Should Scientists Use "Race" as a Proxy for Genetics?

Nature, last month (paywall):

The US Food and Drug Administration (FDA) will soon require [based on new federal legislation] researchers and companies seeking approval for late-stage clinical trials to submit a plan for ensuring diversity among trial participants…. The diversity requirement arrives in the wake of a 2022 report from the US National Academies of Sciences, Engineering and Medicine, which found that, although the representation of white women in clinical trials has improved, progress has "largely stalled" for minority racial and ethnic groups.

New York Times, yesterday:

Grappling with the deep history of racism in Western science, the National Academies of Science on Tuesday released guidelines recommending that scientists not use race as a category in genetic studies.

The guidelines, produced in response to a directive from the National Institutes of Health, noted that racial categories were poor proxies for genetic diversity and that social and environmental factors, like poverty and injustice, were often overlooked.

One possible distinction between these scenarios is that while everyone might agree that "race" is not a good proxy for genetics, it's important to have "representation" of minority groups in late-stage clinical trials so that so members of these groups have confidence in the results, while there is no similar concern in genetic studies. However, as I've pointed out before, there is no reason Americans should think that "Hispanic" or "Black" or "Asian" are medically significant classifications, but not much genetically "narrower" (because historically more genetically isolated) groups like "Ashkenazic Jews" or "Icelanders" unless the authorities told them to be worried about the former categories.

Indeed, even if one thinks that "race" correlates sufficiently with genetics to be medically significant, "Hispanics" can be of any race or any combination thereof, "Asian American" includes Austronesians, Caucasians, and East Asians, and "Black" Americans have a large admixture of European ancestry. In other words, the standard American "racial" categories that were adopted for entirely different purposes but are used by default in medical and scientific research are at odds with what racialists consider to be "race."

For more on related matters, including how FDA and NIH came to require many biomedical companies to use crude "racial" classifications to begin with, see chapter 6 of Classified.

Comments

[Rev. Arthur L. Kirkland]

Boy, this guy seems to disdain classifications.

With one glaring, life-defining exception.

[Frank Drackman]

DSM-5 302.1 Vs. DSM-5 302.2
pretty much defined your "Life" ("Life" get it? as in "Life Sentence")

[AmosArch]

The current racial classification system is nonsense of course. But race as a general concept of interrelated large populations that tend toward certain characteristics obviously exists.

The NAS like many old school establishments that want to jump on the woke train but are getting whiplash by the speed at which the Twitter bluecheck influencers that make the rules change their minds, appear to be struggling with between two mutually contradictory impulses. The former orthodoxy of the concept of race being racist and the newer orthodoxy of not using race (against the right type of people) being racist.

[Rev. Arthur L. Kirkland]

The current racial classification system is nonsense of course.

Have you informed the Republican Party and its redistricting teams?

[Frank Drackman]

Have your informed your Probation Officer?

[John Rohan]

Have you informed the Democratic party?

They have a black Congressional caucus, currently 58 members. All of them Democrats.

https://en.wikipedia.org/wiki/Congressional_Black_Caucus#Current_members

[the questing vole]

This contradictory approach is typical of the DIE/CRT approach to policy because it is all based on the old Leninist question, 'who, whom?'. If it might turn out that the thing in question makes blacks look bad (e.g. intelligence tests of any sort), then race shouldn't count and cannot be mentioned (those that mention it are racists). If it might turn out that blacks benefit in any way from the thing in question (e.g. giving them special preference in medical experiments), then race is the most important thing in the world (those who refuse to mention it are racists). Heads I win, tails you lose, sucker.

[Ed Grinberg]

Racial discrimination is awful!
(Unless it benefits me -- then it's A-OK!)

[Currentsitguy]

“poor proxies for genetic diversity and that social and environmental factors, like poverty and injustice were often overlooked.”

Is it me, or does this stink of eugenics?

Bad genes/breeding anyone?

[David Bernstein]

No, this is right. When researchers use "race" in studies, if they find statistically significant results that vary by race, they are tempted to conclude that "race" is the causal factor, when in fact (removing the tendentiousness) cultural, sociological, etc. differences that *correlate* with race but are not "racial" may be the causal factor. Thus, if the Mediterranean diet works, Greeks and Italians may live longer than Swedes, but it would be a mistake to think this is because they have Greek or Italian genes.

[Currentsitguy]

I'd take issue with that. Greeks and Italians have had countless millennia to have selectively adapted to their climate and the foods that are in abundance there. Presumably those not well adapted did not survive long enough to pass on their genes.

In the larger picture, though I think Ansel Keys' Seven Countries Study was deeply flawed in its methodology.

[Noscitur a sociis]

Greece and Italy have been inhabited by modern humans for a very countable number of millennia. And of course there’s been both significant in- and outmigration among those population as well as changes in the food, fish, and vegetables available. Plus, isn’t the main increase in life expectancy due to the “Mediterranean diet” one that accrues long past the prime reproductive years?

[EscherEnigma]

But that has nothing to do with using racial categories, and everything to do with jumping to conclusions.

If you see that "race" was a factor, you identify this correctly in your conclusion as a relationship above the level of significance, and then, if it's interesting, pursue a follow-up that tries to tease apart the relationship better. That's where you would try to identify if there was something genetic going on, or if it's social/dietary/whatever.

Jumping straight from a single study to a causal conclusion is bad science. Identifying the relationship as something interesting is not.

[Sarcastr0]

When researchers use “race” in studies, if they find statistically significant results that vary by race, they are tempted to conclude that “race” is the causal factor

Not really. Science has learned from that mistake pretty hard.

