National Institutes of Health violates academic freedom, restricts dissemination of taxpayer-funded research

October 21, 2022 • 9:20 am

This article just appeared in the (conservative) City Journal, and is written by James Lee, a behavioral geneticist at the University at Minnesota.  What Lee reports made steam issue from under my collar, for he claims that the National Institutes of Health, a U.S. government science institute, has a huge genetics and “trait” database of several million Americans. The genetic data appear to be thorough, based on genome scans, and the traits associated with each person’s genome include education, ethnicity (“race”), intelligence, income, and occupation. You can imagine how rich that dataset is for mining. And yet the NIH is restricting scientists’ access to the data to projects it apparently considers ideologically kosher.

Remember that the NIH is completely funded by the American taxpayers, so those data were accumulated with our money. To me, this means that any researcher with a valid project should have access to the data. But apparently some projects are more valid than others.

Click to read.

Here’s Lee’s description of the hard time geneticists have in getting the data when their project sounds “iffy”, and by that I mean any project that has to do with heredity and intelligence (presumably IQ or a similar measure). Note that none of the attempts to get the data have been to do projects on ethnicity and IQ, which of course are considered taboo by many (readers may want to either echo or refute that taboo). Check out the second paragraph of the excerpt below, which I’ve put in bold.

American geneticists now face an even more drastic form of censorship: exclusion from access to the data necessary to conduct analyses, let alone publish results. Case in point: the National Institutes of Health now withholds access to an important database if it thinks a scientist’s research may wander into forbidden territory. The source at issue, the Database of Genotypes and Phenotypes (dbGaP), is an exceptional tool, combining genome scans of several million individuals with extensive data about health, education, occupation, and income. It is indispensable for research on how genes and environments combine to affect human traits. No other widely accessible American database comes close in terms of scientific utility.

My colleagues at other universities and I have run into problems involving applications to study the relationships among intelligence, education, and health outcomes. Sometimes, NIH denies access to some of the attributes that I have just mentioned, on the grounds that studying their genetic basis is “stigmatizing.” Sometimes, it demands updates about ongoing research, with the implied threat that it could withdraw usage if it doesn’t receive satisfactory answers. In some cases, NIH has retroactively withdrawn access for research it had previously approved.

Note that none of the studies I am referring to include inquiries into race or sex differences. Apparently, NIH is clamping down on a broad range of attempts to explore the relationship between genetics and intelligence.

It’s hard to believe that the NIH is restricting data that might be used to show any relationship between genes and intelligence, even within one ethnic group.  We already have data on genes implicated in academic achievement (which is correlated with IQ); those data are a big part of Kathryn Paige Harden‘s book The Genetic Lottery: Why DNA Matters for Social Equality, a book I reviewed for the Washington Post and also discussed on this website. As I recall, Harden’s genome-wide association study found nearly 1300 genomic sites associated with variation in academic achievement among the American European (“white”) population. Intriguingly, many of those sites were active in the brain. That in itself is of considerable interest, though Harden’s claim that this variation would help us create “level playing fields” for secondary-school students seemed unjustified.  But even finding genes associated with intelligence would tell us a lot about the developmental genetics of an important human trait.

Lee also explains why the NIH should NOT be a censor of valid research projects:

What is NIH’s justification? Studies of intelligence do not pose any greater threat to the dignity of their participants than research based on non-genetic factors. With the customary safeguards in place, research activities such as genetically predicting an individual’s academic performance need be no more “stigmatizing” than predicting academic performance based on an individual’s family structure during childhood.

The cost of this censorship is profound. On a practical level, many of the original data-generating studies were set up with the explicit goal of understanding risk factors for various diseases. Since intelligence and education are also risk factors for many of these diseases, denying researchers usage of these data stymies progress on the problems the studies were funded to address. Scientific research should not have to justify itself on those grounds, anyway. Perhaps the most elemental principle of science is that the search for truth is worthwhile, regardless of its practical benefits.

NIH’s responsibility is to protect the safety and privacy of research participants, not to enforce a party line. Indeed, no apparent legal basis exists for these restrictions. NIH enforces hundreds of regulations, but you will search in vain for any grounds on which to ban “stigmatizing” research—whatever that even means.

