by Evan Balaban

Racists all over the world are getting very excited about a growing debate in the scientific community over the acceptability of using racial categories in biomedical research. While there may be legitimate scientific and social issues driving some of this discussion in biomedical circles, racists, following in the footsteps of scientific creationists, are quick to turn informed scientific disagreements into validations for their belief systems.

The source of all the fuss is the biomedical community, however. It has become preoccupied with the concept of tailored medicine, the ill-defined, and still evolving idea that it is good to identify groups of people who are likely to benefit the most, or be hurt the least, by particular therapeutic treatments. For example, reasonable people agree that it is prudent not to give foods containing the amino acid phenylalanine to someone who does not have enzymes that can efficiently metabolize it. This is the case in individuals at risk for PKU, which was the earliest scientifically acceptable example of a complex human health condition linked to genetic variation. So, it would also seem prudent to consider individual biological variation with regard to drugs and other therapeutic agents, even if implementing such considerations raises difficult issues about the conflicts-of-interest, protections, rights and responsibilities entailed in gathering and using such information. While the science supporting such an endeavor is still in its infancy, one can imagine a time in the not-too-distant future when structural variations of molecules in the human body, which may produce either adverse reactions or particularly beneficial responses to compounds, can be routinely identified. Social responses to such technological innovations are typically sluggish; this is why current efforts at addressing the complex legal, social and ethical issues involved with individually-tailored medicine are so important.

Before Individualized Medicine

More contentious is what should be done in the meantime, before scientifically valid individualized medicine is realizable. Suppose you are someone who has a severely adverse reaction to a particular kind of heart medication. Current scientific knowledge about the way that medicine works is rudimentary, so it is not immediately possible to say why you react badly to it. Your younger sibling is identified with a heart condition for which the medication could be beneficial. Should they try it or not? Furthermore, you are an immigrant who grew up on the Caribbean island of St. Lucia, and you have dark skin. Could it be that any dark-skinned person coming from St. Lucia should be worried about taking this medication? What about anyone coming from St. Lucia, regardless of skin color? Should any person from the Caribbean in general be worried about it? How about people from any group that ever immigrated to St. Lucia? The information that you react badly to the drug could also be part of an opportunity to better understand the way the drug works. If information about you and others who react badly to it were to be appropriately analyzed, it could help with the identification of particular biological characteristics that correlate with the drug reaction. But what information is important and what isn't?

Generalizing this example into its larger context, how and under what conditions should group-tailored medicine be seriously pursued as a temporary stand-in for individually-tailored medicine? Group-tailored medicine attempts to assess the risks and benefits to people based on their similarities and differences compared to a reference set of data. If growing up on St. Lucia meant exposure to a common environmental or toxicological feature that permanently alters cardiac physiology, you would want to know this before taking particular drugs. Growing up dark-skinned on St. Lucia could also mean that your ancestors came from a particular place in Africa and your mother and father are both descendants of people from that place, so that you are very likely to have some molecules in your body that uniquely interact in a particular way with your heart medication. Growing up poor and malnourished can also alter the physiology of your heart. Some aspect of your personal history may have wider significance for medical decisions by other people, but we don't know what aspects these may be. Social justice considerations add another dimension to group-tailored medicine, since they provide one avenue through which minority populations can draw attention to and garner resources for particular issues that may be common in their communities, but rare on a wider geographical basis. Even though racial categories are not biological categories in the sense that "bird" and "mammal" are, should race nevertheless be considered as a potentially valid criterion for group-tailored medicine?

Of course, you could not worry about the best way to categorize human variation, but leave it to the experts. For a mere $219, DNAPrint Genomics, Inc. of Sarasota, Florida will sell you AncestryByDNA, a “genome-based test for determining your BioGeographical Ancestry admixture percentages.” The company claims to “have identified sequences of DNA that are more prevalent in people from one continent than another. Using complex statistical algorithms, the test can determine with confidence to which of the major bio-geographical ancestry groups, Sub-Saharan African, Indo European, East Asian or Native American, a person belongs, as well as the relative percentages in cases of admixed peoples.” Note that race isn't mentioned here, rather, geography is. This is because modern studies that have attempted to identify natural clusters in human genetic variation find those clusters occurring along geographical rather than traditional racial category lines (although sometimes the two can agree). As the first author of one such high-profile study subsequently told an interviewer from Science News: “In a lot of classical anthropological views of race, race is thought to be a quality predictive of a large variety of traits about a person. We found that for any given person, it's not possible to predict accurately which [variant] they have at any particular site in the genome based on their group membership.”[1,2] Skin, hair and eye color, which vary within almost all of these geographical regions, and which form the basis for many classical racial distinctions, are therefore less reliable predictors of, say, an adverse drug response, than possession of particular genetic variants independent of a person's ancestry.

Race and Group-Tailored Medicine

It has recently been argued that, in spite of the fact that there is much more variation among human individuals (about 85% of total genetic variation) than there is among traditional racial groupings of individuals (about 15% of total genetic variation), the pattern of variation across the minority of genes that differ among racial groups could nevertheless be indicative of important genetic divisions in human populations.[3] Given the fact that humans have rather low genetic diversity compared to other animal species (about half the variation found in chimps, and 1/10 that found in species of fruitflies), and the widespread movements and admixtures characteristic of early human history, it is unclear what significance these relatively weak patterns of genetic variation would have for group-tailored medicine. 

