Race and the Clinic: Good Science?

Humans have long embraced the idea of grouping and naming people who have distinct, genetically determined physical characteristics, like almond-shaped eyes or different skin color. It made sense, from a social standpoint (think safety, politics, and business) to align one's self with kin. However, studying race from a biological point of view, in the hopes of learning about specific diseases or developing new drugs, is a different matter altogether. "Race is generally not a useful consideration

Feb 18, 2002
Ricki Lewis
Humans have long embraced the idea of grouping and naming people who have distinct, genetically determined physical characteristics, like almond-shaped eyes or different skin color. It made sense, from a social standpoint (think safety, politics, and business) to align one's self with kin. However, studying race from a biological point of view, in the hopes of learning about specific diseases or developing new drugs, is a different matter altogether.

"Race is generally not a useful consideration in a clinical decision," says medical ethicist Susan Setta, professor of philosophy and religion at Northeastern University in Boston. "It is sometimes used as a substitute for considerations of lifestyle, which are often essential components of clinical decision-making." Harold Freeman, director of the National Cancer Institute's Center to Reduce Health Disparities, said at a recent meeting, "Race disappears when you look at the human genome."

But scientists know that they cannot ignore the clinical data that show, for example, that African Americans die at a higher rate from coronary heart disease than do whites. Moreover, population genetics has long shown that certain single-gene disorders are more prevalent in some populations, such as Tay-Sachs disease among Ashkenazi Jews. Polygenic disorders also tend to be more common in some population groups. So, it isn't surprising that epidemiological studies show that certain drugs have a better efficacy rate in some groups than others. The controversy arises over what to do with this type of information. For some scientists, the question now is, "Do different ways exist to organize people?" So far, researchers are exploring a few ideas, including studying the human brain and identifying gene combinations that control drug responses. Says Freeman, "Race doesn't exist, but yet it does."

Race and Geneticists

Geneticists have long considered racial designations to be more sociological than biological. In 1999, the American Anthropological Association concurred with the statement that "... human populations are not unambiguous, clearly demarcated, biologically distinct groups.... Throughout history, whenever different groups have come into contact, they have interbred. Any attempt to establish lines of division among biological populations is both arbitrary and subjective."1 The Human Genome Project has so far revealed that what people consider racial differences comprise only 0.01% of the body's estimated 35,000 genes.

Human patterns of migration and mating have unevenly distributed what little genetic variation the global human population has, and people have classified each other based on the most visible of those variable traits. The consequences of racial classification are numerous and interconnected, with differential availability of health care services being just one. Rick Kittles, codirector of molecular genetics at the National Human Genome Center at Howard University, calls it "the enigma of the race and health disparity. To what extent are these differences due to the unequal distribution of resources, or the result of inherent characteristics ... of individuals defined by race or ethnicity?" Adds Robin Andreasen, assistant professor of philosophy, University of Delaware, Newark, "The conditions required to maintain racial distinctness—reasonable reproductive isolation—may no longer exist. Hence, races may be on their way out." She defines races by kinship, rather than physical similarity.

A Short History of Race

Categorizing people by skin color is probably recent; hunter-gatherers didn't wander far enough to encounter peoples of different colors. "The whole idea of race appeared after the age of exploration by ships," says Leda Cosmides, professor of psychology, Center for Evolutionary Psychology, University of California, Santa Barbara. "People at very different geographical points looked different enough that it made it seem like humans came in different morphs, or complexes of traits."

Courtesy of Leda Losmides

Leda Cosmides and John Tooby

In the United States, slavery cemented the importance of skin color. "The 1890 US Census distinguished mulatto (half black and half white) from quadro (one-fourth black) from octaroon (one-eighth black). And there was the 'one drop' rule. If you had one black ancestor, you were black," explained Freeman. The 2000 Census for the first time allowed people to check more than one of six races, including "some other race," producing 63 combinations. About 6.8 million individuals (2.4% of the US population) consider themselves multiracial.

