August 1, 2011

Race and Medicine, Part LXXV

This Washington Post article illustrates that the widespread conceptual confusion over what race is can be bad for health care:
Race reemerges in debate over ‘personalized medicine’ 
By Rob Stein 
Federal examiners have rejected patents for genetic screening tests because the applicants did not explore their effectiveness for different races, adding to the debate about whether race has scientific validity in modern DNA-based medicine. 

Presumably, Patent Office staffers got a memo encouraging them to make sure that genetic tests work on minorities and aren't just being optimized for whites. But this upsets the Race Does Not Exist crowd.
Some geneticists, sociologists and bioethicists argue that “black,” “white,” “Asian” and “Hispanic” are antiquated categories that threaten to revive prejudices. Others, however, say that meaningful DNA variations can track racial lines and that ignoring them could deny many benefits of “personalized medicine,” which aims to develop tests and treatments tailored to a person’s genetic makeup. ...
Jonathan Kahn, a law professor at Hamline University in St. Paul, Minn., discovered the patent rejections when he began sifting through applications, prompted by a 2008 patent office presentation that raised the race issue. 
“Constructions of race as genetic are not only scientifically flawed, they are socially dangerous, opening the door to new forms of discrimination or the misallocation of scarce resources needed to address real health disparities,” Kahn wrote in a report in the journal Nature Biotechnology in May. ...
Similarly, in 2009, an examiner rejected a patent for a test for a propensity for prostate cancer because it did not specify the risk the variation posed among different races, Kahn found.
And in 2010, an examiner denied a patent for a test for a genetic marker for asthma and eczema because it was vetted only in whites and Asians. 
The prostate cancer and asthma rulings were reversed on appeal. But the colorectal cancer applicant narrowed the application to win approval. 
“There’s no telling how many people will just give in and use race in a way that the scientists clearly do not think is an appropriate way to use race,” Kahn said. 
Just the fact that patent applications are including such information is disturbing, he and other critics say. 
“This gives almost scientific legitimacy to the false categories of race — that somehow being white or being European is a strong category you can use in research,” said Troy Duster, who studies the racial implications of scientific research at New York University. 
For decades, demagogues — and even some scientists — argued that racial groups were genetically distinct and, in some ways, biologically inferior or superior, justifying laws barring interracial marriage and other discriminatory practices. 
Genetic predispositions — such as for sickle cell anemia, which occurs more frequently among African Americans, and Tay-Sachs disease, which is found more often in descendants of Ashkenazi Jews — clearly can pass down through generations. But as scientists developed modern tools of molecular biology, they produced ever more convincing evidence that genes vary as much among people who identify themselves as the same race as among groups segregated along traditional racial lines.

Except that they don't. Statistically, genes vary a lot within races, just as they vary a fair amount among siblings within a nuclear family, but they vary even more among individuals across races.
“What we are learning is that ancestry is really the key here,” said Charles N. Rotimi, director of the center for research on genomics and global health at the National Human Genome Research Institute.

Because ancestry and race don't have much to do with each other, I guess.
“The labels for race, at least as we currently use them, distort some of the things we want to understand in terms of ancestry.” 

Then perhaps we need for doctors to use more accurate terms. For example, Professor Kahn is up in arms about a Patent staffer who supposedly treated "Hispanic" as a racial group. This suggests that the medical profession ought to revive more genetically useful terms such as "mestizo" and "mulatto." Doctors use technical terms for lots of things that are considered inappropriate to mention in polite society, so why shouldn't they use "mestizo" and "mulatto?" It's their job, after all.
For example, although sickle cell anemia is more common among African Americans, the blood disorder is also rare in some parts of Africa and common in some predominantly Caucasian populations.

