So now we know a little bit about what schizophrenia is, a little bit about its epidemiology, now we can really jump into, begin to look at the genetic research on schizophrenia. Today we're going to talk about twin and adoption studies. And really twin and adoption studies are basic bread and butter research in behavioral genetics, but it is the type of study that changed the field's view about the origins of schizophrenia. Changed it from believing that schizophrenia was caused by schizophrenogenic parents to that there was some biological component. So we begin with twin studies and I'm going to show you in the next slide a summary of twin studies of schizophrenia. But when you see that slide, you're going to see a very familiar pattern that I highlighted in the second week of this course. So first of all, the, the way the studies are going to be summarized, and it, this also goes back to the second week of the course, how do we measure similarity for a categorical or diagnostic variable? How do we measure twin similarity? We measure that in terms of what's called the concordance coefficient. And what a concordance coefficient is, is the risk of having the disorder in the relative of that individual. Concordance is going to vary from 0%, they have no risk, to 100%, they all developed the disorder. So, we're going to look at monozygotic and dizygotic concordances for schizophrenia. What we're going to see, and this is the typical pattern that we saw before, the MZ concordance is consistently greater than the DZ concordance for schizophrenia. Genetically identical twins are more similar than genetically non-identical twins. But secondly, and this is actually very important to understanding schizophrenia from a behavioral genetic perspective, the MZ concordance is never 100%. In fact, it's far removed from 100%. So here now is a summary of ten twin studies of schizophrenia in terms of the concordance, for. Yeah, I guess, in the light bar for monozygotic twins. In the dark bar, this is actually same sex dizygotic twins. There's no unlike sex dizygotic twins here. And you can see there for multiple countries but unfortunately there's kind of a very western bias here most from Europe. I guess one or two maybe from the US. There is one from East Asia. Now, the typical study. We are not going to go into any of these states in great depth but I will just give you a flavor for how these studies take place. 2% of the population is a twin. 1% will suffer from schizophrenia. Both of those are relatively common, but when you take the intersection between the two, it becomes a rare event. So you need some sort of systematic way. You're not going to advertise for people with schizophrenia in a newspaper and expect it to do a, or twins with schizophrenia in a newspaper and get a sample. So the reason for this western bias is that most of these are based on twin registries in western European countries, where they actually maintain a registry of, a registration of everybody who's a twin in the country, and also a registration of your hospitalizations or the different diagnosis you might have. And so, in these countries, what the researchers were able to do is identify twins who had a diagnosis of schizophrenia. Usually they wouldn't just depend on that diagnosis. They would actually go out and re-interview the twins, but by doing that they could systematically ascertain, identify, reasonably large samples of individuals who were both twins and suffered from this disorder. The light bars concordances in the MZ's, the dark bars concordances in the DZ's. You can see that there was all of a sudden a flurry of twin studies in the late 1960's, early 1970's. And if you think back to the first module, er, first unit in this in this course, that's really when behavioural genetics was reemerging from it's association with the eugenics movement. People began to come back to look at these things. And there's another flurry in the late 20th century as people wanted to understand more about the genetic epidemiology of schizophrenia. There actually are some studies prior to this, but they're methodologically kind of weak. So these are probably the strongest studies. The pattern, the light bars are consistently higher than the dark bars. That suggests a genetic influence. Very consistent finding. In fact, in every study that's what you see. Secondly, the light bars are never 100%. In fact, they cluster at about 40 to 50%. So the majority of the time, genetically identical twins are discordant for schizophrenia. Even though it is a heritable trait, the environment is clearly, clearly very important and that's part of understanding the behavioral genetics that we're going to go through here. So we find this very consistent pattern. As we discussed in week two, we might find that consistent pattern of greater monozygotic than dizygotic twin concordance because there's a failure in the assumptions underlying the twin method. Well, there are two key assumptions here that we should be worried about. One we've already addressed. And the first is that maybe twins are at heightened risk for developing schizophrenia. We touched on that in the second week and I gave you data that suggests that there's really no evidence of that. It happened to be from Sweden, but there are other countries that suggest the same thing. The other possibility is failure of the equal