person other than his or her genetic endowment (Keith and Herring 1991), which thus confounds social and biological facts.

Even were it not demonstrable that skin color measures factors other than biological differences, it stretches one's credulity to come up with a genetic hypothesis to account for the variety, including intra-ethnic variability, of the health inequalities documented. That is, the health inequalities documented here recur across such a wide variety of diseases and, in fact, are recapitulated in measures of access to health care to such a degree that a genetic hypothesis seems implausible. Indeed, research on genetic factors in disease is moving in the direction of the identification of very specific chromosomal differences (Cox 1992); the kinds of broad phenotypic differences implied in the use of the term "race" have little bearing on moder genetic research.

Finally, precise research on diseases in which genetic factors are known to be important confirms the irrelevance of racial categories. For example, David and Collins (1991), discussing birth outcomes, note that the rates of infant mortality due to known genetic causes do not differ by ethnic group. Similarly, sickle-cell anemia is now known to have three different types, each with a specific set of chromosomal variations; each arose in different parts of Africa; each is differen- tially distributed across black populations in the West; and, sickle-cell anemia is found in nonblack ethnic groups in the West (Zago et al. 1992). Again, racial categories are of little relevance.

Therefore, as an explanation for the general phenomenon of ethnic health inequalities, a racial-genetic hypothesis is of little utility. This further reinforces the utility of the exclusive use of the term "ethnic group."

333

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

A Health Behavior/Lifestyle Model A second model to account for ethnic inequalities in health is a health

behavior/lifestyle hypothesis. This model posits ethnic differences as a function of unhealthy behaviors, such as smoking, excessive alcohol use, high fat and high- sodium diets, and lack of physical activity, that are differentially distributed among groups. This kind of hypothesis has somewhat more plausibility than a genetic model, at least insofar as it specifies ethnic differences in discrete variables, rather than remaining vague and global as a genetic explanation does.

There are, however, some problems. To the best of my knowledge, in a specific disorder such as essential hypertension, no one has been able to demonstrate empirically that controlling for health behaviors causes black-white differences to disappear. For example, adjusting black-white differences in blood pressure for obesity ought to go a long way in reducing those differences, because obesity ought to be a composite indicator of an unhealthy diet and low physical activity; but controlling for obesity does little to reduce ethnic group differences (Heymsfield et al. 1977). Similarly, adjustments for obesity, alcohol consumption, and physical activity have little impact on ethnic differences in serum lipids (Freedman et al. 1992). Therefore, while it is a plausible alternative, the health behavior model simply fails empirically (see also Duelberg 1992).

There is another problem with this model, a problem that is reminiscent of the old "culture of poverty" debate. That is, a health-behavior hypothesis strongly presumes that the sole locus of relevance for health is the individual. This presump- tion is reinforced by the unfortunate application of the term "lifestyle" to health behaviors, with its connotation that these behaviors are blithely (and happily?) chosen by the individual. It then becomes the victim's own fault that he or she was so ignorant or misguided as to have chosen to engage in that behavior. This model tends to ignore the larger social context; in specific terms, it ignores social agenda such as the differential targeting of ethnic groups in the marketing and sale of tobacco or alcohol products. While individual behaviors certainly do predispose individuals to one disease or another, the scope of health inequalities is unlikely to be explained by these differences alone.

A Socioeconomic Status Model

The criticisms of both the racial-genetic and the health behavior/lifestyle models lead naturally to a consideration of the socioeconomic status model. Many have argued that ethnic group membership is confounded with differences in socioeconomic status such as income, occupation, and educational attainment; therefore, apparent racial differences in health may simply be a function of this confounding. Similarly, surveys of diet, smoking, alcohol use, and physical activity demonstrate a consistent association of low socioeconomic status and deleterious health habits. If controls for socioeconomic status reduced health inequalities, it could mean that African Americans are overrepresented among persons with unhealthy habits.