This is a no brainer. The NIH has NO business vetting the “political correctness” of research, and since nobody is investigating The Taboo Question—racial (or “ethnic” differences in intelligence—that issue doesn’t even come up. The only reason to prohibit “genetics of IQ” studies is a strict (almost Marxist) anti-hereditarianism based on the fear that there may be a genetic basis to differences in IQ. But we already KNOW that from studies of adoptions and relatives, which show that about 50-60% of variation in IQ among people is due to variation in their genes. The NIH appears to be afraid of being canceled. That is a hell of a way to do scence!

And I can’t imagine why the NIH would even think of restricting the data for any other studies. It seems to be IQ that’s the sticking point here, and that’s unconscionable. The data belong to the American public, and to American scientists, because the American public paid for it.

I’ve always object to the demonization of research that gives results that are politically or ideologically unpalatable, but this goes beyond the pale.  The government cannot withhold data paid for by us on the grounds that it might yield results that could offend people.

If a researcher has a valid reason to request these data, and the NIH refuses because of possible “stigmatizatization,” then I would say that a lawsuit is in order.

36 thoughts on “National Institutes of Health violates academic freedom, restricts dissemination of taxpayer-funded research

  1. the National Institutes of Health, a U.S. government science institute, has a huge genetics and “trait” database of several million Americans. The genetic data appear to be thorough, based on genome scans

    I think I agree with their decision. Too much of this constitutes Personally Identifiable Information, meaning it is open to a FOIA request. Imagine if you opted out of any of this research, but your family members chose to participate. Enough of your data would be available, simply through proximity.

    I would be afraid of how this dataset could be weaponized, even if it is currently anonymous. The data might be secure now, but there is no guarantee they couldn’t be personally contextualized (de-anonymized, etc.). Imagine health insurance providers getting access to genetic records of millions of people, and the extrapolations they could make. Even if the data are only available to researchers, the free market will get access soon enough.

    I’m unfamiliar with the precise details, though, so my opinion co

      1. I think it’s reasonable to worry about things one doesn’t know; until one does, of course.

        One thing which comes to mind with respect to this is that recently I read where the police solved a crime by comparing DNA sequence from the crime scene to a private DNA database (I believe 23andMe, but I am not sure). From this they identified the suspect’s family, which ultimately lead them to the criminal.

        This does suggest that DNA sequences obtained from dbGaP might then be used in a similar search of the same private DNA database, or something similar. I am quite sure that safeguards are in place, including HIPAA requirements and other protections to prevent this from happening. But it’s a reasonable concern (IMO), so it would be good to know what protections are in place.

        1. That identifiable info might be revealed under a FOIA type request, at best (if true, which is not at all clear), might constitute a reason not to release the data for anyone to use; it constitutes no reason whatsoever to be selective about which projects get to use the data. In fact, “vulnerability to FOIA” would be a reason for NIH to destroy the data and never collect it again, because NIH itself is subject to a FOIA request (and it does not have to be FOIA per se; any capability on anyone’s part of identifying individual participants either preempts use across the board or it doesn’t; there is no justification for selectively permitting some studies but not others).

          1. I’m not sure I understand what you are saying. My apologies.

            Perhaps I wasn’t clear. I wasn’t really concerned about people who want to play by the rules and I wasn’t suggesting they could obtain personal info from NIH.

            Any DNA sequence extracted from the dbGaP database may be all the information one needs to ID a person. Copy and paste the sequence into text editor and you have the basis for a search string which, using a different database, could identify the person who that DNA sequence came from. All it takes is for someone who has access to the sequence from dbGaP, the ability to press “ctrlC” and access to another DNA database. It isn’t clear to me what is there to stop that.

            1. Exactly. That risk is identical for *any* study that uses the data and, therefore, cannot explain NIH selectivity in permitting access to the data for some studies but not others. A constant (in this case “possible discovery of individually identifiable info”) cannot possibly explain variability (in this case, some proposals being permitted and others not).

        2. For solving crimes, DNA left at the scene is compared to whatever databases will co-operate (Edit: or can be induced to). It is not expected that the perpetrator would be so foolish to send his own specimen in, which would yield an exact match. But identifying a kindred one or more of whom did send in a specimen provides the detectives with a list of related people. Then they have to get DNA from the ones who had opportunity to commit the crime. If his DNA matches the crime scene DNA, they have their guy. The privacy of the kindred member who sent the specimen innocent of any knowledge of the crime is never compromised.