The Santa Barbara group argues that classification by skin color makes little evolutionary sense, compared to natural selection for such features as age and sex, which directly relate to reproduction abilities. Cosmides and colleagues Robert Kurzban and John Tooby wanted to know if people grouped each other by color to identify coalitions. They showed photographs of individuals, who were paired with a sentence, to student volunteers. Then, the sentences were shown out of order, and the volunteer was asked to identify each speaker. "Mistakes reveal encoding," says Cosmides. "A person who had categorized each individual by race would make more within-race errors, confusing a black person with another black person, than between-race errors."

Next, participants were shown two "teams," each half black and half white. The sentences reflected conflict. When all team members were dressed alike, the volunteers encoded individuals by skin color. But when one team wore gray shirts and the other yellow, encoding was by shirt color, which Cosmides calls "a visual marker of coalition, even though it is arbitrary. And people encoded coalition more strongly than they encoded race in the other experiment." Conclude the researchers, "Despite a lifetime's experience of race as a predictor of social alliance, less than four minutes of exposure to an alternate social world was enough to deflate the tendency to categorize by race."2 (Perhaps the experiments validate the "shirts and skins" approach to distinguish basketball teams.)

Racial Differences in the Clinic

Racism may have arisen as a way to detect ancient alliances, but today the implications of classification by color resonate in the clinic.3 According to the Office of Research on Minority Health at the National Heart, Lung and Blood Institute, age-adjusted deaths per 100,000 people from cardiovascular disease number 172.8 for whites,180.4 for all races, and 265.3 for blacks. The elevation is similar for heart disease, stroke, and congestive heart failure.

"It is well documented that African Americans have higher prevalence and a greater clustering of risk factors that contribute to heart disease-related deaths and disability, including hypertension, dyslipidemia, obesity, diabetes, and physical inactivity," says Zhi-Jie Zheng, an epidemiologist in the cardiovascular health branch at the Centers for Disease Control and Prevention in Atlanta. At the American Heart Association's annual meeting in November, Zheng reported that the decline in deaths from coronary artery disease among blacks is considerably slower than in whites. "The racial disparity in the death rate decline is influenced by a number of elements, including social and environmental conditions, economic resources, geographic and institutional and policy factors," he adds.

Zheng considers race based on skin color as strictly a social label. Several other recent studies approach the race-health disparity enigma from a genetic viewpoint.4 That is, segregation into races may have also grouped drug metabolism gene variants. "The descriptor of African American allows for the clustering of a given phenotype of illness that may define a genotype," says Clyde Yancy, the Dallas Heart Ball Chair in Cardiac Research, University of Texas Southwestern Medical Center, Dallas. "With further studies, that clustering can be used to probe the general population."

The US Carvedilol Heart Failure Trial tested the beta-blocker Carvedilol in 217 blacks and 877 whites, with each participant given either the drug or placebo for 15 months.5 Researchers found that ventricular ejection fraction and clinical status were about equal in both races, with a slightly higher response rate among blacks. In another study, the Left Ventricular Dysfunction trial, each of 800 blacks was matched to four whites for sex, age, and treatment with the ACE inhibitor enalapril or placebo. Participants were assessed for left ventricular ejection fraction and monitored for three years.6 Unlike the case for Carvedilol, enalapril significantly lowered blood pressure and risk of hospitalization for heart failure in whites, but not in blacks. Although the authors concluded that race is a surrogate marker for biological differences in drug response, they also suggest further research into which drugs to use specifically for black patients.

How to extrapolate population data on drug response to individual patient care is not clear, because the numbers are fuzzy. "None of the identifiers in the literature are absolutely inclusive within an entire group," says Yancy, who headed the Carvedilol trial. "It would be a clinical disservice to limit consideration to race. We need to use clustering of phenotypes to identify that part of the genome that explains individual differences."

Eric Young/Washington University School of Medicine

Identifying Drug Metabolism Genotypes

The area of pharmacogenetics probes variability in drug response due to individual genes. Variants of the genes that encode cytochrome P-450 enzymes, for example, affect the way that the body metabolizes codeine, tricyclic antidepressants, beta-blockers, and three-dozen other drugs.7 Some P-450 variants are more common in certain population groups.