This is the kind of race-does-not-exist talking point that's more likely to confuse nonspecialist doctors than to help them make more accurate diagnoses. For the purposes of figuring out which tests to run on sick African American children, it doesn't particularly matter that sickle cell anemia "is rare in some parts of Africa" because traditional African-Americans (i.e., the descendants of American slaves) are a blended population with no ability to accurately tell a doctor something like, "My baby can't have sickle cell anemia because all 512 of my great-great-great-great-great-great-great-grandparents were from parts of Africa where sickle cell anemia is rare." The point is that if your baby is African-American, sickle cell anemia should be a concern for your pediatrician. Now, if you and your spouse just got off the plane from, say, the highlands of Ethiopia, well, maybe not, but you are the exception.

Likewise, it would be good for doctors to know that if your baby is, say, 100% Sicilian, then there's a small chance of sickle cell anemia because there was some falciparum malaria in Sicily.
The ultimate goal of genetic-based personalized medicine is to match care to each patient’s genetic makeup, Rotimi and others say. 
“You are truly going to be looking at that individual, whether black, white or Asian. It’s the individual’s genome that becomes important to their disease risk as opposed to their socially identified race or ethnicity,” said Vence L. Bonham Jr., an associate investigator at the institute, which is part of the National Institutes of Health.

But in the mean time ... Look, this individualized medical genomics thing hasn't working out as fast as people thought it would. What is progressing fast is racial genomics. Scientists are getting very good at figuring out people's racial backgrounds from their DNA.
Injecting race back into the mix carries myriad dangers, critics say. On a practical level, it may result in doctors using tests or treatments on one ethnic group and not another, denying people care based on the color of their skin.

Because less information is better when making diagnoses.
... On a more disturbing level, it could fuel racism. 
“It has the social consequence of making it seem that differences among groups are fundamentally biological,” said Barbara A. Koenig, a medical ethicist and anthropologist at the Mayo Clinic in Minnesota. “Inevitably, in our history, that leads back to the idea that one race is better than another.” 
But others say that although race is far from perfect, some genetic variations with meaningful implications for health can be much more common among certain groups. 
For example, the anti-seizure drug Tegretol produces a life-threatening skin rash more frequently among certain Asians than others; the best dose of the common blood thinner Warfarin varies by race and African Americans appear to be at an increased risk for kidney failure because they more often carry certain mutations. 
“I don’t think race/ethnicity and personalized medicine are mutually exclusive,” said Neil Risch, a professor of human genetics and epidemiology at the University of California at San Francisco. “You can call it sociological, cultural — whatever. It’s all of the above. That doesn’t mean it’s devoid of genetic meaning.”

In other words, racial medicine doesn't work in theory, but it does work with human beings. That suggests that we need a better theory.
In fact, recent analyses have indicated that many common diseases probably are caused by genetic variations in different populations, making it crucial to assemble diverse databases, researchers said in an article published online July 13 by the journal Nature. 
Two large genetic analyses published July 20 by the journals Nature and Nature Genetics found hundreds of genetic discrepancies between people of African American and European descent. And two papers published online Sunday by Nature Genetics found four unique genetic variations associated with asthma in people in Japan and people of African ancestry. Until scientists learn more about individual genetic predisposition, race provides a useful proxy, some say. 
“I think there’s a healthy debate right now about the role of race in medicine,” said Noah A. Rosenberg, a professor of biology at Stanford University.

One of the reasons that this debate has dragged on in a confused fashion for so many years, probably killing a few patients along the way, is that doctors aren't given a solid concept of race. Doctors are busy, practical people. They need the conceptual heavy lifting to be done by intellectuals, but the intellectual class has overwhelmingly failed when it comes to understanding what race is.

The problem is that because it's easy to poke holes in the crudest forms of old-fashioned American racial concepts, such as the one-drop rule, that means you can jump all the way to Race Does Not Exist, which is even cruder and stupider. What we need instead is a more sophisticated way for doctors to think about race. Fortunately, I invented* that way back in the 1990s: a racial group can most profitably be thought of as an extended family that is partly inbred. This is very close to being tautological, and, not surprisingly, lots of recent genetic data supports this insight.

