The Problem of Race in Medicine

Philosophy of the Social Sciences

Volumn 31, Issue 1, 2001, Pages 20-39

  Root, Michael ^a  

^a UNIVERSITY OF MINNESOTA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 84937338798     PISSN: 00483931     EISSN: None     Source Type: Journal    
DOI: 10.1177/004839310103100102     Document Type: Article

References (51)

1. The New York Times, 21 January 1999, A11.
2. Sho-Ban News: Ethnic News Watch, 6 March 1998, 1.
3. C. P. Jones, T. A. LaVeist, and M. Little-Blanton, “‘Race’ in the Epidemiologic Literature: An Examination of the American Journal of Epidemiology, 1921-1990,” American Journal of Epidemiology 134 (1991): 1079-84.
4. Oliver Fein, “The Influence of Social Class on Health Status: American and British Research on Health Inequalities,” Journal of General Internal Medicine 10 (1995): 577-86.
5. Carol J. Rowland Hogue and Martha A. Hargroves, “Class, Race and Infant Mortality in the United States,” American Journal of Public Health 83 (1993): 9-12.
6. Richard Cooper, “Health and the Social Status of Blacks in the United States,” Annals of Epidemiology 3 (1993): 137-44.
7. For example, see H. W. Matthews, “Racial, Ethnic and Gender Differences in Response to Medicines,” Drug Metabolism and Drug Interactions 12 (1995): 77-91; and Milo Gibaldi, “Ethnic Difference in the Assessment and Treatment of Disease,” Perspectives in Pharmacotherapy 13 (1993): 170-76.
8. For example, see R. S. Cooper and David R. Cooper, “The Biological Concept of Race and Its Application to Public Health and Epidemiology,” Journal of Health Politics, Policy and Law 11 (1986): 97-116; J. S. Kaufman, R. S. Cooper, and D. L. McGee, “Socioeconomic Status and Health in Blacks and Whites: The Problem of Residual Confounding and the Resiliency of Race,” Epidemiology 8 (1997): 621-28; and T. A. LaVeist, “WhyWe Should Continue to Study Race… ButDoaBetter Job:AnEssay on Race, Racism and Health,” Ethnicity and Disease 6 (1996): 21-29.
9. This leaves open how the race of the mother was determined; in some cases, it might be self-identification, but the National Center of Health Statistics (NCHS) assumes that however it was determined, the result matches the way the NCHS's rule assigns race to the infant, that is, on the basis of descent.
10. Robert A. Hahn, “The State of Federal Health Statistics on Racial and Ethnic Groups,” Journal of the American Medical Association 267 (1992): 268-71.
11. “Race and Ethnic Standards for Federal Statistics and Administrative Reporting,” in Statistical Policy Handbook (Washington, DC: Office of Federal Statistical Policy and Standards, U.S. Department ofCommerce, 1998). Directive 15 divides the United States into five racial categories: black or African American, white, Asian, Native Hawaiian or other Pacific Islander, and American Indian or Alaskan Native.
12. The standard of validity of archival data, the data most often used in epidemiology, is similar to the standard for test data, the data used in psychology and education. Test data, a set of scores X on a test T, are valid only if X strongly correlates with a criterial variable Y. Where T is the National Center of Health Statistics's method of assigning race,Xthe set of racial assignments based on T, andYthe criterion of race (the criterion in use within the community),Xis valid only ifXstrongly correlates withY.
13. L. Cavalli-Sforza, Paoli Meonozzi, and Alberto Piazza, The History and Geography of Human Genes (Princeton, NJ: Princeton University Press, 1994), 79-80.
14. There is some disagreement over the existence or extent of any statistical variation but no disagreement that the variations, such as they are, are independent and do not strongly correlate. Masatoshi Nei and Arun K. Roychoudhury, in “Genetic Relationship and Evolution ofHumanRaces,” Evolutionary Biology 14 (1982): 1-59, argue, for example, that there are some statistical variations, whereas Richard Lewontin, in “The Apportionment of Human Diversity,” Evolutionary Biology 6 (1972): 381-98, questions the variations; but neither maintains that there are genetic differences that cluster together with a significant degree of predictability. Moreover, the studies of variation draw their samples from groups that are believed to be most reproductively isolated in order to maximize intergroup variation, but the groups that are customarily classified as different races are not isolated, and as a result the samples on which the population studies are based are not representative of the groups we count in the United States as different races. In other words, the methods Cavalli-Sforza, Meonozzi, and Piazza and Nei and Roychoudhury use to compare gene frequencies between races assume that the populations sampled (e.g., “Caucasians” and “Negroids”) correspond to the groups classified as distinct races in the countries that classify by race (e.g., “blacks” and “whites” in the United States), but they do not. The samples are from genetically isolated populations (that is why they were chosen). The five official races in the United States are not so isolated, and consequently blacks and whites in the United States are not the Caucasians and Negroids of Cavalli-Sforza, Meonozzi, and Piazza's and Nei and Roychoudhury's studies.