Another gene that affects drug response, MDR1, encodes the protein portion of P-glycoprotein, a drug pump on intestinal lining cells and T cells.8 In times past, natural selection favored overactive MDR1 genes whose products ejected enterotoxins common in Africa. But yesterday's poison protection is today's drug resistance; some P-glycoproteins jettison drugs to treat cancer, AIDS, or tissue rejection. "High expression of P-glycoprotein in CD4 lymphocytes may lead to a poorer response to HIV protease inhibitors," says pediatrician Matthias Schwab of the Dr. Margarete Fischer-Bosch Institute for Clinical Pharmacology in Stuttgart, Germany. "This may be of great importance in populations with a high frequency of the C3435T polymorphism" associated with overexpression. To test this hypothesis, Schwab's group genotyped 172 West Africans, 41 African Americans, 537 whites, and 50 Japanese, all healthy, for the C3435T variant of MDR1. They found it in 83% of West Africans, 61% of African Americans, 26% of whites, and 34% of Japanese. People with this genotype may need higher doses of protease inhibitors.

On the pharmacogenomics front, SNP (single nucleotide polymorphism) profiles can foretell how well certain painkillers, blood thinners, and chemotherapies will work. SNPs are the sites where DNA sequence differs among individuals. To confirm that it makes more biological sense to consider drug-metabolizing genes rather than skin color in drug choice, James Wilson, David Goldstein, and colleagues at University College, London, compared 23 markers for such genes among 354 people representing eight classically defined races: white (Norwegian, Ashkenazi Jews, Armenians), black (Bantu, Ethiopian, and Afro-Caribbean), and Asian (Chinese and New Guinean).9 Using a technique called hierarchical cluster analysis, they found that the genetic markers form four natural groupings that do not correspond to any of the appearance-defined categories. Concludes Howard L. McLeod, a professor of medicine at the Washington University School of Medicine in Missouri, "There is no clear link between skin pigment and drug metabolism genes. Skin pigment is a lousy surrogate for drug metabolism status or most any aspect of human physiology."10

At the same time that researchers are correlating SNPs to disease risks and drug responses, classification by skin color continues to influence medical decision-making. At least one company, NitroMed Inc. in Bedford, Mass., is planning to market a heart disease drug specifically to African Americans, according to its Web site (www.nitromed.com).

With researchers demonstrating that the human brain can easily classify people in varied ways, and human genome data revealing the gene combinations that control response to drugs, the concept of race, particularly as it relates to clinical decision-making, is at the very least headed for reconsideration. Says McLeod, "There are already a lot of SNP data to allow this to happen. The rate-limiting step is the data showing that a particular marker or set of markers is informative enough to individualize medical decisions."

Ricki Lewis (rickilewis@nasw.org) is a contributing editor.

1. American Anthropological Association, "American Anthropological Association statement on race," American Anthropologist, 100:712-3, 1998.

2. R. Kurzban et al., "Can race be erased? Coalitional computation and social categorization," Proceedings of the National Academy of Sciences, 98:15387-92, 2001.

3. A.M. Epstein, J.Z. Ayanian, "Racial disparities in medical care," New England Journal of Medicine (NEJM), 344:1471-3, May 10, 2001.

4. R.S. Schwartz, "Racial profiling in medical research," NEJM, 344:1392-3, May 3, 2001.

5. C.W. Yancy et al., "Race and the response to adrenergic blockade with Carvedilol in patients with chronic heart failure," NEJM, 344:1358-65, May 3, 2001.

6. D.V. Exner et al., "Lesser response to angiotensin-converting enzyme inhibitor therapy in black as compared with white patients with left ventricular dysfunction," NEJM, 344:1351-7, May 3, 2001.

7. A.J.J. Wood, "Racial differences in the response to drugs - pointers to genetic differences," NEJM, 344:1393-6, May 3, 2001.

8.E. Schaeffeler et al., "Frequency of C3435T polymorphism of MDR1 gene in African people," The Lancet, 358:383-4, April 4, 2001.

9. J.F. Wilson et al., "Population genetic structure of variable drug response," Nature Genetics, 29:265-9, November 2001.

10. H.L. McLeod, "Pharmacogenetics: more than skin deep," Nature Genetics, 29:247-8, November 2001.

Supplemental Materials