environmental similarity assumption that we talked about in the second week. Maybe monozygotic twins are more similar for schizophrenia, not because they're genetically more similar than dizygotic twins, but somehow they have a more similar environment. What makes us believe that it's their similar identical genotypes, not their more similar environments? Well, there is other lines of evidence that support a genetic interpretation here. First of all, there are reared-apart monozygotic twins studies. These are pretty rare and there aren't a large number of these reported in the literature. And they're summarized in, in, in a, in a paper by these two individuals back in 1982. But if you look at, again, it's, you have to take this with a little grain of salt, because it's pretty small sample, admittedly. But if you like at monozygotic twins that were reared apart, so they didn't really share other than that early uterine environment, you don't see a concordance that's lower than the concordance you're seeing in reared-together twins. Maybe that gives us some greater confidence that it's a genetic influence, not a failure of the equal environmental similarity assumption. What gives your, gives us actually even greater confidence that there's a genetic influence here, it's not genetically determined, but a genetic influence, is findings that come out of adoption studies. A completely different design. I'm just going to highlight in this next slide three adoption studies. Two done in Scandinavia, one done in the US. All three of these studies are consistent with there being some genetic influence being transmitted in risk of schizophrenia. These first two studies, one done in the 60s and then second done in Finland published in 2000, what they did is, they looked at the reared-away adopted offspring of mothers who had schizophrenia. And they compared that to a sample of, of adopted individuals whose birth mothers did not have schizophrenia. What did we see? We saw elevated rates in both studies, elevated rates in schizophrenia, in, the adopted children, who had a birth mother with schizophrenia, than in the adopted children, whose birth mothers did not have schizophrenia. That's consistent with a genetic influence. This third study, done in Denmark, is consistent again with a genetic influence. The design is just a little bit different. In this case, they looked at the biological relatives, they kind of went the other way, the reverse of, of the way that was gone here. In this case, they looked at the first degree biological relatives, like the birth parents, of an adopted individual who had schizophrenia, versus the first degree biological relatives of an adopted individual who didn't have schizophrenia. Again, among the biological relatives, a greater risk for schizophrenia, if the adopted individual had schizophrenia than if that adopted individual did not. If you look at what you might call the basic clinical or genetic epidemiology of schizophrenia studies of twins, study of reared-apart twins, and adopted, they're very consistent in suggesting two things. One is that there's an important genetic contribution here. And secondly that the environment is also important. That schizophrenia is not a genetically determined trait. This next slide summarizes the risk of, to various classes of relatives, of an individual with schizophrenia. Recall that in the general population, the lifetime morbid risk for schizophrenia, is one out of a hundred. So a randomly selected individual has a chance of about 1% of developing schizophrenia in his or her lifetime. Among third degree third degree relatives of an individual with schizophrenia, the risk is about two times that, 2%. Among second, these are all second degree relatives here, the risk goes up again. If you average those, it's about 4% or 5%. First degree relatives, 9 to 10%. It varies a little bit, depending upon the type of first degree relative, but on average, it's about 9 to 10%. And then, what we might call zero degree relatives, monozygotic twins, it's about 50%. 1%, 2%, 4 to 5%, 9 to 10%, 50%, your risk of developing schizophrenia, that lifetime morbid risk, increases actually exponentially with your degree of relationship with an individual with schizophrenia. And in fact, for somebody who has a genetically identical twin with schizophrenia, their risk of developing schizophrenia is 50 fold times greater than the risk in the general population. This exponential increased risk of schizophrenia with genetic relationship is something that, that we see with many genetically influenced or heritable disorders. You would see something very similar if you looked at risk for bipolar disorder. We're not going to talk about bipolar disorder in this class, but you'd see a very similar exponential increase. And that's actually a very important pattern. You'd see a very similar pattern even if you looked at a non-psychiatric disorder. Something like insulin-dependent late onset diabetes shows a very similar increase of risk with degree of relationship. And I, I've illustrated this exponential increase. I've flipped it a little bit in this slide, but it's the same pattern. I've aggregated the, the MZ twins, which is about 50%, the first degree relatives, second degree relatives, and third degree relatives. And I've just plotted it here as a degree of relationship with someone who has schizophrenia. It