As the BBC has reported here, the first baby has been born in the UK whose birth was a function of preimplantation genetic diagnosis (PGD) for–or, rather, against–a gene, some of whose alleles have mutations strongly correlated with certain forms of breast and ovarian cancer. What follows are the basics, and two viewpoints on this, and a poll for you to participate in.
Basics: PGD is used in conjunction with in vitro fertilization when a number of eggs are fertilization, and so multiple embryos formed before any are implanted in the womb. This then allows for some testing on the embryos, and selection to be made amongst them in the light of the outcomes of the tests. One primary use of PGD is to screen for conditions seen as disabilities, when tests can be formed on embryos that are only a few days old; another is to test for sex, especially in countries where there is a decided preference for males (despite the use of those tests for this purpose being against the law there, e.g., China). BRCA1, touted early on (and misleadingly) as “the breast cancer gene”, was first detected in the early 1990s, and is known to play an important role in tumour suppression. Sometimes, it takes forms where it does not play this role. The BRCA1 gene is on chromosome 17 (band 21), and is associated with 5-10% of all breast cancers. In some forms that BRCA1 takes (alleles), the gene can reduce the risk of breast cancer developing in women who carry it; in other forms, it correlates strongly with mutations that give high risks of developing breast (and ovarian) cancer later in life. BRCA1 was the first gene associated with a risk for cancer that was approved for testing to be delivered in conjunction with PGD in the UK in 2006. All such approvals in the UK must be given by the Human Fertilisation and Embryo Authority; it has currently licensed PGD for about 60 genetic conditions.
First view: This is the view that you’ll find represented most prominently in the BBC article, and at H+biopolitics (which deserves a h/t here for posting on this: thanks!; there’s also some discussion at the Women’s Bioethics blog). This is a medical triumph that increases parental choice and lessens future suffering. It means that parents who have a history of susceptibility to a disease–at least 4 immediate female family members on the husband’s side in this case have developed the forms of cancer associated with BRCA1–can signficantly reduce the chances that children that they have will have children who go on to develop those cancers. Not all of those with the allele will develop those cancers, and there is some chance that the use of this testing in conjunction with PGD will lead to the discard of an embryo that would have a higher overall level of health, other things being equal. But medical intervention is never about certainty, only about improving the chances for health and reducing the chances of disease.
Second view: This is an objectionable use of genetic technology in itself that is also the thin end of the eugenic wedge. One might think that all combinations of PGD with genetic and other testing on embryos is morally objectionable, but put such views to one side here. Even if one thinks that some uses are justified (e.g. for cystic fibrosis), the decision to expand those uses to include cases like the BRCA1 gene–where the science has been hyped and the susceptibilities cannot be reliably estimated–is to both buy into a crass form of genetic determinism and to open the door to a massively expanded use of PGD that amounts to a dangerous form of eugenics. There are hundreds of correlations as strong and as questionable–as that between BRCA1 and breast and ovarian cancer. It is difficult to see how, once one of these becomes acceptable, that the others will follow suit. But that’s just negative eugenics all over again, now not on fetuses but on embryos.
The Wikipedia article on PGD provides very good coverage of PGD for those interested in following through; the article on the HFEA is sketchier but can be linked from the PGD article.
Living Archives on Eugenics Blog 
“is to both buy into a crass form of genetic determinism”
I don’t see how anyone could reasonably think this. It’s very clear (and is made clear by genetic counselors and doctors for patients) that this is a probabilistic intervention. Genetic differences between people often *do* determine outcomes in conjunction with other factors. Can you give a definition of ‘crass genetic determinism’ such that it is both false and implicitly endorsed by the actors involved with this procedure?
If the concern is the limited effect of an individual mutation, then I would make an analogy to a single cigarette. The odds that a single cigarette will induce lung cancer are very low, and yet it is a good thing to cut your lifetime cigarette consumption by one. It is better to cut it by 5% or 20% or 100%, but that’s no argument against refraining from any particular cigarette.
My worry is that people will select against variants that have unrecognized benefits. For instance, some BRCA variants may be associated with increased intelligence (faster cell growth both in brain development and in the breasts).
http://www.futurepundit.com/archives/003694.html
But I would argue that the way to avoid such problems is by conducting the large-scale studies to learn about all the effects of variants, both advantageous and disadvantageous, so that parents can take both into account.
