Genetic privacy and "Big Brother"

By Razib Khan | July 15, 2011 1:21 pm

Several people have pointed me to Mary Carmichael’s piece for Nature, Newborn screening: A spot of trouble. It’s free, but you have to register. The subheading is: “By raising hell about newborn blood-spot screening, Twila Brase could jeopardize public-health programmes and derail research. The problem is, she has a point.”

The broader issue is “genetic privacy,” and the nature of consent in our public health system. Here’s an interesting paragraph which gets to some of the things we’ve discussed here before:

Twila Brase was not always the kind of person who hands out politically charged propaganda in airports. On a first meeting at her modest office in a shopping plaza in St Paul, Minnesota, she seems more like the unassuming nurse she was back in 1995 — before she began her second life as a bioethical gadfly, and before she had started making YouTube videos that accuse her state of commandeering the DNA of children as “government property” through widespread newborn screening programmes. Her voice is quiet and level. It is difficult to write her off as a conspiracy theorist: she simply doesn’t sound like one, even when, 4.5 minutes into making the case against screening, she suggests that “some researchers” might be trying to convince the state to test day-old infants for genes linked to “a tendency towards violence”.

Here’s the problem, as I’ve noted before the government already has an excellent predictor of your predisposition to violence: your birth certificate. Genetic information only adds value on the margin, many of the things we can predict about people, from height to I.Q. to personality, are reasonably predicted by easy to access genealogical and sociological information.

But I do think that there are real concerns with the way large institutions take our genetic material, and then don’t tell us what’s they’re doing with it. I’m not scared about government or corporate biobanking, there are many more easy ways to “get me” than to be devilish with my DNA. But transparency and trust should go both ways. They better know whose DNA is whose, and they better willing to give me a sample if I want it if it is mine or my child’s. Mischa Angrist has covered this area thoroughly, so I’ll just let it go.

Addendum: Yes, I do think that as a practical matter security protocols are going to be such that there will be constant breaches and people will have access to your or your children’s genomes. I don’t think this is that worrisome, but it’s a real issue, and these institutions should own up to it, and provide some real value to the individuals whom it wishes to recruit for the greater good of science and public health. There are real trade-offs here, and we shouldn’t pretend otherwise.

CATEGORIZED UNDER: Genetics, Genomics, Personal Genomics

Comments (7)

  1. Your point about genealogical and sociological information already revealing about us seems well-founded. But — and this is probably no news to you — I suspect the collection of genetic information bothers people more not just because it seems (seems) more exacting, but because it seems and is more invasive: it’s pulled from out of us, from the center of our cells. It feels like a much more intimate taking of information.

  2. david, yeah. this is the kind of cognitive intuition that paul bloom has spent a lot o time discussing about. we believe in ‘essences.’ our genes are a fragment of our indelible essence. the realized behaviors of kith & kin are simply proxies for that mystical aspect of who we are.

  3. One of the big issues is forensic DNA identification based on material collected from crime scenes. The risk of false positives is pretty modest when you only collect this material from people who have run ins with the criminal justice system. The risk of false positives is significant when you have a comprehensive DNA database of everyone who is born.

    There is also real risk that incorrect inferences about the relationship between genotype and phenotype will be used to the disadvantage of the child and not cleared up until harm is done. For example, there is a “bad husband gene” possibly related to oxycontin expression, that is associated with higher divorce rates in Sweden. But, if the link is misstated and disclosure of an already taken genotype becomes socially the norm in relationships, individuals misidentified as having the trait could have their marital prospects greatly diminished.

    The difference between a birth certificate and a genotype in terms of predictive value is first, that the use of the former is legally prohibited for most purposes by the constitution (the “no corruption of blood” clause), and secondly, that gathering accurate information about parents that is disclosed by genotype is impracticable in a society that doesn’t have a centralized health record authority the way that Iceland and several other countries do.

  4. “Genetic information only adds value on the margin”

    I disagree with this premise. The media and many researchers are focused only on the weak difference between the 3R and 4R MAOA VNTR. On the far-off chance that you could discover a baby with a spontaneous mutation for Brunner syndrome, you would know that the child will likely have behavioral problems without genealogical information offering any clues. (If you are looking at everyone’s DNA, the chance that someone has it will not be so far off.) Then again, if you are looking at, say, 5 genes and using an extremely rudimentary calculation, as Beaver et al did, then you will predict adult violent behavior in black men better than having intimate knowledge of the child’s relationship with his mother. Could having some genealogical information offset the value of this information? Maybe so, but how are you going to get that? Plus, developing useful algorithms for that kind of information would take a lot of effort for a result that would not be so scalable with advancing knowledge as a simple genetic test, so it stretches a reasonable person’s capacity for paranoia. What happens when rudimentary genetic calculations give way to advanced algorithms?

    The debate over the relative value to the public of knowing who will be violent versus individual privacy rights and due process is a separate issue.

  5. . On the far-off chance that you could discover a baby with a spontaneous mutation for Brunner syndrome, you would know that the child will likely have behavioral problems without genealogical information offering any clues.

    i agree with you on the facts here, but this seems frankly kind of a dumb objection. when i say it “adds value on the margin” that’s an average. obviously this is not so for all people, some people carry large effect mendelian risks which they couldn’t have known about through family information. i guess i should have added the caveat only adds information on the margin for most people? what am i missing?

    . What happens when rudimentary genetic calculations give way to advanced algorithms?

    when will advanced algorithms show up? give me an estimate with some made up error bar. i’ve gotten more skeptical over the years as to the possibilities in the short term, but can be convinced otherwise.

  6. First, I want to stress that I was not addressing the value of newborn screening or biobanking, per se, but rather your premise. I was not actually objecting to these programs. You raise the question of the relative consequence of individual genes or individual persons with rare genetic syndromes. When advanced algorithms are available, the rareness of single-gene behavioral syndromes will be an obsolete point. Yet, even now a single case of Brunner syndrome could have enormous consequences if the government acted on the information because the media would hone in, and a legal challenge could make its way through the courts. (Whether the government would act is also a separate issue, but many, myself included, would argue that the government should act within reasonable boundaries to prevent violence.)

    As to when advanced algorithms will show up, advanced is relative, and if I could have the time, resources, and data, I could give you a much better algorithm in a jiffy. If you read the study, you will see what I mean. They took 5 genes and gave each 1 point. I think I could write a program to take the data and calculate constants that could weight each gene differently to maximize predictive value. I also would not have conflated all versions of MAOA, the best-studied gene, into just 2 groups because the 2R version doubles the rate of violent delinquency without needing an environmental-interaction variable, and 5% of black men have this allele. So, putting out a much better algorithm is possible right now. Sorry, if pointing this out isn’t as good as an estimated time of arrival, but, as far as I can tell, the limiting factors are how sensitive this subject is and how few researchers are focusing on it. According to Scopus, only 5 studies have cited Beaver et al, and all 5 were by Dr. Beaver and his team.

  7. i will check the paper. i still don’t see the point of bringing up brunner syndrome, as i obviously wouldn’t deny the fact that in a minority of cases large effect QTLs offer up actionable information. but i’ll let that rest.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


Edifying books

Collapse bottom bar