The good news is that doctors shouldn't have too much trouble grasping my concept because it fits nicely as an extension of a concept they use all the time: the family medical history. The Surgeon GeneralAMA and the Mayo Clinic advocate that patients draw up a family medical history for themselves.

Race fits into the notion of a family medical history by allowing your family medical history to be extended beyond relatives whose medical histories you happen to know. Thinking of race as a partly inbred extended family means implies that statistical tendencies should also be garnered from large numbers of members of your more extended families.

The bad news is that almost nobody is explaining this concept to doctors. Thus, we see confused and confusing articles like this one.

----------------------
* I'm sure lots of other people invented it before me.

18 comments:

eh said...

...adding to the debate about whether race has scientific validity in modern DNA-based medicine.

I would expect there's not much debate about that amongst the "scientific". Elsewhere, perhaps.

RS said...

>> genes vary as much among people who identify themselves as the same race as among groups segregated along traditional racial lines.

> Except that they don't

Yeah... Lewontin's fallacy is highly misleading but literally correct, whereas this is just nonsense.

Thomas said...

There's a policy org in the Bay Area focused heavily on this issue: http://www.geneticsandsociety.org/

Ludditism and Lysenkoism live.

It's telling how many of the people nattering on about this have no background in hard science. They're lawyers or social scientists (if we can even still call anthropology a science, even anthropologists aren't sure about that anymore). It's also telling how often the consequentialist argument gets used. Maybe they should just up and say "DNA be racist."

Anonymous said...

My private DNA analysis service didn't seem to have too many PC qualms. Many statements were suffixed by '...for a person of European descent.'

Also there was a pie chart which came out reassuringly pink. I kid you not.
Gilbert P.

Inkraven said...

The concept of HBD must be gaining some traction, if in a very rudimentary way, to be showing up on Yahoo's front page:
http://news.yahoo.com/gene-discovered-raises-asthma-risk-blacks-202443130.html

If we can find the genes for asthma, certainly we can find those for intelligence, among other things

AMac said...

For a patent to be granted, an invention must meet three criteria:

1. It must have some utility.

2. It must be novel.

3. It must be non-obvious (to a skilled practitioner of the art).

4. It must meet or exceed the EEOC's Four-Fifths Rule.

Let's see, three criteria, numbered one through four, with a fraction to boot. Dang, advanced mathematics is hard!

Marlowe said...

Is it really so hard for the medical profession? How do veterinarians deal with different breeds of dogs, cows, sheep and horses? I've seen stories recently about the medical problems of pure breeds. It seems to me that the concept is employed all the time in the animal breeding world and only becomes a problem when one tries to include humans as part of the animal kingdom.

Shawn said...

I thought there was only just as much variation within racial categories because of junk genes, the genes that do not do anything.

Anonymous said...

Steve,
The whole thing is merely an exercise in futility and sophistry, literally arguing that 'black is white' - and declaring yourself the winner because you've used more big words and shouted down the opponents loudest.
Doctors are practical and intelligent people.They will mouth off enough of the prerequisite verbiage to cover themselves, but apart from that will do what they need to do.

Luke Lea said...

FWIW: Dienekes on Race and Reproductively Isolated Populations (posted April 07, 2003):

"There seems to be a confusion between race and 'inbred population', or 'reproductively isolated population'. For example, Steve Sailer defines race as follows: A racial group is an extended family that is inbred to some degree. goddless capitalist uses races as a synonym for "reproductively isolated groups".

It's true that relative inbreeding (within the race) is a necessary condition for a race to exist. But, it is not sufficient as we will show by means of examples.

Example #1. The Amish are an inbred population. Occasionally foreigners marry into the Amish population, but not often. The Amish are a population that is inbred not to some degree, but to a great degree. Are they a race? Of course not. They are simply a genetic isolate within the Caucasoid race.