Unfortunately, the socioeconomic status model does little to account for ethnic health inequalities. The evidence noted previously in studies of essential hyperten- sion and low birthweight is particularly striking in this regard; controlling for

334

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

conventional measures of socioeconomic status does little to reduce ethnic differ- ences in disease rates. This presents a conundrum to the investigator in social epidemiologic research who is aware of the confounding of ethnicity and socioeco- nomic status. Yet, as I have argued elsewhere (Dressler 1991a), the notion that ethnicity and socioeconomic status are simply confounded is theoretically too anemic to account for the relationship of ethnicity and social class. While many writers attempt to eschew any strong theoretical inclination in the definition of socioeconomic status in favor of the simple inclusion of indicators of status in quantitative models, it is important to recognize that the mere selection and inclusion of variables in a model is theoretically driven, however hidden or implicit that theory may be.

The assumption implicit in the confounding hypothesis is derived most likely from a general functionalist perspective on socioeconomic status. In this perspec- tive, the vertical, unequal distribution of power, resources, and status (or prestige) is viewed as necessary for the smooth operation of a society. Persons "qualified" on the basis of intelligence, commitment, motivation, and the like must be recruited to fill socially necessary occupational roles, so these unequal rewards must attach to the roles. This vertical distribution of resources can then be described by the roles themselves (occupation), the credentials needed to fill the roles (education), and the rewards that accrue to the roles (income) (see Davis and Moore 1945).

From this perspective, the association of African American ethnicity and lower socioeconomic status is seen as a result of historical forces once related to social class processes, but now largely tangential to them. That is, under slavery and segregation, black persons were denied, by force and by statute, the potential of filling valued roles, and hence denied the potential for upward mobility. With the advent of the civil rights movement, however, these political forces have been removed. Therefore, as the mobility associated with the functionalist perspective sets in, more black people will be represented in higher socioeconomic statuses, thus attenuating the confounding of ethnicity and socioeconomic status. The crucial assumption here is that now, in a post-civil rights movement era, processes of socioeconomic mobility are the same for blacks and whites in this society. By implication, then, adjusting for socioeconomic status, under these assumptions, would account for ethnic health inequalities.

The failure of such an adjustment process to account for ethnic differences has already been documented. But is the critical assumption, that socioeconomic processes are equivalent for black Americans and white Americans true? Ample evidence exists that the promise of the civil rights movement has remained unfulfilled (Dressier 1991b; Farley 1985). On a variety of criteria, it can be demonstrated that African Americans have not had the economic and material conditions of life substantially altered relative to the white majority. The fact of change associated with the civil rights movement is undeniable, but the degree of change remains very much a moot point. For example, hourly wages paid to black workers have nearly converged with those of white workers, yet the ratio of black median family income to white median family income has remained nearly constant at around .60 for 40 years (Farley 1985). The structure of income production for black and white households is also very different; an individual wage earner (usually the head of household) contributes a large proportion of white household income, while in black households total income tends to be proportionately distrib-

335

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

uted across several wage earners (Dressier 1991b). Futhermore, higher education does less to protect African Americans from the risk of unemployment than white Americans (Willhelm 1986).

This difference in socioeconomic process is nowhere more clearly demon- strated than in a number of recent articles examining the association of skin color (from darker to lighter) with higher income, occupational attainment, and educa- tional levels within the black community (Dressler 1991 a; Hughes and Hertel 1990; Keith and Herring 1991). A direct association between lighter skin color and higher socioeconomic status persists within the black community, an association that replicates across a variety of studies. In short, to profit from the traditional avenues to upward mobility in American society, an African American not only must attain more education and achieve a higher occupational status, but he or she must also phenotypically conform to some externally-imposed criterion (which, by the way, provides empirical support for the social basis of racial categories). In this sense, then, "controlling for" socioeconomic status would seem to be logically impossible, because skin color itself becomes a criterion of status. Or, as Cooper has observed:

Explaining racial differentials by education, income, etc., could in a causal sense be considered "overcontrol:" race is not confounded by the other variables; it is antecedent to them. It is race that influences class standing. [ 1984:721 ]

What is needed is a model of ethnic health inequalities that takes into account the fact of continuing discrimination against people of color in color-conscious socie- ties, a fact that automatically assigns people of color to a lower-ranked position in the social structure. I will refer to this as "a social structural model."