The thought is something like this: to extend the range of approvals from cases like that of (certain kinds of) cystic fibrosis to those like tha of (certain kinds of) breast cancer is to assimilate the former cases to those that are monogenetic and for which the causal relationship between gene and disease is relatively direct, well-established, and a major pathway to the disease. By moving to cases that are much more deeply probabilistic, where the pattern of causation is not well understood (e.g., there are over 600 mutations that have been found on BRCA1, only some of which are correlated, 30+ years later, with breast and ovarian cancers), and where these account for a small minority of the cancers identified, we’re in much riskier territory, something masked by the hype associated with early pronouncements about BRCA1 as “the gene” for breast cancer. The concern here is as much one about what’s going on with those most directly involved in the practice (and this particular case) as it is about the public perception of what the practice involves.
“we’re in much riskier territory”
Every time a child is conceived there’s risk, with both new mutations and massive genetic reshuffling. PGD just picks out an embryo conceived by the union of parental gametes. The DNA in question was present in the parents and would be present in half the kids if PGD was not used. And if you’re selecting against rare mutations in mutation-selection balance, then side effects are quite unlikely. If you systematically select variants with health benefits, then you will only get average negative effects elsewhere if there is actual pleiotropy, e.g. genes that cause disease but provide some other non-medical benefit. Do you think the ‘riskier territory’ lies in the area of negative non-medical effects? Would you support the collection of extremely large databases on the genetics of non-medical traits (personality, aptitudes, dexterity, neurodiversity) so that we can watch out for side-effects of variants targeted for selection based on their influence on disease?
Or by ‘risky’ do you just mean the thin end of the eugenic wedge? Sure, there’s a direct path from this approval to selection for things like higher IQ or reduced obesity that would both be beneficial for health and have broader social ramifications, but as long as we set aside personality (which affects politics) but what’s wrong with that? A future society with lower medical costs and more high-IQ individuals will be better able to solve global warming, protect the environment, eliminate poverty, reduce inequality, and generally make the world a better place.
“the hype associated with early pronouncements about BRCA1 as “the gene” for breast cancer.”
The usual pattern these days is that the headline is nonsensical to the scientifically literate, the first paragraph of the article improves things greatly, and by the end all the relevant points have been hit (that selection is for one allele of a gene, rather than for the presence or absence of a gene, that effects are probabilistic, that the portion of the variance explained by a particular variant is small, etc). Anyone who actually reads the articles will get the picture. Admittedly, some people will only read the sensationalist headlines, but people using or providing PGD and policymakers will read past the headline or spend 5 minutes Googling. Services like 23andme are at the beginning of popularizing an accurate, probabilistic understanding of what genetic data can tell us. I don’t think it’s reasonable to oppose medical advances on the grounds that some people won’t read carefully or understand what they are until they become widely used.
“but what’s wrong with that?”
Or rather “what’s wrong with that?”
Utilitarian,
Keep in mind that there are two points of view represented in the initial post; I share some sympathies with both, even though they stand in tension. But there are definitely things that concern me about the first that are brought out by your questions about the second.
You ask: “Do you think the ‘riskier territory’ lies in the area of negative non-medical effects? Would you support the collection of extremely large databases on the genetics of non-medical traits (personality, aptitudes, dexterity, neurodiversity) so that we can watch out for side-effects of variants targeted for selection based on their influence on disease?”
Fair questions. I think the risks are still “medical”, a fairly narrow sense of that term, since we have quite an impressive history of systematically under-estimating the complexities to genetic interventions in the human organism. But it’s true that, when it comes to the kinds of non-medical traits that you list (which have, of course, been medicalized of late), I’m even more sceptical about our collective ability to get things right. A large part of this is that I don’t think we are on very firm ground when it comes to the basic categories–of personality or neurodiversity, for example–that we might, in principle, collect lots of data on. Is borderline personality disorder a solid enough diagnostic category to whack into the equations? Is autism or (now) autism spectrum disorder?