[David Bernstein]

No, they really haven’t. This comes up all the time, including on all sorts of public health, activists and officials were making arguments that members of minority groups should be first in line for Covid vaccines and treatments based on their ethnicity and race because these groups were harder hit by Covid, race. Critics noted that what they were attributing to race was really about socio economic factors correlated with which could be taken into account directly. I give other examples in the book, including treatment, algorithms that differ by race when again the notion that it was race, rather than things that happened to be correlated with race that could be measured directly, was assumed rather than shown. Unlike with Covid, the results of these algorithms is generally unfavorable to aggressive treatment of African-Americans.

[Sarcastr0]

I can't speak for activists, but it looks like you're conflating sociological studies, where race is often a causal factor, with medical studies, where race rarely is.

The Covid treatment thing was not clinically driven, it was public health driven, which is a discipline that includes social cues quite a bit.

For instance, there was a study about the way hospitals treated black folks' self-reported pain levels, and it found a tendency to underrate them.

Maybe I misread you and you were going after studies that find an externally-driven disparity caused by race (those studies seem a largely legit avenue of inquiry to me), but it looked like you were implying that there was a eugenics-like assumption that races are causally different biologically, and I don't think that's supported.

[EscherEnigma]

One possible distinction between these scenarios [...]

Is unnecessary.

The first blurb is saying that while progress has been made, test subjects are still too homogenous.

The second blurb is saying that (A) race is a bad proxy for genetics, and (B) there are other things that need to be looked at more.

Reading some kind of conflict into those is unnecessary, and both are saying the same thing: studies are improved with diversity.

As a side note? Professional journalists who focus on science reporting are often shit at it. Lawyers who pretend to be journalists taking a pot-shot at science are shit at it too.

[SRG]

Skin colour is certainly a poor proxy for race but, for example, one should certainly increase the number of people of recent African ancestry in medical studies because of the greater genetic variation in African populations, and hence increasing the likelihood that any difference in response owing to genetic factors will be spotted in the trial.

[David Bernstein]

The goal of increasing genetic diversity seems like a sound one, but there is nothing in the way that race is reifiec in the relevant rules that is geared towards doing so. Instead, the government requires researchers to use statistical classifications that when placed into federal law came with the explicit caveat that they were not genetic or anthropological classifications! In fairness to Congress, they originally told federal agencies to require raise, but didn’t specify how to do it. On the other hand, however, in practice doing anything, but just using the familiar Classifications rooted in racist American history would have created an unbelievable political fire storm, whereas just using black, Hispanic, etc. Stoped opposition from the biomedical research community, but not from the public and even won the support of the American medical association.

[Stephen Lathrop]

However, as I've pointed out before, there is no reason Americans should think that "Hispanic" or "Black" or "Asian" are medically significant classifications,

Except for this reason. You may be talking about applied medical science. You may expect the medical providers to use their own perceptions of race to adjust medical treatment. You might hope that even if genetic science proved that an unsound method in theory, the research would approximate the method which would actually be used, and might thus deliver practical assistance on that basis.

[StevenF]

If you don't use genetics to DEFINE race, whatever you are doing isn't science.

[the questing vole]

The fact that you don't 'get' something is pretty much an hourly occurrence, isn't it?

As I said above and you just confirmed, the primary policy question for racial grifters like yourself is, does it make my side feel better. Truth has nothing to do with it.

[David Bernstein]

It's not the least bit obvious to me that if people have "deep skepticism" about the medical field because of historical discrimination, that the way to counteract that skepticism is to reify racial classifications that have their underlying roots in historical racism, and have no scientific basis, in scientific research. Even if that was obvious, there was no specific historical discrimination of the sort that was experienced by African Americans in medicine against the other nonwhite groups classified. What historical memory of a hostile medical establishment to Asian Indians, the vast majority of whom arrived here since the 1970s, have? And even aside from that, one would have to consider whether reifying unscientific racial classifications in science, as if they are scientifically salient, will have other negative consequences, as nonsensical racial classification tends to. People have a natural psychological tendency, after all, to think that if the study they heard about on NPR broke down data by whites and black people separately, that there must be some "real" differences between "the races."

[the questing vole]

I do.

[the questing vole]

It's a great deal more pleasant than the real one.

[the questing vole]

But I don't ask the government to acknowledge, support, and promote my feelings.

If you want feelings, though, here they are: https://www.youtube.com/watch?v=g6vI0uE9iqM

[Sarcastr0]

Saying 'I hear and understand' is a better bedside manner than hitting them with this post of yours.

Sometimes the best thing is to meet people where they are, even if technically if you dig back in history yada yada yada.

[David Bernstein]

There is little if any grassroots demand for this stuff. It started bc when men’s health activists pointed out that women are actually different than men, and Congress should therefore require more female representation in studies. Then some politicians and interest groups at the last minute, jumped on the race bandwagon. Until that point, in the early 1990s, civil rights groups have been almost universally against race, being a factor in medical studies for all the obvious reasons. But the populist Appeal of “all groups should be represented” won out, even though in practice all groups means using scientifically nonsensical classifications rooted in racism.

[David Bernstein]

I agree, but I think that positive step would be to eliminate the bogus use of race entirely in medical studies. That would actually show that we learned our lessons.

[David Bernstein]

Wimen’s health advocates, not when men’s

[Sarcastr0]

I get drinks sometimes with some folks in the NIH whose job it is to make sure that drug studies include different demographics just in case there’s a biological component.

Are you saying that’s not a good idea? There were some famous examples that drive that push. Ambien was one, I think. Studies found that dosages were off for women across the board, and the side effects were therefore higher.

But this seems different than the original conversation, which was about genetics, and this more race than gender. Racial diversity has been added to drug trials as well, I think just following on the findings about women. But the main thing I was talking about was the difference between bedside manner accuracy and scholarly accuracy.