A Social Structural Model

The theoretical foundation for such a model is provided in the work of Parkin (1979) and Murphy (1986), sociologists who have led in the reformulation of the traditional interpretation of Weber's concept of social closure. Conventional inter- pretations of Weber underlie the use of socioeconomic status indicators. Occupying different points on the continua defined by these indicators provides individuals different "life chances" relative to the marketplace in capitalist societies-that is, being higher in the status hierarchy improves one's chances in life. Parkin, Murphy, and others, while recognizing the importance of these factors in relation to the market, emphasize in addition Weber's concept of "social closure." This refers to the active mobilization of power to enhance or preserve one's place in the hierarchy. Enhancement or preservation may proceed by excluding certain classes of persons from competition for social rewards, or in the case of lower status persons or groups, it may proceed by attempts to usurp others' prerogatives.

The major criteria by which, for example, individuals are excluded from entry into higher status groups are the usual ones adopted in measures of socioeconomic status. That is, rarely are unskilled laborers allowed into exclusive strata of society; or, it is unlikely that an individual lacking educational credentials will be accepted as a political or opinion leader. But, as the social closure theorists make clear, "any convenient and visible characteristic, such as race, language, social origin, religion, or lack of particular school diploma, can be used to declare competitors [for status] as outsiders" (Murphy 1986:23). Therefore, in a skin-color-conscious society like

336

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

the United States, the phenotypic characteristic conventionally thought to be indicative of a genetic-racial trait is actually a criterion used, along with others, in the fundamental definition of social class. Furthermore, as Kemper and Collins (1990) and Bourdieu (1984) argue, these processes defining, as it were, "suitability" for status group membership can get played out at the individual, as well as the group, level. The theory of social closure subsumes conventional notions of socioeconomic status and alerts us to the social significance of biological traits such as skin color.

How does this perspective on ethnicity and social class further an under- standing of health inequalities? Recently I used this orientation to examine findings that have been interpreted in other studies as evidence of a genetic etiology for essential hypertension in the African American community. Darker-skinned black people have higher blood pressure than lighter-skinned black people, which some have argued is a genetic effect (Boyle 1970; Harburg et al. 1978; Klag et al. 1991). Others have argued, however, that darker skin color automatically assigns African Americans lower social status (Blackwell 1975), which in turn could be related to autonomic nervous system activity and higher blood pressure (Thomas and Dob- bins 1986). The problem, of course, lies with developing an operationally specific model that would allow a choice to be made between alternatives.

As a way of specifying a model, I suggested that skin color as a measure of social status could be incorporated into the operational definition of status incon- gruence, a factor that has been found to be associated with blood pressure in several settings (Dressier 1993). Briefly, status incongruence is a circumstance in which an individual occupies inconsistent positions on two dimensions that define social prestige. In Western society, one status dimension of primary importance, because it can communicate the possession of wealth in mundane social interaction, is lifestyle. Lifestyle is defined here in its original sense as the accumulation of material goods and the adoption of status-enhancing behaviors. If, however, an individual attempts to maintain a lifestyle that is higher in rank than his or her occupation or education, he or she may be perceived by others to be fraudulently laying claim to that higher status; in such a situation, the individual is unlikely to receive confirmation of that claimed social status in mundane social interaction. This can lead to a vigilant style in social interaction, heightened cardiovascular reactivity, and sustained elevated blood pressure.

Ethnicity was incorporated into this model in a study carried out in a small city in Alabama. It was argued that darker skin color would allocate persons to lower social status just as lower occupational or educational status would; therefore, the association of higher blood pressure with darker skin color should be absorbed entirely by a category defined by darker skin color and higher lifestyle. To test this hypothesis, data from a random sample of 186 persons, age 25 to 54, were employed, a sample that had been selected from two African American neighbor- hoods. The sample had been selected to test the status incongruence hypothesis. The hypothesis was confirmed using the discrepancy between lifestyle and occu- pational status (Dressier 1990a) and the discrepancy between lifestyle and educa- tional credentials (Dressier 1990b).

Skin color was assessed by having African American interviewers allocate respondents to one of three categories: dark, medium, or light. To form a measure of status incongruence, the lifestyle scale (a 23-item scale of reported ownership

337

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

TABLE 3 Cross-tabulation of categories of skin color and lifestyle, illustrating the measurement

of status incongruence.

