Embryo Selection for Mental Health

A conversation with Ives Parr

I originally released this conversation as a podcast in audio form, but I am re-releasing it as an edited transcript and removing the audio to ensure Ives' anonymity.

Ives Parr is an independent scholar known for his writing on the science and ethics of genetic enhancement. I invited him on the podcast to talk about how polygenic screening and embryo selection could be used to prevent or reduce the risk of mental illness.

Ives and I discuss whether there are trade-offs between mental well-being and other traits, potential arms races in selection for mental and physical abilities, which kinds of psychological diversity we should value, whether parents with mental health issues or physical disabilities should be permitted to intentionally create children with those same problems.

The Science of Genetic Enhancement

Jonah: Most people, when you say genetic enhancement, are probably thinking of CRISPR. The kind of genetic enhancement that you’ve been writing about has mostly been a different kind called embryo selection. Can you start by explaining how you see genetic enhancement and what the current state of genetic enhancement looks like?

Ives: Sure. You’re correct that CRISPR gene editing is higher in the public consciousness. In 2018, there was an incident where a scientist actually edited two embryos of young girls, Lulu and Nana. They eventually grew up to become two babies, and this was the first public case of that. And so a lot of people are more familiar with that incident than they are with the current available technology.

The most interesting technology at present is the ability to use polygenic screening. Polygenic screening takes place when a couple is undergoing in vitro fertilization. They want to decide some criteria to select the embryos on the basis of, so most commonly couples are deciding to select for embryos that are healthier.

In the recent past, it’s been very common for couples undergoing IVF to decide that they want to implant an embryo, and when they were looking into which embryo they wanted to implant, they would look for a condition called aneuploidy. Aneuploidy is an abnormal number of chromosomes. The most common example of this is Down syndrome. People with Down syndrome live very short lives that are much more difficult, and they suffer from cognitive impairment. So generally people want to avoid implanting an embryo that has any form of aneuploidy. Other aneuploidy conditions are also typically pretty bad, and most of the time aneuploidy pregnancies actually don’t result in successfully carrying the child to term. So it’s been common practice that couples would do screening for aneuploidy.

More recent developments and understanding of genetics have allowed couples to also screen for what are called monogenic conditions. Monogenic conditions are when a couple may be concerned about having a disorder associated with a single gene. It’s possible that one is recessive or dominant. So you could have some cases where future offspring may not have the condition, and some offspring may have the condition. A famous example of this is Tay-Sachs disease. When couples are deciding if they’re at risk of having a child who may suffer from these monogenic disorders, they may be inclined to undergo monogenic screening. That’s been common for quite some time now.

But the most recent and exciting development is polygenic screening. This polygenic screening looks at a larger portion of the genome and it looks for polygenic traits. Monogenic is associated with one gene. Polygenic is associated with many genes, and so this allows couples to select on the basis of diseases that are associated with many different genes. These are much more common, and more people are at risk of these sorts of things. It’s more like diabetes, heart disease, Alzheimer’s, cancer, and these sorts of conditions.

Now, this was only possible due to recent advances in technology that made sequencing and genotyping much less expensive. So now we can actually study our genome and understand which genes are associated with which conditions, and it’s now affordable for couples to biopsy the embryos and have them genotyped to determine whether or not they’re at higher or lower risk. And the company that’s using that sort of technology currently that’s most prominent is called Genomic Prediction.

The first baby born using that technology was born in 2020. The baby is Aurea Smigrodzki, and there is a video I have written about showing the father and his wife taking care of the baby. More recently, there’s also been a public case in which Malcolm and Simone Collins selected on the basis of both health and psychological well-being, and I also posted about them more recently.

Jonah: Polygenic versus monogenic screening gets a lot of people confused because oftentimes when social scientists or geneticists say oh, well this thing is genetic, being gay is genetic or depression is genetic, the retort to that from a lot of people has been, well, where’s the gay gene or where’s the depression gene? It’s common to see science write-ups where some journalists are saying, oh, it turns out depression isn’t genetic because there is no depression gene. You see this sometimes in the antipsychiatry community too.

The important thing to understand is that most of these conditions that are of social interest, they’re not one hundred percent caused by genes, but to the extent that they are, it’s going to be many genes and very rarely just one.

