should science be colourblind?

catalyst, september 2007

In 2005 the American government gave its official blessing to BiDiL, a drug used to treat congestive heart failure. The decision kicked off a huge controversy because BiDiL was the first officially-sanctioned racially-targeted drug. It had proved ineffective when tested on the general population. But when given to African Americans it appeared to cut death rates from heart failure by 43 per cent. So it was marketed as a black drug, transforming at a stroke the relationship between race and medicine, and opening the way, in the words of one medical journal, to 'a new era of race-based therapeutics'. Last year, for instance, the pharmaceutical company Schering Plough organised a clinical trial for a hepatitis C drug from which African Americans were controversially excluded.

The debate about BiDiL - and the Schering Plough trial - gets to the heart of one of the most explosive issues in current medicine. Does race matter? Or should medicine be colourblind? Since 'race is biologically meaningless', argued the New England Journal of Medicine so 'instruction in medical genetics should emphasise the fallacy of race as a scientific concept and the dangers inherent in practicing race-based medicine'. A paper in Nature Genetics claimed that 'commonly used ethnic labels are both insufficient and inaccurate'.

Others begged to differ. The geneticist Neil Risch suggested that 'there is great validity in racial/ethnic self-categorisation both from the research and public policy points of view'. In a widely reported essay in the New York Times, psychiatrist Sally Satel admitted that 'In practicing medicine I am not colorblind. I always take note of my patient’s race. So do many of my colleagues'. She added that 'When it comes to practicing medicine, stereotyping often works'.

At the Washington drug clinic where she works, Satel prescribes different amounts of Prozac to black and white patients because 'clinical and pharmacological research show that blacks metabolise antidepressants more slowly than Caucasians and Asians. As a result, levels of medication can build up and make side effects more likely'.

So, who is right? As with much else in the debate on race, the answer is both and neither. Different populations exhibit distinct risk profiles to diseases and disorders but the differences are not necessarily racial. According to Sally Satel 40 per cent of African Americans metabolise anti-depressants more slowly than do Caucasians. The majority of black Americans, in other words, respond in the same manner as most Caucasians. Similarly with BiDiL. One of the dangers of marketing it as a black drug is that it may be given to American Americans who won't respond to it, but denied to non-blacks who would benefit.

We know that all but a tiny proportion of genetic variation exists within a population. Ideally, doctors would like to genetically map (or 'genotype') every individual within a population, and hence be able to predict the medical problems he or she may face and how each might respond to any particular drug. Such a procedure may well become as commonplace in the future as, say, vaccinations are today. Currently, though, individual genotyping is both practically unfeasible and too costly. Doctors, therefore, often resort to using surrogate indicators of an individual's risk profile - such as his or her race. Knowing the population from which an individual’s ancestors originally came can provide clues as to what genes that individual might be carrying. It is what Sally Satel calls a 'poor man's clue'.

Race provides medical clues because there are clearly genetic and social differences between population groups that have medical consequences. But it's a poor man's clue because the way we divide up society into different groups is not necessarily the most useful way to understand a disease or disorder - as we saw in the first part of this essay in the discussion of sickle cell anaemia.

When we ordinarily talk about human differences, we are often vague about the terms we use. We may talk about races, cultures, ethnic groups, or populations. We generally refer to whites or Europeans rather than Caucasians even though many Caucasians are neither white nor European. On the other hand, we use the term blacks and Africans interchangeably, even though there are many blacks who are not African.

If we are trying to sort out the problems of life over a pint such vagueness and confusion generally does little harm. We would expect a scientist or physician, however, to think with greater precision. In their book Sorting Things Out, Geoffrey Bowker and Susan Star point out that any scientific classification must possess three properties. First, there must be 'consistent, unique classificatory principles in operation'. So, when biologists order the living world, the rules they use to define humans (Homo sapiens) as a species are the same as the rules they use to define chimpanzees (Pan troglodytes) as a species. Second, 'categories must be mutually exclusive'. A chimpanzee cannot belong to two distinct species. And third, a classification system must be complete and able to absorb even those entities not yet identified. If we discover a new species we can slot it into the system we use to classify all other known species.

Racial classifications possess none of these properties. Races are difficult to define and there are no objective rules for deciding what constitutes a race or to what race a person belongs. People can belong to many races at the same time. And, finally, new races are not 'discovered' and slotted into the existing classification system; they are 'created' by carving up the classification system in a different way.

In the absence of a scientific classification of race, medical researchers are forced to import the racial categories we use in everyday life. The result is a striking contrast in many medical papers between the tightness and technical quality of the language when the authors are discussing genes, diseases and physiological processes and the looseness of the language about racial differences.

These problems are exacerbated by the use of self-identification as a way of assigning of racial categories. Most medical studies allow a person to define his or her own racial category. In the Schering Plough hepatitis C trial, for instance, self-identified African Americans were excluded. But those who possessed African ancestry, yet who did not identity themselves as African American, were included. This raises important questions about what the selection criteria for the trial could possibly tell us about the impact of the drug on genetic differences.

In any case, much research shows that self-identity can be unstable. One study compared data from matched birth and death certificates for infants who had died in the first year of life. While the overwhelming majority of children who were identified as black or white at birth were identified the same way at death, for other populations there were significant shifts in racial identity. Another study revealed that more than one in 10 adolescents gave a different answer when asked to define their race at school and at home.