Lifestyle Skin Color Low Moderate High Total Dark 19 34 9 62

(B) (C) (C) Medium 10 43 17 70

(A) (B) (C) Light 3 32 18 53

(A) (A) (B) Total 32 109 44 185

of consumer goods and reported behaviors such as television watching and travel) was first trichotomized, and then lifestyle and skin color were cross-tabulated, as shown in Table 3. The status incongruence hypothesis is directional; that is, only lifestyle exceeding the other status measure is considered to be problematic. The opposite sort of incongruence (i.e., higher occupational status with lower lifestyle), or what might be considered colloquially to be living "modestly," is not problem- atic. Therefore, three groups from Table 3 are important: persons with a lighter skin color and lower lifestyle (those cells labeled "A"); persons who fall into commen- surate categories on both skin color and lifestyle (cells labeled "B"); and, those persons with darker skin color and higher lifestyle (cells labeled "C"). Note that allocation to these groups depends on relative rank on the two variables, not some absolute rank.

The association of systolic and diastolic blood pressure, and hypertension (diastolic blood pressure > 90 mm Hg), with these categories of status incongruence is shown in Table 4 (in which dependent variables have been adjusted for age, sex, body mass index, education, independent effects of skin color and lifestyle, and interaction of education and lifestyle incongruity). The association is significant for both systolic (p < .02) and diastolic (p < .03) blood pressure, and the effect is independent of education. When hypertension is used as the dependent variable in a logistic regression analysis, the risk odds ratio associated with the darker skin color/higher lifestyle category is 6.24 (p < .02) (see Dressier 1991 a for more detail).

As anticipated, the association of skin color and blood pressure can. be accounted for entirely with a model in which the biological effect of skin color is assumed to be socially, as opposed to genetically, mediated. Evidence from a related study (Dressier 199la) suggests that a black person who is maintaining a higher status lifestyle and who has darker skin meets with more frequent frustrating social interactions, interactions in which he or she is not treated in a manner commensurate with the level of social status claimed. These repeated, frustrating social interactions can lead to repeated autonomic arousal and sustained blood-pressure elevation.

338

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

TABLE 4 Systolic (SBP) and diastolic (DBP) blood pressure and percent hypertension (% Htn) by categories of status incongruence within educational categories (adjustedfor age,

sex, body mass index, independent effects of skin color and lifestyle, education, and the interaction of education and status incongruence).

Education < 12 years > 12 years

Status incongruence SBP DBP % Htn SBP DBP % Htn

Lighter skin color/lower lifestyle (n = 43) 114 78 14.3 112 76 8.7

Skin color/lifestyle equivalent (n = 79) 123 82 20.5 121 82 22.9

Darker skin color/higher lifestyle (n = 59) 128 85 42.9 125 86 26.7

Note: For SBP, the statistical effect of status incongruence from ANOVA is F = 4.63; df= 2,180; p = .011. For DBP, the effect is F = 3.75; df= 2,180; p = .025. For % Htn, the effect is calculated from logistic regression: ROR = 6.24, Wald statistic = 6.12, p = .013.

The processes by which social closure leads to health inequalities need not be viewed solely in terms of a status incongruence model; rather, the process can occur at different levels. LaVeist (1992) recently examined infant mortality rates in 176 U.S. cities. He focused on a measure of black political participation in city government as a predictor of infant mortality rates. As the ratio of black elected officials to the black voting population increases-indicating proportional repre- sentation-infant mortality rates decline. This effect is observed while controlling for a variety of variables, including poverty levels, degree of residential segrega- tion, rates of low birthweight in the population, and others. Furthermore, LaVeist found that the beneficial effect of black political participation on infant mortality rates was unmediated by expenditures on health services, although other municipal expenditures were positively associated with participation. This suggests that higher black political participation, municipal expenditures, and lower infant mortality rates are all related to an underlying, latent variable of social structure. To the extent that community organization in the minority community assists persons in usurping some of the social and political prerogatives of the majority community, better services and health result.