Ives: That’s exactly right. People tend to think in simplistic terms about genetic architecture, thinking that one gene equals one trait, whereas really these traits are very complicated, and many genes make small contributions. Psychological traits such as depression, schizophrenia, and bipolar are polygenic traits.

This should go along with people’s intuitive sense of these traits because almost everyone recognizes that there are levels to this. Some people are very depressed most of the time or experience a brief bout of depression. Some people experience milder depression. The genetic contribution to these traits is polygenic, and so the result is that these traits end up looking like bell curves. They form normal distributions because people tend to fall somewhere in the middle of the average of schizophrenia or propensity for depression, these sorts of things.

Embryo Selection for Mental Health

Jonah: Oftentimes, when people are talking about genetic engineering through embryo selection, they’re often talking about traits like IQ, health, and happiness. But as you’ve just said, by the very same logic, we could also make it so that people have the choice to select an embryo that is less likely to have a mental illness. That could involve screening different embryos to see if one is in the 99th percentile for likelihood of having schizophrenia or something like that.

That’s an extreme example, but maybe even something like anxiety, which is less heritable than the other mental illnesses, a person could then say, okay, I’m not going to choose that embryo to continue with and instead I’m going to choose an embryo that has very low polygenic risk score for these mental illnesses. Right?

Ives: Absolutely. I know one concern that people have is that, say you’re selecting strongly against schizophrenia, you may inadvertently select for depression; or you’re selecting against depression, you may inadvertently select for a very anxious person. Fortunately, there’s more positive pleiotropy. These negative psychological outcomes tend to correlate together. They tend to have shared overlapping genetic architecture so much that some believe we could characterize psychiatric conditions with a general p factor, and if we were to select against the p factor, we could probably reduce the incidences of all mental illnesses.

Most people agree that that would improve psychological well-being, but they raise specific objections and many concerns to these interventions that they would not normally raise towards an intervention that would make children psychologically healthier. They’re very concerned about the genetic angle. If you knew that you could reduce your child’s risk of psychiatric conditions through breastfeeding, people would have not many objections. They wouldn’t say, well, we need kind of crazy people or something like that. I think they would be fully on board. But as soon as you start talking about genes, people get very concerned

Jonah: We’ll talk about those concerns in a bit. But just to stay on the p factor, this is an important and under-discussed idea in psychology. We know that so if your father was schizophrenic or bipolar, you’re more likely to be schizophrenic or bipolar. That’s well-established and intuitive. But what we also see is that all these different mental illnesses are somewhat genetically related. Sometimes they’re weaker or stronger than others, but they all kind of cluster together. The p factor is based on the g-factor of intelligence, where all the different kinds of smarts correlate strongly. We see that with mental illness too.

There’s also been some recent work looking at the d factor, which is a disease factor, and so some people are arguing now that not only do we see mental illnesses clustering together, but also physical illnesses and mental illnesses cluster together. We can say that some people are just, unfortunately for genetic reasons, more prone to all types of illness. That’s on the one hand quite sad because it suggests that some people are getting the short end of the stick in life through no fault of their own. They are just born extra susceptible to mental and physical illnesses. On the other hand, it does raise the possibility that if couples could select against this d factor, it would be easy to select against that and improve the health of those future children, and perhaps over time the population.

Ives: This may be the way to go in terms of marketing: There’s a general overall health factor, and if you just select for overall health, you’ll make people mentally healthier, you’ll make them physically healthier, and probably have higher average cognitive ability. That may be a good framing as all of those are aspects of health. But the good thing about these sorts of positive correlations and these general factors that are positive is that it makes the task of selection much easier.

One concern, like I was saying earlier, is you have this sort of antagonistic pleiotropy wherein you try to make someone smart, you end up making them very sickly; you try to make them healthier, you end up making them mentally ill. It looks like that’s really not a concern when you’re doing selection. The good things tend to go together reducing how complex the embryo choice problem is for couples.

Jonah: You point out that if you sell it on physical health, that seems much less controversial than selling on mental health. This is kind of strange because mental illnesses can be just as debilitating as physical illnesses. But for some reason, people don’t get as upset and are less sensitive about most physical problems than mental problems. If you say we’re going to try and use genetic engineering to outbreed diabetes, everyone would say, oh, of course, great. But if you said, oh, well, we’re going to try and get rid of depression, ADHD, something like that, people would have a lot more reservations, maybe along the lines of, I identify with my mental illness, it’s in some way a part of me, and I don’t like that you are trying to get rid of it.