          In the administration of justice, everything is discoverable. I expect the police would seek production orders from all DNA databases if it helped them solve a cold-case rape and murder of a child. HIPAA and other privacy laws are to prevent snooping by the idle curious and the gratuitous and careless release by data custodians. If the Court approves it, the police (and defence) will get what they need, whether it’s DNA, tax records, medical information, your middle-school diary, the works. Everything but solicitor-client privilege and even your lawyer can’t hide evidence.

          This isn’t really the point. The issue is the NIH is refusing to release properly safeguarded data for some studies that they would release if the studies addressed other hypotheses. This is wrong.

  2. There is a difference between nature and nurture (heredity v. environment) when it comes to human psychology and morality. It is one thing to say that you may not do as well in math because of how you were raised, but it is yet another to say that your were born with a mathematical handicap. We can also point to how we view genetic modification differently than medication, even though they could result in the same outcomes. For better or worse, we view our DNA as a part of our being and our environment/upbringing as our experience in life. I don’t think it is fair to say that they are one in the same.

    1. Eric, I wish you were more specific in your argumentation. Jerry did not say that DNA and environment/upbringing are the same. What is your point? Blank slatists (most of whom are close to the left end of the ideological spectrum) don’t want to hear any talk about the genetic basis of intelligence since for them the irrelevance of genetics for understanding social behavior is an article of faith (see Steven Pinker’s book The Blank Slate), adopted because acknowledgment of the facts in this matter interferes with the belief in the infinite malleability of humans by social engineers (who the French economist Thomas Piketty calls the Brahmin left).
      Here’s something funny: the heritability of IQ is about the same as the heritability of obesity (50-60%). But leftist will eagerly point to the heritability of obesity to convince us that we should NOT stigmatize the obese (about 43% of the adult US population is obese now, and experts predict no end of the upward trend of the obesity rate). So leftists are actually saying: Given our modern environment, most of the obese adults in the US are born with a genetic handicap (which is correct and is known as the evolutionary mismatch explanation of obesity). So why is recourse to genetics okay in one case (obesity) but should be suppressed in another case (IQ and educational achievement)?
      Personally, I find both discrimination against the obese and against low-IQ people objectionable. But I agree with Jerry that the NHI should be taken to court.
      Note also that it is the opinion of Harvard university political philosopher Michael Sandel (who’s probably a leftist), that disdain for the less educated is well ensconced among leftists:
      Michael Sandel: Disdain for the Less Educated Is the Last Acceptable Prejudice. It’s having a corrosive effect on American life — and hurting the Democratic Party.
      Do you remember the cutting and dismissive comments Obama and Hilary Clinton made about supporters of the Republican party (who, nowadays, tend to be less educated than supporters of the Democrats)?

      1. It was in the quote from Lee:

        “What is NIH’s justification? Studies of intelligence do not pose any greater threat to the dignity of their participants than research based on non-genetic factors. With the customary safeguards in place, research activities such as genetically predicting an individual’s academic performance need be no more “stigmatizing” than predicting academic performance based on an individual’s family structure during childhood.”

        Lee conflates genetic and environmental variables which just isn’t how humans view it.

        Whether we like it or not, subjective human biases do not hold these two as equal. Don’t get me wrong, I am all for objective approaches to these studies, but science doesn’t happen in a cultural vacuum. The NIH is a political entity by its very nature, however much we hate it (and I say that as someone who has been funded by NIH research grants). I am also against woke culture, but there is a balance in there somewhere.

        1. You state: “Lee conflates genetic and environmental variables which just isn’t how humans view it.” Do you know of any data that support the “which just isn’t how humans view it” part of your statement? For instance, it’s pretty obvious that less comely and obese people are discriminated against. But poor people and face less discrimination? The less educated (Hilary Clinton’s “deplorables”) face less discrimination (I know that education reflects both nature and nurture)? I’d like to see some data.

    2. I think it would be fair to say, however, that we are the result of complex interactions between our genes and our “environment/upbringing”. Makes little sense, then, to forbid inquiries into part of that complexity, no?

  3. Is the problem NIH is trying to avoid something like this? “Study shows genetic traits that link variation in human intelligence and education to variation in disease.” “Different study shows variation in test scores, education, and disease between black and white people.” Everyone knows there are some small genetic differences between human racial groups. Some racists then confound the two patterns and argue that black people are genetically less smart than white people. Is that it, or is there some other way to account for NIH trying to avoid GWAS or PheWAS to study these traits.

    Seems like a shame because the dbGaP database has a lot more genetic variation in it than some others like the UK BioBank or the Iceland database that sample a less diverse population.