Michael Bamshad and his colleagues attempt a nuanced discussion of these issues in an article entitled “Deconstructing The Relationship Between Genetics And Race,” which appeared in a 2004 edition of Nature Reviews Genetics. They point out that existing racial categories are validated by some comparisons of genetic variation and invalidated by others, and that former studies have been biased by assigning the identity of individuals and groups a-priori.[4] In general, divisions based on genetics do not agree very well with distinctions based on morphology (including things like skin color) or on cultural features (including things like language). While future studies may lead to a better scientific consensus on the historical population structure of humans and what this says about defining present-day populations for medical purposes, this is likely to remain controversial for some time. Thus, even if people think that group-tailored medicine might be advisable under some set of circumstances, there may be no single best way of dividing up human populations for this purpose.

The Current Debate

The use of racial designations in biomedical research has persisted, under fire, for the last thirty years. The trend among users has been to agree with critics that racial categories have little biological reality, but to insist that in particular instances they may correspond with differences of medical significance that justify their use in those narrow circumstances. Yet another such cycle of criticism and response of this kind was published in The New England Journal of Medicine in 2001, when an editorial entitled “Racial profiling in medical research” attacked studies using racially-based indicators of response to heart medications; the authors of those studies defended the practice in this instance.[5] However, this exchange provoked two opinion papers from Neil Risch, Esteban Burchard and colleagues, passionately arguing for the scientific acceptance of racial categories, on the basis that any groupings based on a combination of genetic, economic and cultural differences deserve independent biomedical status, and that such status furthers the goal of social justice in medicine. [6,7]

Indeed, Risch and colleagues have begun to publish studies that attempt to map differences in the distribution of genetic marker sequences onto the categories “white, African-American, Mexican American and East Asian.”[8] This blatant conflation of group-tailored medicine with race-based medicine seems particularly counterproductive, especially since Risch, Burchard, and their colleagues fallaciously present it as a contrast between "paying attention to" and "ignoring" human diversity in biomedical research. The meaningful contrast in group-tailored medicine concerns what information is the best temporary stand-in for detailed individual information. If the value of group-based designations is defined by their temporary utility, and not the long-term implications of such systems, this at least opens the door to the possibility of a pluralistic and flexible system of referencing human variation, so that patterns of variation that do not coincide with preconceived racial categories can be detected. In contrast, a rigorous focus on racial categories may actively hamper the detection of such patterns.

We need to be collectively mobilized and challenged by Burchard and his colleagues' assertion that, “Although there are potential social costs associated with linking race or ethnic background with genetics, we believe that these potential costs are outweighed by the benefits in terms of diagnosis and research.”[9] The response on racist Web sites makes it clear that the social costs are very poorly conceived and enunciated; these advocates have also failed to adequately consider how Balkanizing medical research may reduce its effectiveness. The complex issues of protecting public interests and individual rights while promoting social justice and allowing maximum common benefits to be achieved from biomedical knowledge deserves a deeper and more thoughtful treatment before trying to make a silk purse from the sow’s ear of racial distinctions.

Evan Balaban is an assistant professor of psychology at McGill University and a member of the CRG Board.


1. Brownlee C (2005) Code of many colors: Can researchers see race in the genome? Science News 167: 232, 247.
2. Rosenberg NA, Pritchard JK, Weber JL, Cann HM, Kidd KK, Zhivotovsky LA, Feldman MW (2002) Genetic structure of human populations. Science 298: 2381-2385.
3. Edwards, AWF (2003) Human genetic diversity: Lewontin's fallacy. Bioessays 25: 798-801.
4. Bamshad M, Wooding S, Salisbury BA, Stephens JC (2004) Deconstructing the relationship between genetics and race. Nature Reviews Genetics 5: 598-609.
5. Schwartz, RS (2001) Racial profiling in medical research. New England Journal of Medicine 344: 1392-1393.
6. Risch N, Burchard E, Ziv E, Tang H. (2002) Categorization of humans in biomedical research: genes, race and disease. Genome Biology 3: comment 2007.1-2007.12.
7. Burchard EG, Ziv E, Coyle N, Gomez SL, Tang H, Karter AJ, Mountain JL, Perez-Stable EJ, Sheppard D, Risch N. (2003) The importance of race and ethnic background in biomedical research and clinical practice. New England Journal of Medicine 348: 1170-1175.
8. Jorgenson E, Tang H, Gadde M, Province M, Leppert M, Kardia S, Schork N, Cooper R, Rao DC, Boerwinkle E, Risch N (2005) Ethnicity and human genetic linkage maps. American Journal of Human Genetics76: 276-90.
9. Burchard EG, Ziv E, Coyle N, Gomez SL, Tang H, Karter AJ, Mountain JL, Perez-Stable EJ, Sheppard D, Risch N. (2003) The importance of race and ethnic background in biomedical research and clinical practice. New England Journal of Medicine 348: 1170-1175.

GeneWatch: Current Issue
Volume 30, Issue 1: CRISPR & Gene Drives
Remembering the pioneering scientist, ethicist, and founder of the Council for Responsible Genetics.
From a longer letter, "A Call to Conservation With a Conscience," initiated by
the Civil Society Working Group on Gene Drives.
GeneWatch: Archives