This social structural model also can be useful in understanding other findings, such as the greater likelihood of referral for coronary artery bypass graft surgery if one is white. It is important to remember that this finding cannot be rejected on the basis of differences in pathology or ability to pay. On the one hand, it could be that black patients refuse surgery at a substantially higher rate than white patients. This seems unlikely. On the other hand, it could be that black patients are actively discriminated against by referring physicians. One hopes this is not the case. A third alternative, consistent with the social structural model, could be that providers

339

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

attach extraneous meaning to social categories such as ethnicity. The risk of cardiovascular disease among African Americans is not widely appreciated by many physicians, despite the clearly higher prevalence of essential hypertension (Young et al. 1991). Similarly, those elusive factors that add up to define "quality of life" may not be applied to blacks and whites in the same way, such that African Americans may be thought not to be "appropriate" candidates for surgery, but rather get offered medical intervention. Eisenberg (1979) has discussed these and similar cultural assumptions that may lead to specific patterns of clinical decision making, and Schwartzbaum et al. (1990) have shown that physicians' decisions are influ- enced by the ethnicity of their patients. My point here is that, whatever the phenomenology of the decision-making process, the structural result of many individual decisions (which are of course based on socially patterned meanings regarding ethnicity) can add up to the exclusion of whole categories of persons from health care, based on criteria such as skin color. A social structural model of health inequalities can thus facilitate the search for mechanisms. Such a model helps us to direct our thinking away from the usual focus on the individual and, instead, to focus on the social-relational structure, which in turn can generate the large-scale group differences observed in health and health care.

Discussion

The aim of this article is to suggest a theoretical orientation for studying ethnic iniqualities in health. The specific focus here has been on differences in health between African Americans and European Americans in the United States, although with some modification, the model proposed here can be specified for the study of all ethnic health inequalities. The need for a new model for ethnic differences in health stems precisely from the generality of the phenomenon. That is, for each of the health problems examined here, there certainly are specific etiologic factors for that particular problem. Yet ethnic differences persist across all problems after specific influences have been accounted for in data analyses, which is precisely why ethnicity is such an enduring, although untheorized, variable in epidemiologic research (Jones et al. 1991). Therefore, in addition to specific etiologies, there is a general pattern in need of explanation (see Marmot et al. 1984 for a similar discussion with respect to the issue of social class differences).

To say the least, the study of ethnic differences in health is a contentious issue, but that contention revolves around a fairly consistent set of findings. That is, blacks and whites differ on a variety of measures of health status, with black people generally evincing poorer health than white people. Research in which health behaviors (such as diet, physical activity, and alcohol or cigarette use) or conven- tional measures of socioeconomic status (such as education level or income) are controlled for in ethnic comparisons has demonstrated that these variables do little to account for health inequalities. A common, but often implicit, inference is that these ethnic differences are genetic in origin.

It was argued here that a genetic hypothesis for ethnic health inequalities can be faulted for two reasons. First, the anthropological critique over several decades of the concept of "race" suggests that there are few biological differences between these large population groups that are of explanatory relevance. There is simply too much variation within groups. Second, in diseases in which there is a clear genetic

340

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

effect, gene frequencies at the population level described by the race concept are irrelevant; intrapopulation diversity in these gene frequencies is such that smaller groups than so-called racial groups are necessary to understand the genetic epidemiology of the diseases. (Of course, the question of the persistence of racial categories in Western thought remains unanswered, although Wade (1993) offers some interesting notions.)

For both these reasons, and as Crews and Bindon (1991) argue, ethnic group was chosen as the more appropriate theoretical term, because of the possibility of greater operational specificity. In the precise comparison examined here, skin color was emphasized as the primary determinant of ethnic status, although of course in other kinds of comparisons, other measures of ethnic status would be chosen (see Kaw 1993 for a fascinating example of a very different indicator of ethnic status and attempts to alter it). In many cases, skin color may be appropriate, but along with other indicators such as language, religion, or even stature. The point is to be operationally specific, and skin color was chosen here because, as others have pointed out (Blackwell 1975; Thomas and Dobbins 1986), it is darker skin color that is automatically assigned lower status in American society.