The same is true for the Samaritans, a genetic isolate of about 650 members in Syria and Lebanon. The Samaritans are also a genetic isolate within the Caucasoid race.

It was recently proven that Caucasoids, Mongoloids, Negroids, Australoids and Amerindians could be distinguished from one another with an objective test. That is because these groups are inbred populations that have had enough time to evolve group specific genetic profiles. That is what makes a race.

Consider the example of a hypothetical island on which 500 Swedish and 500 Nigerian settlers intermix to form a hybrid population. After a few generations of interbreeding, the population will remain what it is: a racially mixed population. Its individuals will be random assortments of genes in the parental gene pools.

But, they won't be a race distinct from others in the sense that Caucasoids and Mongoloids are distinct. Over time, this inbred population may also develop into a race, as it evolves separately from the rest of mankind and develops its own genetic profile.

These examples should caution us against using "reproductively isolated group", or "partially inbred extended family" as synonyms for race."

Kylie said...

From the Washington Post article:

"On a practical level, it may result in doctors using tests or treatments on one ethnic group and not another, denying people care based on the color of their skin...On a more disturbing level, it could fuel racism."

Shades of General Casey's response to the Fort Hood massacre: "And what happened at Fort Hood was a tragedy, but I believe it would be an even greater tragedy if our diversity becomes a casualty here."

Right. Because living human beings are not as important as empty abstractions.

Anonymous said...

You seem to suggest that a lot of very smart people are in fact very very dumb because they deny race or at least they deny race in the sense that you believe it to be important. Could they all be wrong and you and only you are right?

Kylie said...

"You seem to suggest that a lot of very smart people are in fact very very dumb because they deny race or at least they deny race in the sense that you believe it to be important. Could they all be wrong and you and only you are right?"

It isn't a question of "you and only you", as you know--or should know. So drop the straw man already.

Steve is not the only one who holds these views re race. He is, however, one of the few publicly to admit to holding such views.

So could they all be wrong and those who hold opposing views right? Yes.

Anonymous said...

Personalized medicine. How about personalized education?

Randall H said...

' “It has the social consequence of making it seem that differences among groups are fundamentally biological,” said Barbara A. Koenig, a medical ethicist and anthropologist at the Mayo Clinic in Minnesota. “Inevitably, in our history, that leads back to the idea that one race is better than another.”

But others say that although race is far from perfect, some genetic variations with meaningful implications for health can be much more common among certain groups. '

- It is a disturbing sign of the times that 'ethicists' promote flat-earthism for political reasons even at the expense of human lives.

ben tillman said...

Yeah... Lewontin's fallacy is highly misleading but literally correct....

It is not literally correct. Here's what he said:

"The results are quite remarkable.
The mean proportion of the total species diversity that is
contained within populations is 85.4% . . . . Less than 15%
of all human genetic diversity is accounted for by differences
between human groups! Moreover, the difference between populations within a race accounts for an additional 8.3%, so that only 6.3% is accounted for by racial
classification."

It can't be literally correct because IT HAS NO LITERAL MEANING!

Reg Cæsar said...

Now, if you and your spouse just got off the plane from, say, the highlands of Ethiopia, well, maybe not, but you are the exception.

Not in my neighborhood!

When I take my kid to the local storytime, sometimes he's the only non-Ethiopian child there. This leads to strange situations. For Father's Day week, the white lady director read a typical picture book about fathers to a small crowd of rapt Ethiopian and Somali toddlers. (And our sandy-haired token, too.) All the kids knew exactly what a father was.

There are other black neighborhoods in the area in which she wouldn't dare read that book!

Reg Cæsar said...

...a medical ethicist and anthropologist at the Mayo Clinic in Minnesota.

It's "Mayo Clinic", not "the Mayo Clinic". After that boner, who can trust anything the article says?

And what do they know about race in Rochester? That has to be the largest city in America that's 99% white.

What do you call a non-white person in Rochester?
His Honor...