This is sort of the argument people say with Down syndrome will make against these more basic kinds of embryo screening to get rid of Down syndrome. They will say, why would anybody not want to be alive just because they’re like me? I have a great life. This is part of who I am. So I mean, do you think the sell, if there is a sell is going to be mostly around physical health, or do you think there’s a way to make it more palatable around mental illness?

Ives: I think that people are probably most susceptible to physical health, and the monogenic testing and the aneuploidy testing go largely uncriticized. People don’t necessarily care. People care in the case of abortion, but abortion is quite morally different to people than in vitro fertilization, and I don’t see huge public backlash for monogenic disorders or aneuploidy. People are hesitant towards polygenic screening for health and will be for a while, but eventually people will fully accept that and there’ll be high rates of acceptance for that.

Now, things like mental health, I think people will also be largely receptive to, but there’ll be some concerns about improving cognitive ability and maybe changing personality to some extent. People form their identity around their personality and trying to mess with that is kind of concerning. There’s a stigma around the association between psychological traits and genetics that makes people feel very uneasy, but I think eventually this technology will be totally embraced and will hopefully be considered an ethical thing to do. I think that mental illness will probably not face too much backlash if it’s stuff like schizophrenia, depression, or anxiety. Hopefully parents are receptive to that.

Selection against those traits will actually be more important than selection for physical health because a lot of the diseases that the company Genomic Prediction is selecting against have late onset. You don’t experience Alzheimer’s until you’re quite old or you’re not necessarily getting prostate cancer until you’re quite old, whereas these sorts of psychological conditions can come early. You can become schizophrenic or experience depression and anxiety early on, and I would think that these sorts of conditions reduce welfare more than physical illness except in extreme cases. So I think selection against mental disorders is a moral priority and more important than selection against physical health, although people will be more receptive to the physical health stuff.

Jonah: Yeah, I mean one kind of counter-example I could think of is take something like psychopathy, right? Psychopathy seems to be quite a successful strategy in some cases. If you were born a psychopath and you’re surrounded by people who are not psychopaths, who are just kind of pleasant and nice, you could prey on them quite easily. Psychopathy is usually something like 1% of the population, and I think it’s kind of stayed around that number consistently because that seems to be an equilibrium number. So I wonder if that’s an example of a trait where keeping that trait might confer some advantages to the people who have it over not having it.

Ives: Yeah, there’s a problem where if you genetically engineer everyone to be nice they’re more susceptible to being taken advantage of. There are a lot of these game theoretical issues that the author Jonathan Anomaly addresses in a book called Creating Future People. Another point about this is, for example, height in men. Women tend to prefer men who are taller than them, and they tend to prefer the taller of the men generally speaking. But if you were to genetically engineer everyone to be taller and taller and taller, it’s this sort of runaway problem where people end up being way too tall, physically unhealthy, and so forth. Now, if you were to make everyone as nice as possible, you would open yourself up to exploitation by people who are psychopaths.

The point about whether we should intentionally make people somewhat psychopathic if it confers an advantage in some cases, maybe we could capture the advantage of certain psychological traits without having the necessary downsides because the human genetic architecture is so complex, so it has so many dimensions to it that we may be able to have sort of the ruthless, hardworking person who’s trying to get to the top, but also make them really not prone to violence. We want CEOs and leaders of industry, but we don’t want serial killers and psychopaths who take advantage of others. I think that the genetic architecture of people is complex enough that at some point we’ll be navigating through this space, and we’ll be able to create people that are very driven but not prone to violence and hurting others.

Art and Culture Post-Mental Illness

Jonah: Thinking about psychological diversity brings up kind of an interesting question. People with mental illness add a lot of cultural richness. So you think about artists like Van Gogh, Ian Curtis of Joy Division, Woody Allen, people who are psychotic, depressed, anxious, they add a lot of richness to art and culture. They look at things in a different way than psychologically healthy people do. Personally, I would find life boring if people who didn’t have similar experiences to me weren’t creating culture and weren’t putting out works that looked at the world in a more depressing way. Psychologically healthy people often appreciate those kinds of works too. Will curing or preventing mental illness at a population level lead to a leveling of culture? Do we lose something when we do not have the same psychological diversity interpreting the society around us?