The comment “A future society with lower medical costs and more high-IQ individuals will be better able to solve global warming, protect the environment, eliminate poverty, reduce inequality, and generally make the world a better place.” indicates one difference between us. I’m not sure that manipulating the genome WILL produce lower medical costs, and even less sure that a world with “more high-IQ individuals” will produce a better world, as opposed to a world that allows more efficient killing of one another. I’d be curious what evidence there is for either claim. There’s all sorts of estimates of savings here and there, and smarter people can certainly solve certain kinds of problems more effectively than others. But what’s the basis for the global, across-the-board optimism here? Are medical costs lower now than they were 100 years ago? Has the Flynn Effect (generational increasing IQ) over the past century been accompanied by the solution to any of the problems you list?
“Keep in mind that there are two points of view represented in the initial post; I share some sympathies with both, even though they stand in tension. But there are definitely things that concern me about the first that are brought out by your questions about the second.”
I understand, please take what follows as an attempt to clarify those concerns.
“A large part of this is that I don’t think we are on very firm ground when it comes to the basic categories–of personality or neurodiversity, for example–that we might, in principle, collect lots of data on. Is borderline personality disorder a solid enough diagnostic category to whack into the equations? Is autism or (now) autism spectrum disorder?”
The way I see this, there are two big problems. First, people miss the upsides of neurodiversity, e.g. Asperger’s may make you a better computer programmer, or make it easier to see and raise socially taboo ideas. Gay men may have an unusual combination of spatial ability (mean and variance) and motivations that leads to their domination of the fashion industry, with excellent results. Again, my solution is collecting lots of fine-grained data, which is possible, but many of the people who oppose ‘risky’ PGD also oppose funding the behavioral genetics that would reduce the risks (and increase the benefits). In light of that common and disingenous position, having a modest number of people use PGD now (BRCA selection is not going to seriously change the gene pool) may be a good thing: it could stimulate opponents to shift from blanket opposition to actually supporting research into the risks.
Second, a genotype may be valuable to society, but be comparatively burdensome to parents. If parents want grandchildren, then they may prefer that a son be heterosexual rather than homosexual, even if society benefits from the presence of gay men. Similarly for other types of neurodiversity. But I would strongly oppose the idea of prohibiting PGD to *force* women to give birth to neurodiverse children. Instead, in such cases a subsidy could be offered, i.e. a baby bonus for parents who create children with valuable but unpopular neurodiversity.
If the concern is simply with the perpetuation of neurodiverse ingroups, then I would tend to think that such groups should pay the costs of reproduction. If some virus disproportionately rendered Caucasians infertile, it would be a travesty to require women of other ethnicities to bear Caucasian children as surrogate mothers in order to perpetuate an ethnicity. Likewise, if straight parents tend to select straight kids, and gay people want to maintain the size of their community for non-universalistic reasons, then they should select gay kids themselves. Similarly for deafness or the autistic spectrum or other types of neurodiversity.
My own neurodiverse ethnic group is rapidly dissolving into the American population through intermarriage and assimilation, and I see this as a great thing. When co-ethnics demand that their kids mate ‘within the tribe’ and perpetuate the ethnicity, this seems parochial and atavistic. I guess my suspicion of some of the ‘What Sorts’ project ultimately stems from a worry that it reflects the same parochialism.
“But what’s the basis for the global, across-the-board optimism here? Are medical costs lower now than they were 100 years ago?”
Medicine’s cost-benefit ratio is vastly better than it was 100 years ago (it was probably still a net negative at that time). Anyway, this is a non sequitur, genetic variation affects demands on the medical system, as it has in the past and will continue to do so in the future.
“even less sure that a world with “more high-IQ individuals” will produce a better world, as opposed to a world that allows more efficient killing of one another. I’d be curious what evidence there is for either claim.”
A rather large amount of evidence actually.
On global warming and the environment, the connection to faster scientific research is obvious, but here we go:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=11519655&dopt=Abstract
High IQ is associated with progressive political values in individuals.
Click to access Deary%20(2008)%20Psychological%20Science%20iq%20enlightenment.pdf
You can spend time with the General Social Survey, which includes huge amounts of value/attitude and social indicators, as well as a verbal IQ test. Smarter people are more tolerant of foreigners, more supportive of foreign aid, more positive towards other races or ethnicities, more supportive of equal status for men and women, more supportive of science, more supportive of the Americans with Disabilities Act, etc.
http://www.norc.org/GSS+Website/Browse+GSS+Variables/
Students at colleges with higher SAT scores are more cooperative in Prisoner’s Dilemma experiments.