Many fear not just that racial categories are impossible to define, but also that linking race and disease could entrench prejudice. 'A detectable genetic hallmark', the bioethicist Ernst Juengst worries 'could serve as an indelible "yellow star" marking for oppression those with indigenous ancestry.’ For instance, studies have shown that Jewish women have a higher prevalence of a gene associated with some forms of breast cancer. This could, some fear, turn breast cancer into a 'Jewish disease' while, at the same time, confirming the prejudice that Jews constitute a distinct race.

There is indeed a long history of using science to 'racialise' diseases. The designation of sickle cell anaemia as a 'black disease', for instance, was a weapon wielded by colonial administrators in Africa and racist politicians in the USA to brand black people as unhealthy and unclean. Until the 1980s the US Air Force and many commercial airliners banned black pilots with sickle cell for fear of the effects of the disease. Might not associating breast cancer with Jewish women lead to similar problems?

Such fears have led to calls for race and ethnicity to be excluded from scientific and medical research. One survey of medical research called for the establishment of special committees to scrutinise all papers and decide whether their use of racial categories is socially acceptable.

Researchers, in fact, already face a variety of restrictions in practice. The US National Human Genome Research Institute has established a databank of DNA polymorphisms, or gene variations, based on 450 samples from African, Asian, European, North and South American and Native American individuals. The samples, however, come without any information about their population of origin or about the individual who provided it. Researchers who want to use the database have to sign the scientific equivalent of the Official Secrets Act. They must promise not to try to determine the race or ethnicity of the people who contributed the DNA or even to cite papers that might have speculated on this.

Concepts of race and the use of such concepts in medical research are, as we have seen, unsatisfactory. Yet we should be wary of calls to ban such research. All scientific papers are subject to peer review. But, especially when the concept of race is so contested by scientists themselves, the idea that they must also be scrutinised for their social consequences smacks of placing them on an Inquisitorial Index. Who would sit on the scrutinising committee? Those who deny the validity of race? Or those who think that the use of racial categories is crucial for scientific and medical advances? In any case, banning scientific research will not necessarily prevent the stigmatisation of particular social groups. Do we really believe, for instance, that had references to sickle cell anaemia as a black disease been banned in the first part of the twentieth century, there would have been less discrimination against African Americans?

What really needs challenging is not research into population differences as such, but the meaning that researchers and others often impute to such differences. It is a fact that populations differ in their genetic profiles. Such differences can often be important in medical research. Suppose you want to study the impact of a new drug. It is likely that you would want to sample genetically distinct populations, to be sure that the drug does not produce an adverse reaction in people with particular genetic profiles. Since we cannot sample every population in the world, researchers often use race and ethnicity as proxies for genetic difference, as a means of creating a rough and ready approximation of worldwide genetic diversity.

Or suppose you are looking to see if a particular condition or disease - say breast cancer or high blood pressure - is linked to specific genes. One way to conduct such 'association studies' is to compare the genomes of people with and without that particular condition or disease. If the two groups are genetically similar, then it becomes easier to spot the genetic difference that might be responsible to producing breast cancer in one group and not the other. Whereas in the previous example we wanted to sample as wide a range of human variation as possible, here we want to find populations as homogenous as possible. But human populations are rarely homogenous. So race and ethnicity, once again, are often used as proxies, this time not for genetic difference, but for genetic relatedness.

There is no such thing as a 'natural' or homogenous human population. Migration; intermarriage; war and conquest; forced assimilation; the embrace of new identities; any number of social, economic, religious, and other barriers to interaction - these and many social other factors impact upon the character of a group and transform its genetic profile. Yet, human groups can act as surrogates, however imperfectly, for biological relatedness. Many of the ways in which we customarily group people socially - by race, ethnicity, nationality, religious affiliation, geographic locality and so on - are not arbitrary from a biological point of view. Members of such groups often show greater biologically relatedness than two randomly chosen individuals. Such groups have often been ghettoized by a coercive external authority, or have chosen to self-segregate from other groups. Many have a distinct history, perhaps deriving from a small founder population.

Categories such as 'African American', 'people of Asian descent' and 'Ashkenazi Jew' can be important in medical research because they are a social representation of certain aspects of genetic variation. This is why race is a 'Poor man’s clue' in medicine: not because races are natural divisions of humankind but because investigating socially defined populations provides a practical means for geneticists of dividing humans into groups that show different degrees of biological relatedness.

In order to study human genetic diversity, scientists ironically need socially defined categories of difference. The danger is that by using socially defined groups for medical or other scientific research, biologists will endow differences between such groups with greater importance than is warranted. A contingent, pragmatic division of the world into populations useful for medical research can all too easily turn into the argument that science has 'demonstrated' the reality of race.

All this points to the care needed in handling racial categories in medicine. But the fact that race is not a real biological entity does not mean that science or medicine should necessarily be colourblind. Population differences are important and can have medical consequences. It makes little sense from a biological point to view to regard these differences as 'racial'. But it also makes little sense to ignore population differences or to ban the use of racial or ethnic categories in such research. Whether or not science and medicine should be colourblind is a pragmatic question, not one rooted in scientific or political principle.