A compelling research agenda.

Abstract

Two articles in this issue [1, 2] and another published recently in Annals [3] remind us that race and ethnicity are powerful predictors of health-related outcomes [1-3]. The reminder is ironic at this point in the history of the United States, when fierce battles are being waged against affirmative action; the essential thrust of opponents of race-conscious strategies is that race is not a relevant way of distinguishing persons from one another. Yet these articles argue otherwise. Carter and coworkers [1] show that in all racial and ethnic minority groups, except Alaska Natives, prevalence of noninsulin-dependent diabetes mellitus is two to six times greater than that in white persons. Moormeier [3] found that black women are less likely to survive breast cancer even though they have a lower incidence of this condition. Although the lower survival rate partially results from diagnosis in the latter stages of disease, black women also have lower stage-specific survival rates. Burns and colleagues [2] show that elderly black women, including those in the highest income quintile, have a substantially lower rate of mammography use than white women, even after adjustment for the number of visits made to primary care providers. Into what context should we place these articles? One relevant context is that of history. Ten years ago, three articles related to the health of minority groups would not have been clustered in one journal. In my view, a landmark event leading to the current level of attention to minority health was the 1985 Report of the Secretary of the U.S. Department of Health and Human Services (DHHS) Task Force on Black and Minority Health. Most of the staff for the 3000-page report was drawn from the National Institutes of Health (NIH), and the report detailed the then-current state of knowledge of the health status of Asians and Pacific Islanders, Native Americans and Alaska Natives, black persons, and Hispanic persons [4]. The report had several unique aspects. First, it focused on all four major racial and ethnic minority groups, reflecting the fact that, as we approach the 21st century, blacks make up only about half of the minority population in the United States. Second, it tried to address the role of social class, whereas most previous reports used race and ethnicity as a proxy for lower socioeconomic status or examined impoverished populations, regardless of race or ethnicity. Third, the role played by the NIH gave the report both scientific rigor and the appearance of scientific rigor. Finally, subsequent creation of an Office of Minority Health in the DHHS led to the institutionalization of minority health-related research and service initiatives in the various units of the U.S. Public Health Service and in some state departments of public health. A second context is that of high-quality data from which to draw epidemiologic conclusions. Perhaps the single most important change spurred by the DHHS Task Force was increasingly fine-grained collection of data on the national level with respect to race and ethnicity and social class. State and local health agencies also improved their methods; these local data are aggregated to produce our national statistics. The collection of high-quality data ensured that we would, at a minimum, be able to measure health problems of minority groups with increasing sophistication. Until recent years, health-related data on nonblack racial and ethnic minority groups were not systematically collected. Our basic census data have only recently begun to catch up with our mosaic demographics. The other race category accounted for more than 1% of the population for the first time in the 1970s. However, Hispanic persons (who were first counted in 1980 and at that time represented 6% of the U.S. population) can be of any race. Because only a tiny percentage of Hispanic persons are self-described as black, it is likely that the vital statistics of Hispanic persons previously were aggregated with the white population. A third context in which we should understand these articles is what I call therapeutic discrimination. In the study by Burns and colleagues [2], differences in mammography rates persisted even after adjustment for primary care visits and at each income quintile. These data suggest that factors known to affect access to care, such as insurance, transportation, and literacy, may not fully explain this difference. If that is the case, these findings would join an increasingly robust body of literature showing that black persons have a pervasively different rate of use of various medical services and procedures [5-7]. Such differences persist even after adjustment for potentially confounding factors, such as insurance or disease burden. Few studies of this phenomenon have been done in minorities other than blacks. No one has suggested that the pattern of discrimination results from deliberate or conscious action. However, it is likely that the health care system contains substantial barriers for minority patients. Such barriers range from the obviouslocation and number of providersto the subtleinterpersonal relationships between health care personnel and their patients and families. These relationships could, for example, easily determine the quantity and quality of time spent with patients and would, in turn, affect whether patients would provide or receive important health-related information. A fourth context into which these articles can be placed is that of the overall status of minority health. It has been known for many decades that the death rate is more than 50% higher in black populations than in white populations. We now know that minority health status is not uniform. Among minority groups, blacks have the highest mortality rates, followed by Native Americans. Hispanics actually have slightly lower mortality rates than white persons, and Asian persons have much lower mortality than white persons. Our data on morbidity are more limited. An examination shows that patterns of specific causes of death by race or ethnicity are even more complex [8]. Poverty does not seem to drive these patterns, because poverty levels are similar among black, Hispanic, and Native American persons. Our limited information indicates that, within groups, persons with higher incomes have better health status than do persons with lower incomes. These patterns suggest that research in minority health should be viewed as a scientific and intellectual challenge and not, as is so often the case, a special interest issue. Careful study of variations of disease by race and ethnicity offers us the opportunity to deepen our understanding of how genetics, biology, culture, health care, psychology, socioeconomics, and intergenerational factors interact to produce health outcomes. Such studies could help us begin to answer important scientific and policy questions, such as those that follow. Why has the inverse relation between mortality rates and social class become stronger in the United States in recent decades [9]? What are the specific mediating mechanisms that make the inverse relation between social class and mortality rates a relatively straight line? For example, why do persons in managerial and technical occupations die at a higher rate than those in professional occupations when there are no obvious differences in their material conditions of life [10]? Why do black women have twice the rate of low-birthweight infants as white women, even when college-educated black women are compared with white women of the same educational background [11]? Why do more acculturated Mexican-American women have worse birth outcomes than female Mexicans who have recently immigrated? Do the putative healthy immigrants who decide to emigrate become less healthy after exposure to life in the United States [12]? How can we influence behaviors that often lead to such adverse outcomes as high rates of obesity, human immunodeficiency virus infection, substance abuse, and homicide? Why do Mexican-Americans have such a low rate of heart disease, despite the high prevalence of the obvious risk factors of diabetes, obesity, and cigarette smoking [13]? Forming special interest groups and vigorously campaigning on behalf of those groups seems to be part of the political and social makeup of United States culture. Specific diseases, women's health, and the health of minority groups are often the subjects of such activities. This Balkanization of advocacy, and the politics related to it, can engender cynicism in the scientific community. However, there is complex and elegant science to be done if the questions set forth here, as well as many other similar questions, are to be answered. Even if the United States were not on a trajectory to have half of its population be members of minority groups by the middle of the next century, unraveling the mysteries related to class, race, ethnicity, and health would be no less important and no less exciting.