Ives: It’s an interesting question. I think psychological diversity will be important. To what extent is creativity and interesting cultural contributions dependent on unusual thinking, and how much is actually driven by mental illness versus personality? There’s an interesting article by Jonathan Anomaly, Christopher Ingal, and Julian Valles, called “Great Minds Think Different Preserving Cognitive Diversity in the Age of Gene Editing” that discusses why we can’t all psychologically be the same. There are some benefits to having people that think differently.

An important question is, how many of these great intellectual contributions are dependent on psychological suffering? To some extent, we would hope that maybe they just have very different personality differences and that you could just select for different personality differences. And the result could be that you get geniuses who make interesting creative contributions while not necessarily suffering from depression or anxiety.

It may be important for some people. When you have super geniuses that make really important intellectual contributions, sometimes they’re really weird and there is an interesting discussion about whether or not that’s necessary. Or is it a consequence of their intelligence? I don’t think it is. I think that being intellectually weird predisposes you towards creative and intellectual success. And I think that maybe for example, if you have someone that’s so strange that they don’t have any friends and they don’t have a girlfriend or boyfriend and they spend all their time on mathematics or something like that, they may be way more likely to make a breakthrough contribution to mathematics. Do we want that sort of thing, and can we avoid if it is necessary?

I think if we allow for reproductive autonomy fully, we’re going to see a diversity where parents who are more similar meet with each other and then have offspring that are more similar to them, and hopefully the interesting, creative, intellectual genius sort of contributions are dependent on personality rather than actually experiencing suffering.

Jonah:  Maybe if you have a mentally healthier population, the audience for that kind of cultural content also drops significantly. Maybe if there are fewer sad people, there’s going to be fewer Kurt Cobains who are making a lot of money off of sad music because people don’t want to listen to sad music. Or if they do, they want to listen to a very mild kind of sadness or something like that. Perhaps as the diversity of cultural creators shrinks so will the audience, and nobody will be complaining. I mean, in the end, there’ll just be a world where everybody’s listening to pop music and they’re watching fun, happy movies, but nobody’s upset about that or nobody cares. Maybe the problem solves itself.

Ives: Yeah, I think it’ll be a long time before everyone is a product of genetic enhancement. That would take a very long time. And then we would probably see if it really is a sort of, I think maybe if we need more sad people, you could use environmental interventions, but I don’t think people would be receptive. Because one concern is you could say, well, we need more sad artists or something like that. Maybe we could expose them to chemicals that distort their mood or give them brain damage and make them act crazy or something like that. But I don’t think anyone would be receptive to that sort of thing. What do you think about that idea?

Jonah: Yeah, that would be wrong. I could imagine some people doing it voluntarily, but I agree with you. It’s more going to be along the lines of people will stop caring. Maybe there’ll just be fewer Dostoevskys, but there’ll be fewer people who want to read novels like that.

Ives: I think that artificial intelligence is getting so good that it’s plausible we can just automatically create any sort of media that we want soon. I think you could go to Midjourney and generate sad pictures all day long or sad novels and maybe it’d be extremely high quality in the near future.

Can Impairment be Ethical?

Jonah: What if there’s someone who in the current year knows about this technology and wants their child to be anxious or depressed like they are? And the reason they want that is because they want to be able to relate to their kid. They want that kid to have similar experiences as they did. They want them to grasp the sadder side of life because that will sort of help them relate to the parent. And even though they know it’s not necessarily in the best mental health interests of their kid, they see it almost as a personality or character trait.

It strikes me as something that a lot of people, even as they are concerned about mental illness and mental health, might consider doing because increasingly we identify with our mental illness or issues. We say, oh, it’s part of my identity, it’s also part of my culture. Increasingly that seems true versus it being something you don’t tell anybody else about. So I’m wondering if that will be something that people want to do. And if so, I mean, is that wrong? Is that morally reprehensible? What do you think?

Ives: That’s morally wrong. It’s probably unlikely that a clinic would provide that service for you to have a child that suffers more. It’s kind of akin to if a person was violently abused as a child, they want to violently abuse their child. Maybe it’s not quite akin to that, but I try to think in terms of analogies with environmental comparisons.