But simply asserting that is not novel, nor is it sufficient to be able to distinguish the social effects of ethnicity from lingering notions, either explicit or implicit, of the biological basis for ethnic health inequalities. Rather, the process by which ethnic differences are translated through social and somatic pathways into biological health differences must be specified both theoretically and operationally. The formulation here of a social structural model of ethnic health inequalities is just such an attempt. It has been demonstrated that differences within the black community in blood pressure can be accounted for in part by the discrepancy between skin color and consumer lifestyles (Dressier 1991a). Similarly, LaVeist (1992) showed that intra-ethnic variability in infant mortality rates could be accounted for by political participation.

Earlier in this article, I suggested that the association of ethnicity and health status reinforced a socially constructed tendency to think about ethnic differences in racial terms. By the same token, treating an indicator of ethnicity, as I have in research on blood pressure, as another measure of social status may strike some as odd, or even as absurd. But any indicator or variable has meaning only in relation to some larger theoretical system. When examined in terms of the social structural model of ethnic health inequalities, the use of skin color as a social variable, to be combined in various ways with other measures of social status, becomes straight- forward. This is, after all, the job of theory, to make the obscure straightforward. The job of empirical research is then to evaluate whether or not that theory, which simplifies our understanding, is useful to account for observed regularities. Unfor- tunately, social science in general and anthropology in particular have not gone far enough in dismantling old notions of race. While the critique of the concept of race as being of little biological utility, with the accompanying recommendation for the substitution of the concept of ethnicity, has been useful, there has remained a need for an alternative and specific theory to account for the regularities of ethnic health inequalities. The social structural model is intended to be one step in that direction.

Acknowledgements. Dr. John Wheat, Dr. Cynthia Cole, Dr. James P. Adams, and especially, Dr. Kathryn S. Oths offered helpful comments on an initial draft of this article. Anonymous reviewers also provided very helpful direction.

341

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

Correspondence may be addressed to the author at Department of Behavioral and Community Medicine, University of Alabama, P.O. Box 879326, Tuscaloosa, AL 35487- 0326.

REFERENCES CITED

Blackwell, James E. 1975 The Black Community: Diversity and Unity. New York: Harper and Row.

Bourdieu, Pierre 1984 Distinction: A Social Critique of the Judgement of Taste. London: Routledge and

Kegan Paul Boyle, E.

1970 Biological Pattems in Hypertension by Race, Sex, Body Weight, and Skin Color. Journal of the American Medical Association 213:1637-1643.

Braithwaite, Ronald L., and Sandra E. Taylor, eds. 1992 Health Issues in the Black Community. San Francisco, CA: Jossey-Bass.

Cox, David R. 1992 Medical Genetics. The Journal of the American Medical Association 268:368.

Chatters, Linda M. 1991 Physical Health. In Life in Black America. James S. Jackson, ed. Pp. 199-220.

Newbury Park, CA: Sage Publications. Comstock, George W., and Knud J. Helsing

1976 Symptoms of Depression in Two Communities. Psychological Medicine 6:551- 563.

Cooper, Richard 1984 A Note on the Biologic Concept of Race and Its Application in Epidemiologic

Research. American Heart Journal 108:715-723. Crews, Douglas E., and James R. Bindon

1991 Ethnicity as a Taxonomic Category in Biomedical and Biosocial Research. Eth- nicity and Disease 1:42-49.

David, Richard J., and James W. Collins, Jr. 1991 Bad Outcomes in Black Babies: Race or Racism. Ethnicity and Disease 1:236-244.

Davis, Kingsley, and Wilbert E. Moore 1945 Some Principles of Stratification. American Sociological Review 10:242-249.

Dressier, William W. 1990a Lifestyle, Stress, and Blood Pressure in a Southern Black Community. Psycho-

somatic Medicine 52:182-198. 1990b Education, Lifestyle, and Arterial Blood Pressure. Journal of Psychosomatic

Research 34:515-523. 1991a Social Class, Skin Color, and Arterial Blood Pressure in Two Societies. Ethnicity

and Disease 1:60-77. 1991 b Stress and Adaptation in the Context of Culture: Depression in a Southern Black

Community. Albany: State University of New York Press. 1993 Social and Cultural Dimensions of Hypertension in Blacks: Underlying Mecha-

nisms. In Pathophysiology of Hypertension in Blacks. Janice G. Douglas and John C. S. Fray, eds. Pp. 69-89. New York: Oxford University Press.