Click to access 443.pdf
IQ (adjusted for the Flynn Effect) is very strongly associated with economic prosperity and economic growth.
http://ideas.repec.org/p/wpa/wuwpdc/0507005.html
National IQ strongly predicts national rates of poverty, infant mortality, life expectancy, Gini inequality, and a host of other human development measures:
http://en.wikipedia.org/wiki/IQ_and_Global_Inequality
“Has the Flynn Effect (generational increasing IQ) over the past century been accompanied by the solution to any of the problems you list?”
First of all, in areas where intelligence has been substantially boosted in controlled experiments, such as with iodine supplementation of salt (iodine deficiency can cost 15 IQ points, as well as causing goiter and other problems) there have been large improvements in prosperity and social indicators. The UN, Kiwanis Club, and Gates Foundation have devoted hundreds of millions to work for the spread of iodized salt and iron-fortified bread to raise the IQ of kids suffering from malnutrition (and thus their adult IQ). Likewise with mercury and lead poisoning.
Second, Flynn himself doesn’t think that the Flynn effect has been an increase in general intelligence. The gains are confined to a few subtests, e.g. vocabulary and arithmetic have not improved. Flynn thinks this is a result of greater societal emphasis on some skills rather than an increase in general intelligence.
Third, there is reason to think that the gains do not primarily reflect an increase in the g-factor (the component shared across skills, which predicts car accident rates, time for military recruits to fix a machine, reading comprehension, brain size, etc).
http://dx.doi.org/10.1016/j.intell.2006.07.006
“It’s very clear (and is made clear by genetic counselors and doctors for patients) that this is a probabilistic intervention.”
I’m not so sure about this. The existing sociological literature on amnioscentesis and other forms of prenatal testing suggests that prospective parents have a lot of trouble grasping the difference between tests which diagnose whether some condition will occur (e.g. amnio can determine Down syndrome) and tests which only refine the likelihood that some condition will occur (e.g. nuchal fold translucency can estimate the risk of Down syndrome, but can’t determine whether or not the risk is present). The tendency is for prospective parents to treat all tests as diagnostic and/or to assume that a positive result on a probablistic test (higher than average risk that a condition will occur) is the same as a positive result on a diagnostic test. I don’t have a lot of faith in the ability of doctors or nurses, especially in overstretched public health systems, to effectively communicate these statistical nuances. The difference between diagnostic and probabilistic may seem obvious in theory, but in real life it’s often not that well understood.
I just wrote an extended post, which seems to have been devoured by wordpress. I will therefore only cover the links to IQ evidence you requested.
The human development effects of boosting IQ by iodine supplementation in salt (universal in the West today, but not in Africa or India) are clearly positive.
IQ predicts progressive values.
Click to access Deary%20(2008)%20Psychological%20Science%20iq%20enlightenment.pdf
The General Social Survey has IQ, social indicator, and values/attitudes data. More IQ is associated with social liberalism and more progressive and tolerant attitudes with respect to foreigners, other ethnicities, and the disabled.
http://www.norc.org/GSS+website
IQ is strongly associated with growth and prosperity between countries:
http://ideas.repec.org/p/wpa/wuwpdc/0507005.html
Plus it predicts nearly every positive human development indicator:
http://en.wikipedia.org/wiki/IQ_and_Global_Inequality
The Flynn Effect isn’t what it’s hyped as. Ask Flynn, who says that it was not an increase in general intelligence, but the development of particular skills that are disproportionately helpful on some tests:
Gains have not primarily been on the g-factor, the general ability that is related to motor accident rates, the performance of military recruits, brain size and MRI neuroanatomy scans, etc. There has been essentially no Flynn effect on reading, vocabulary, or arithmetic.
Stopping mercury or lead poisoning, or giving iodine-fortified salt or iron-fortified bread to kids with vitamin deficiencies, DOES result in g-factor improvements.
What happened to your original post was it got spammed by Akismet, probably due to all the links in it, though maybe for some other reason. It’s recovered now, but I’ll leave up the secondary post.