15. The color of human skin is determined by the amounts of three chemicals whose presence is determined by genes.
16. J. S. Jones, “How Different Are Human Races,” Nature 293 (1981): 188-90.
17. Julian S. Huxley and Alfred Cort Hadden, We Europeans: A Survey of “Racial” Problems (New York: Harper, 1936), 82.
18. David Hollinger, Post-Ethnic America: Beyond Multiculturalism (New York: Basic Books, 1995).
19. Henry Louis Gates, Jr., “Writing, ‘Race,’ and the Difference It Makes,” in Loose Canons: Notes on the Culture Wars (Oxford: Oxford University Press, 1992).
20. Anthony Appiah, “Illusions of Race,” in InMyFather's House: Africa in the Philosophy of Culture (Oxford: Oxford University Press, 1992).
21. Ashley Montague, The Concept of Race (New York: Free Press, 1964).
22. Orlando Patterson, “The Race Trap,” The New York Times, 11 July 1997, op-ed.
23. Some social scientists maintain that the categories “male” and “female” are themselves due (at least in part) to us (to how we choose to label each other) rather than to nature. See, for example, Suzanne Kessler andWendy McKenna, Gender: An Ethnomethodological Approach (NewYork:Wiley, 1978).Onsuch a view, the traditional distinction between sex (nature) and gender (convention) is spurious. When I say that sex is due to nature, I amreferring to the reproductive differences biologists find when studying members of sexually reproducing species (e.g., differences in their reproductive organs, gametes, or hormone levels).
24. For a history of the concept of race, see Hannah Franziska Augustein, ed., Race: The Origins of an Idea, 1760-1850 (Bristol, UK: Thoemmes Press, 1996).
25. One might wonder whether such a category would deserve to be called race, for racial classification has been based on descent rather than consent for so long that speaking of naturalized whites or black sounds otiose. But here I believe in family resemblance; if our old and new talk of race overlap enough, our new categories, although different, would still be race.
26. That is, if a child were labeled “mixed race” if she had at least one black and one white ancestor, few children today would be black or white; most would be mixed race.
27. However, in using the data to compile racial statistics, the Bureau of the Census will only recognize one checked box, and if the box is “black,” the person will be officially black. Given that the data are used to monitor racial discrimination and a person who checks “black” is likely to experience racial discrimination, the Bureau of the Census’ decision to only recognize one box is reasonable.
28. To the extent that black in the United States today is still defined by the one-drop rule, the National Center of Health Statistics's definition in 1988 traces the U.S. racial divide better than the current one does.
29. Sampling can provide some estimate of the error. That is, the epidemiologist can study a sample of the respondents who check the “black” box and determine what percentage is black by a customary criterion, but, of course, she would first have to choose a customary criterion and know that her sample was representative.
30. J. S. Kaufman and R. S. Cooper, “Epidemiological Research on Minority Health: In Search of a Hypothesis,” Public Health Reports 110 (1995): 662-66.
31. Anthony Polednak, Racial and Ethnic Differences in Disease (Oxford: Oxford University Press, 1989), 6.
32. Their method of reasoning is old. John Stuart Mill, in On the Logic of the Moral Sciences (Indianapolis, IN: Bobbs-Merrill, 1965), describes the method this way: “Subduct from any phenomenon such part as is known by previous inductions to be the effect of certain antecedents, and the residue of the phenomenon is the effect of the remaining antecedents” (p. 165).
33. J. C. Kleinman and S. S. Kessel, “Racial Differences in Low BirthWeight: Trends and Risk Factors,” New England Journal of Medicine 317 (1987): 749-52.
34. R. J. David and J.W. Collins, “Different BirthWeight among Infants of U.S. Born Blacks, African-Born Blacks and U.S. Born Whites,” New England Journal of Medicine 337 (1997): 1209-14.
35. A. F. Naylor and N. C. Myrianthopoulos, “The Relation of Ethnic and Selected Socio-Economic Factors to Human Birth-Weight,” Annals of Human Genetics 31 (1967): 71-83.