Dressier, William W., and Lee W. Badger 1985 Epidemiology of Depressive Symptoms in Black Communities: A Comparative

Analysis. Journal of Nervous and Mental Disease 173:212-220. Duelberg, Sonja I.

1992 Preventive Health Behavior among Black and White Women in Urban and Rural Areas. Social Science & Medicine 34:191-198.

342

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY

Eisenberg, John M. 1979 Sociologic Influences on Decision-Making by Clinicians. Annals of Internal

Medicine 90:957-964. Farley, Reynolds

1985 Three Steps Forward and Two Back? Recent Changes in the Social and Economic Status of Blacks. Ethnic and Racial Studies 8:4-28.

Ford, Earl, Richard Cooper, Angel Castaner, Brian Simmons, and Maxine Mar 1989 Coronary Arteriography and Bypass Surgery among Whites and Other Racial

Groups Relative to Hospital-Based Incidence Rates for Coronary Artery Disease: Findings from NHDS. American Joural of Public Health 79:437-440.

Ford, Earl S., and Richard S. Cooper 1991 Risk Factors for Hypertension in a National Cohort Study. Hypertension 18:598-

606. Freedman, David S., David S. Strogatz, David F. Williamson, and Ronald E. Aubert

1992 Education, Race, and High-Density Lipoprotein Cholesterol among U.S. Adults. American Journal of Public Health 82:999-1006.

Harburg, E., L. Glieberman, P. Roeper, M. A. Schork, and W. J. Schull 1978 Skin Color, Ethnicity, and Blood Pressure I: Detroit Blacks. American Journal of

Public Health 68:1177-1183. Harrison, G. A., J. M. Tanner, D. R. Pilbeam, and P. T. Baker

1988 Human Biology. 3d ed. Oxford: Oxford University Press. Heymsfield, Steve, et al.

1977 Race, Education, and Prevalence of Hypertension. American Joumal of Epidemiology 106:351-361.

Hughes, Michael, and Bradley R. Hertel 1990 The Significance of Color Remains: A Study of Life Chances, Mate Selection, and

Ethnic Consciousness among Black Americans. Social Forces 68:1105-1120. Jaynes, Gerald David, and Robin M. Williams, Jr., eds.

1989 A Common Destiny: Blacks and American Society. Washington, DC: National Academy Press.

Jones, Camara Phyllis, Thomas A. LaVeist, and Marsha Lillie-Blanton 1991 "Race" in the Epidemiologic Literature: An Examination of the American Journal

of Epidemiology, 1921-1990. American Journal of Epidemiology 134:1079-1084. Kaw, Eugenia

1993 Medicalization of Racial Features: Asian American Women and Cosmetic Surgery. Medical Anthropology Quarterly (n.s.) 7:74-89.

Kemper, Theodore D., and Randall Collins 1990 Dimensions of Microinteraction. American Journal of Sociology 96:32-68.

Keith, Verna M., and Cedric Herring 1991 Skin Tone and Stratification in the Black Community. American Journal of

Sociology 97:760-778. Kessler, Ronald C., and Harold W. Neighbors

1986 A New Perspective on the Relationships among Race, Social Class, and Psycho- logical Distress. Journal of Health and Social Behavior 27:107-115.

Klag, Michael J., Paul K. Whelton, Josef Coresh, Clarence E. Grim, and Lewis H. Kuller 1991 The Association of Skin Color with Blood Pressure in U.S. Blacks with Low

Socioeconomic Status. Joural of the American Medical Association 265:599-602. Kleinman, Joel C., Lois A. Fingerhut, and Kate Prager

1991 Differences in Infant Mortality by Race, Nativity Status, and Other Maternal Characteristics. American Journal of Disease of Children 145:194-199.