I’ll propose a counter thought experiment: Imagine that a couple wanted a child that experienced depression and anxiety and by the luck of the genetic lottery, the child was born extremely psychologically healthy. Do you think it would be ethical for the parents to make the child’s life bad, to inflict treatment on them that would make them anxious and depressed? Maybe not allow them to have friends, physically hurt them, verbally abuse them, or do any sort of thing to better relate to them?

Jonah: I agree, it seems very wrong. There is a real-life example that is similar to that. Deaf people have these things called cochlear implants, which give some the ability to hear. And if a deaf kid grows up with those cochlear implants, then they will be able to live life as a normal person. They don’t need to learn sign language, although many do. But you’re basically going through life like a hearing individual. Some people in the deaf community are against cochlear implants for kids or anyone, and the reason they give is that they say, well, deafness is its own culture with its own community sign language is its own language. If we have a technological solution to this problem, we are going to lose that culture, community, and language.

From the outside looking in at that community, that seems cruel. How could you ever want to deny someone the ability to hear just for the preservation of your group? But a lot of these groups naturally have this group mentality where maybe they’re united around a defect or a flaw or a problem, but they want to preserve that. They’ve come to identify with it and don’t like other people telling them that it’s something that needs to be done away with.

Ives: Yeah, it’s interesting. There was a case a long time ago that was often discussed among bioethicists where a lesbian couple, both of whom were deaf, wanted to have a deaf child. So they intentionally selected a donor with congenital deafness. They ended up having one child who was deaf, I believe, and then they had a second one. The second child ended up being discovered and it got media attention, and that’s when it was debated. The second child was able to hear a little bit, and they said that they needed to start teaching him English or give him cochlear implants or something along those lines for him to properly develop speaking and hearing successfully. I think they intentionally decided not to give it to him. In my mind, it is unethical to deprive him of that environmental intervention.

But I also think it was unethical of the couple to intentionally have a child that they wanted to be deaf. I know some bioethicists disagree. I like Julian Savulescu, but I believe he defended their right on the grounds of reproductive autonomy. They decided to choose a donor that they knew the child would end up deaf with. But Julian Savulescu interestingly believes that, or I believe created a principle called “procreative beneficence.” He believes among the available embryos, you should pick the one that to experience the best life. But he also believed in reproductive autonomy, and he believed that that couple had the right to have that child and that having that child is not immoral because the child in a counterfactual world wouldn’t have experienced anything whatsoever. They wouldn’t have existed otherwise. So no harm was brought upon the child.

In my view, that’s kind of an irrational preference. I can understand why someone would want that, but that makes the child’s life difficult and it doesn’t prevent them from actually learning sign language and communicating with the deaf community, but it does ensure that they can’t communicate with other people. I think that level of desire is too far, and it reduces the well-being of that child. It is obviously a disability because you have less ability to do things, and we have accommodations in place to take care of people that have disabilities. I don’t think that people should feel bad for having them, but when we’re creating people we should want to create people that have experienced high welfare.

If a couple were to select an embryo they expected to be deaf. It’s not quite equivalent, but morally, I think that’s intentionally deafening your child when others that are not deaf are available. So inflicting that environmental deprivation on the child is somewhat like that. It’s very clearly immoral and most will recognize that taking out your child’s eardrums is morally wrong. But I think also for similar reasons, it’s not quite equivalent, but intentionally having a deaf child is also not morally good. I think it’s wrong to do that.

Conclusion

Jonah: Yeah. Well, I have been following the conversation around embryo selection for a while, and I’m optimistic about it, especially for mental illness. One thing that people have not wrapped their heads around enough is that mental illness to a large extent, especially bad mental illness, is quite genetic. It’s not that trauma isn’t bad, can’t make people unhappy, or can’t make someone go into a depression or something like that; I’m not suggesting that trauma isn’t terrible. But we have to focus more on the genetic aspect because that is doing a lot of the heavy lifting, especially for some of the most debilitating mental illnesses.

Ives, I’m glad there are people like you out there who are writing on this topic and who are trying to push for a better future, a healthier, happier future. Where can people find you?

Ives: I’m on Substack at https://parrhesia.substack.com. You can find me there. That’s pretty much the only place. I’m not active on Twitter or anything like that, so check out my stuff there and I really appreciate you having me on. Thank you very much.

Jonah: Thank you.