36. H.W. Foster, “The Enigma of Low BirthWeight and Race,” New England Journal of Medicine 337 (1997): 1232-33.
37. David and Collins, “Different Birth Weight among Infants,” 1209.
38. The correlation coefficient between X and Y can be high even though X is not the cause of Y if there is a confounding factor Z (a hidden variable) that causes Y and strongly correlates with X. In the birth weight studies, Z is the intergenerational variable.
39. Craig St. John and Nancy A. Bates, “Racial Composition and Neighborhood Evaluation,” Social Science Research 19 (1990): 47-61.
40. Melbourne Tapper, In the Blood: Sickle Cell Anemia and the Politics of Race (Philadelphia: University of Pennsylvania Press, 1999), 12-54.
41. I can sharpen the point by replacing race with occupation. Assume (1)-(3), as before, but, instead of G varying with race, G varies with occupation (e.g., the frequency ofGis greater among lawyers than doctors).We should not conclude that occupation is a genetic factor in disease D, for the link between G and occupation is coincidental; but the relationship between the sickle-cell gene and race is coincidental as well.
42. Studies of the rates of nicotine absorption in black and white smokers, for example, find that black smokers have higher concentrations of cortinine in their blood than white smokers after controlling for confounding variables such as weight, differences in smoking habits (e.g., depth of inhalation and length of cigarette), and smoke exposure at home and work. Cortinine is a by-product of the breakdown of nicotine in the body. The higher cortinine levels indicate that black smokers absorbmore nicotine from the same smoke exposure than whites do. The conclusion drawn is that the differences are due to genetic differences between the races, that a genetic trait linked to protection from nicotine addiction may be missing from blacks. See Eliseo Perez-Stable, Brenda Herrera, Peyton Jacob, III, and Neal Benowitz, “Nicotine Metabolism and Intake in Black and White Smokers,” Journal of the American Medical Association 280 (1998): 152-56. Such findings are often cited to explain why lung cancer rates are higher among black than white smokers.
43. The conventions that make one man white and another black, as I explained earlier, are site specific, and as a result a man who is black in one place or at one time can be white in or at another.
44. At some sites, the races might match clines (geographically separate populations with somewhat different gene pools), whereas at others, they might not match any genetically distinguishable groups at all.
45. The differences in gene frequencies between the races found by Cavalli-Sforza, Meonozzi, and Piazza in The History and Geography of Human Genes and Nei and Roychoudhury in “Genetic Relationship and Evolution of Human Races” are not responsible for the differences studied by the epidemiologist or the pharmacologist, and the populations in which the differences in gene frequencies were found do not match the groups labeled “black” and “white” in these studies.
46. Polednak, Racial and Ethnic Differences in Disease, 4.
47. There is more to race than racial discrimination, but when looking at the risks of race, discrimination should be our focus. Sartre maintains, in Anti-Semite and Jew, translated by George Becker (New York: Schocken, 1965), 67-69, that the anti-Semite makes the Jew, that is, provides the bond that draws a disparate people together. Sartre is wrong, but when trying to explainwhyrates of Jewish morbidity and mortality were so high in Germany after 1935, epidemiologists should label x a Jew if, under the Nuremberg Laws, x carried that label.
48. Epidemiologists would do better to assign different labels to (1) children with a black father and white mother, (2) children with a black mother and white father, and (3) children with both a black father and black mother and compare how rates of mortality and morbidity differ between (1), (2), and (3).
49. I owe this to Alison Wylie.
50. Naylor and Myrianthopoulos, in “The Relation of Ethnic and Selected Socio- Economic Factors to Human Birth-Weight,” conclude that the unexplained variance in birth weight was due a genetic difference between blacks and whites because they did not understand how the effects of past acts of overt discrimination carry forward and years later can adversely affect the health of a grandchild or even great grandchild and, as a result, how institutional discrimination could explain the variance between black and white infants.
51. Achild's risk of sickle-cell disease does not increase if her mother is black (if the child is black according to the National Center of Health Statistics) unless she has a black father.