      1. Feel free to prepare yourself. I won’t be bothering. All such studies of the genetic bases of group diffs in IQ are correlational. They consistently fail to meet the standards necessary for inferring cause from correlation, even though a certain hardcore group has been at it for over 50 years. That is an epic record of methodological failure. Given how weak the methods are for inferring cause from correlation, and how frequently this error has been repeated, one might even begin to suspect that there is more going on there than a courageous search for truth.

        On the difficulties of inferring cause from correlation:
        Although this is not the place to debunk either that paper or prior work similarly failing to meet standards for inferring cause from correlation, I did find this quote from their paper amusing:

        “…we estimate that the known eduPGS can naïvely explain as much as 20%–25% of the African-/European-American intelligence gap.”

        So, even if you take their results entirely at face value, which is not justified by virtue of their failure to meet causality inference standards, ~3.6 of the 14.7pt IQ gap in their data is explained by genes. One can only speculate on how much further it would have been reduced had they meet the standards necessary for inferring cause from correlation.

        1. Thanks for the link to the Rohrer paper. Interesting and one I need to re-read and think on. Very helpful. I have long been bothered by this very issue (from a layman’s perspective), but couldn’t really grasp it.

        2. Thanks for the response!

          “…fail to meet the standards necessary for inferring cause from correlation…”

          Seems like I’ve seen that argument used against the hereditarian position in general. Do you agree that it’s a meaningful general argument against hereditarian interpretations, or is it mostly meaningful just WRT IQ? It seems to me that assuming a genetic causal correlation has been the default assumption used in many other contexts, and I’m certain that in many, many cases the assumption was proven correct.

          “…one might even begin to suspect that there is more going on there than a courageous search for truth.”

          In my case, rather than suspecting, one could simply ask. Prejudices can run either direction: to affirm hate and xenophobia, or to deny the existence of troubling differences altogether. I feel that my stance is perfectly neutral: it just seems obvious to me that the brain could not possibly be exempt from the natural human variation we see *everywhere* else.

        3. Setting aside the issue of causation, do you allow that there are well established differences in various group IQs? That the mean IQ of different groups – including racial groups – differ according to currently available data?

  4. Maybe of interest to some readers here:

    Why parenting matters less than you think. An interview with behavioral geneticist Robert Plomin about his new book Blueprint: How DNA makes us we are (MIT Press, 2019). New Scientist, May 25, 2022, pp.39-41

    Q: Isn’t treating people differently based on genetic tests heading towards eugenics?
    Plomin: The only people who ask me questions about eugenics and the Nazis are journalists such as yourself, not the general public. It’s not an unimportant question, but the belief that environment dictates everything has done worse damage than genetic determinism. No one has ever asked those who hold that belief: “Aren’t you concerned about the abuses of the past, where environmentalism has led to totalitarian regimes that assume everybody is the same and puts people into gulags if they are not?” Even in the West ascribing everything to upbringing has done a lot of harm.

    Q: Can you empathise with those who don’t like your ideas about parenting?
    Plomin: I think my latest book actually has a good message for parents: that they should lighten up and enjoy their children. Because, despite what they think, parents aren’t in control. If you think your kids are clay that you can mould, forget it. I think it’s better if we think of parents as resource managers, whose job it is to find out what their kids like to do and give them opportunities to do it. Why not accept that it’s a relationship – enjoy it as best as you can, and watch who your children become.

    Chapter 7 Interview with Robert Plomin about behavioral genetics. pp.101-124
    in: Thomas Witkowski (ed.): Shaping Psychology: Perspectives on Legacy, Controversy and the Future of the Field. Palgrave Macmillan, 2020
    the first two questions put to Plomin are about how awful it is deemed to be to study the genetic bases of social behavior

    If you like to take a look at these interviews, free download here:!AmJNthmLmMu7klylLcvt1IPYJvHN?e=oUZxJM

    1. Throw in Judith Rich Harris, who’s shown parenting has little to no effect on many important life outcomes, and you have a recipe for driving the blank slaters mad.

  5. >> Remember that the NIH is completely funded by the American taxpayers, so those data were accumulated with our money. To me, this means that any researcher with a valid project should have access to the data.

    So why require a “valid” project, or any project at all? Shouldn’t any curious citizen/co-owner of the data be able to see it?

  6. The NIH administration has evidently been captured by those holding the view that anything that is or might be related to “biological determinism” must be suppressed. In much the same way, anything related to “bourgeois Mendelism-Morganism” was suppressed for two decades in a galaxy far away.