343

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

MEDICAL ANTHROPOLOGY QUARTERLY

LaVeist, Thomas A. 1992 The Political Empowerment and Health Status of African-Americans: Mapping a

New Territory. American Journal of Sociology. 97:1080-1095. Leslie, Charles

1990 Scientific Racism: Reflections on Peer Review, Science and Ideology. Social Science & Medicine 31:891-905.

Littlefield, Alice, Leonard Lieberman, and Larry T. Reynolds 1982 Redefining Race: The Potential Demise of a Concept in Physical Anthropology.

Current Anthropology 23:641-655. Marmot, Michael G., Michael J. Shipley, and Geoffrey Rose

1984 Inequalities in Death-Specific Explanations of a General Patter? The Lancet 1:1003-1006.

McEwan, Peter J. M. 1990 Comment on "Scientific Racism" by Charles Leslie. Social Science & Medicine

31:911-912. Montagu, Ashley

1945 Man's Most Dangerous Myth: The Fallacy of Race. New York: Columbia Univer- sity Press.

1962 The Concept of Race. American Anthropologist 64:919-928. Murphy, Raymond

1986 Weberian Closure Theory: A Contribution to the Ongoing Assessment. British Journal of Sociology 37:21-41.

Navarro, Vicente 1991 Race or Class or Race and Class: Growing Mortality Differentials in the United

States. International Journal of Health Services 21:229-235. Nickens, Herbert

1986 Health Problems of Minority Groups: Public Health's Unfinished Agenda. Public Health Reports 101:230-231.

Neighbors, Harold W., and James S. Jackson 1986 Uninsured Risk Groups in a National Survey of Black Americans. Journal of the

National Medical Association 78:979-983. Oberman, Albert, and Gary Cutter

1984 Issues in the Natural History and Treatment of Coronary Heart Disease in Black Populations: Surgical Treatment. American Heart Journal 108:688-694.

Parkin, Frank 1979 Marxism and Class Theory: A Bourgeois Critique. London: Tavistock.

Robins, Ashley H. 1991 Biological Perspectives on Human Pigmentation. Cambridge: Cambridge Univer-

sity Press. Rushton, J. Philippe, and Anthony F Bogaert

1989 Population Differences in Susceptibility to AIDS: An Evolutionary Analysis. Social Science & Medicine 28:1211-1220.

Schwartzbaum, Judith A., John R. Wheat, and Robert W. Norton 1990 Physician Breach of Patient Confidentiality among Individuals with Human

Immunodeficiency Virus (HIV) Infection: Patterns of Decision. American Journal of Public Health 80:829-834.

Sempos, Christopher, Richard Cooper, Mary Grace Kovar, and Marilyn McMillen 1988 Divergence of the Recent Trends in Coronary Mortality for the Four Major

Race-Sex Groups in the United States. American Journal of Public Health 78:1422- 1427.

344

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions

HEALTH IN THE AFRICAN AMERICAN COMMUNITY 345

Starfield, Barbara, et al. 1991 Race, Family Income, and Low Birth Weight. American Journal of Epidemiology

134:1167-1174. Thomas, James A., and James E. Dobbins

1986 The Color Line and Social Distance in the Genesis of Essential Hypertension. Journal of the National Medical Association 78:532-536.

Wade, Peter 1993 "Race," Nature and Culture. Man 28:17-34.

Wenneker, Mark B., and Arnold M. Epstein 1989 Racial Inequalities in the Use of Procedures for Patients with Ischemic Heart

Disease in Massachusetts. Journal of the American Medical Association 261:253-257. Willhelm, Sidney M.

1986 The Economic Demise of Blacks in America. Journal of Black Studies 17:201- 254.

Young, Rosalie E, John B. Waller, Jr., and Eva Kahana 1991 Racial and Socioeconomic Aspects of Myocardial Infarction Recovery: Studying

Confounds. American Journal of Preventive Medicine 7:438-444. Zago, Marco A., Mauro S. Figueiredo, and Satie H. Ogo

1992 Bantu Bs Cluster Haplotype Predominates among Brazilian Blacks. American Journal of Physical Anthropology 88:295-298.

This content downloaded from 134.114.101.31 on Tue, 27 Oct 2015 19:54:34 UTC All use subject to JSTOR Terms and Conditions