    The Lysenkovshchina over there is widely misunderstood as an imposition of state policy from above on academia. But before the Party, the state, and the NKVD got involved, it began in the 1930s as a dispute within the Soviet academic world, as recounted in detail by Zhores Medvedev and other writers. Lysenko, Prezent, and their faction enlisted the Party and the state (and thus the NKVD) on their side. But had they succeeded only in capturing the academic and research administrations, the outcome would have been rather similar—an experiment that we are apparently running today in our own galaxy. The outcome here, for biomedical science generally, might be as brilliant as it was for the science of genetics in that galaxy far away.

    1. There are any number of genome-wide association studies (GWAS) funded by the NIH, so it isn’t a case of the NIH suppressing all work on genetic predispositions. However, the NIH does consider the ethics of a study before granting access to the data.

      Even if we ignore genetic influences on intelligence, there are still quite a few ethical issues tied up in these concepts. If certain ethnic groups are found to carry alleles that predispose them to colon cancer does this mean insurance companies can increase their insurance premiums based on their ethnicity? Bringing intelligence back in, could top tier private universities screen applicants based on their intelligence alleles?

      However, there is ultimately no way to keep this knowledge suppressed. It’s sitting there waiting to be discovered. Pandora’s box was opened once we gained access to high throughput sequencing, and there is no going back. There will be a certain percentage of the population who will feel betrayed because their tax money has been used to label people as dumb because of their genetics, but facts are tough things to suppress.

      1. If insurance companies knew there was robust genetic linkage between ethnic group and risk of disease, beyond that contributed by the health questions they ask of applicants individually, then yes, they would want to charge higher premiums to members of that ethnic group. But how would they know the ethnic group of the individual applicant? I don’t know about America, but in my 60-odd years of life I have never been asked on any kind of application form what ethnic or racial group I self-identify as. I can self-identify on the Census and on occasional voluntary diversity surveys at work—I always just wrote “Canadian”—but never on any form where identifying as to race was required to access the service. So if Blacks have more high blood pressure and South Asians get more severe Covid, that can’t produce insurance discrimination here.

        Insurance companies know that most of these racial effects are small, and the way to prevent adverse selection is to sign up as many currently healthy people as possible, of whatever race.

        One piece of personal information that does have underwriting impact for life insurance is sex—(which one, not how much.). Since women die at lower rates in every decade than men, they pay less for life insurance. But since they live longer, their life annuities also pay them less a month. No one seems to think this is unfair, but the magnitude of the sex difference and the personal and family health history are so much bigger than racial effects, those are all the insurance companies need to set a fair premium.

        Note that if a gene sequencing done for entertainment showed you had an allele that the insurance company was interested in, you would have to disclose that if they asked, just as they would want to know that your father and all four of his brothers died in their 40s of heart attacks.

        Edit: I don’t see any privacy concerns in releasing anonymized genome results to researchers. If society wants to avoid stigmatizing ethnic groups, it shouldn’t collect that information in the first place.

  7. The charge of “biological determinism” is a perfect example of the logical fallacy known as the “straw-man” committed by “blank-slaters”, since no-one actually believes that biology (genes, heredity) is all that matters in explaining human behavior, while they themselves are, with few exceptions, “cultural determinists”, looking only to economic or social causes (man having no nature, only a history) while completely discounting the importance of genes.

      1. I agree Lee. I feel a system is needed in order for both legit research and privacy. I don’t know how to establish such a system, but I can understand one that works.

      2. Lee, thanks a lot for commenting here – you have real knowledge of these things (unlike me who’s just an interested spectator). How could a researcher deanonymize the dbGaP data in any way that would actually be a risk to participants’ confidentiality? I’d accept that it’s possible but can you explain how?

        1. I am not a geneticist, so I cannot answer that question. My guess is that concerns about anonymity are red herring with respect to Jerry’s essay, because, even if it is possible, it is almost always also possible to make data available in some way that ensures the person is not identifiable. But I say that on general principles, not because I specifically know how to do that with genetic data per se.

    1. @ Michael Sisley: How so? As long as the names and other identifying info are removed from the records to be released, which they are, there is no threat to privacy.

      1. I agree with you and yes, that should happen. But you know what people think and mistakes do happen. But yeah, so long as there is a good system it should befine.

        Don’t take me too seriously.

Leave a Reply