has this exponential rate of increase. That's an important pattern, because it's telling us something that, not only that schizophrenia is likely genetically influenced, but something about the nature of that genetic influence. And the reason it's telling us that is, what would the pattern look like if, for example, schizophrenia was a autosomal dominant disorder? It's not, but what would it look like? If you thought about something that was an autosomal dominant disorder, something like Huntington Disease, which we will talk about later in the course, then the risk to a 0 degree relative would be 100 percent. And the risk to a first degree relative would be 50%. So it'd be half that. A second degree relative would be 25%. A third degree relative 12.5%. And so risk would actually increase, not exponentially, but linearly with degree of relationship. Because the risk is increasing exponentially with degree of relationship, what we know about the genetics of schizophrenia is not only that is genetically influenced, but it must be more than one gene. There must be multiple genes, because multiple genes is what gives you this exponential pattern. If it's one gene, you get something that looks linear. If it's multiple gene, it increases exponentially. When we begin in, in, in a couple modules down the road to try to find those genes, what we need to know is that we're not looking for one thing. We're going to be looking for multiple things. That's also important in understanding another feature about the epidemiology of schizophrenia. So if they're multiple factors, multiple genes that cause schizophrenia, and actually, we're going to get to this, they're likely hundreds, if not thousands of genetic factors contributing to an individual's risk of schizophrenia, the people who develop schizophrenia are carrying a large number of these factors, hundreds of these before they develop schizophrenia. But because there are so many of those, many of us who, many of us who don't suffer this disorder are actually carrying these genes as well. We're not going to suffer schizophrenia, because we're not carrying enough of them to get the disorder. Only a select number of us, who have a large number of these are going to develop the disorder. But, those genes are to, are distributed throughout the population. We carry them. We reproduce. One explanation for the fact that schizophrenia's common, heritable, but has a selective disadvantage, is because there are many genetic factors that underly schizophrenia and because many of those factors are being carried by the general population who reproduce, that, that we're maintaining schizophrenia at this 1% rate, not because schizophrenia, individuals with schizophrenia are reproducing but rather the general population of people with schizophrenia, people without schizophrenia, are reproducing and producing because they carry many of these factors can produce a child with schizophrenia. If you go back a couple of slides, and I told you that the risk to a first degree relative of somebody with schizophrenia is 10%. That means one out of 10 individuals, who has a first degree relative with schizophrenia, will also develop schizophrenia. We think about a typical nuclear family. If the risk is 1 out of 10, then in the typical nuclear family, if somebody has schizophrenia, there won't be another individual with schizophrenia. Right, if the risk is only one out of 10. Most families will only have one person with schizophrenia. That means that the parents don't have schizophrenia, but yet, are able, they must be carrying the genetic factors that can produce an individual with schizophrenia. That's because there are many relevant genetic factors. And it's those many relevant genetic factors that somehow combine to produce the genetic risk in that offspring. The last thing here, though, is that if schizophrenia is a heritable disorder, and there are many genetic factors, factors going on, well, how heritable is it? What is our best estimate of the heritability of schizophrenia? And you recall from the third week in the course, we talked about the heritability of a phenotype. And if you remember that definition, the heritability of a phenotype is the proportion of variance in the phenotype that we can attribute to genetic factors. So if we want to answer the question what the heritability of schizophrenia is, we want to know what the proportion of variance in schizophrenia is that, that we can attribute to genetic factors. There's one little glitch in trying to do that, or one complication. And that is, schizophrenia is an either or trait. And we usually think about variance for a quantitative trait, something like IQ, that we'll talk about a little bit later in the course. Variance doesn't have, at least in this context, a well-defined meaning when we're talking about something either or. But our old friend Falconer actually developed a model that allows us to estimate the heritability, not necessarily of schizophrenia, but what is believed to underly schizophrenia. That is we, from a quantitative genetic perspective that there are hundreds or thousands of genetic factors and maybe a large number of environmental factors that affect our risk of schizophrenia. Then we imagine, or Falconer imagined, that those factors somehow combine to define some sort of quantitative liability that we have to developing schizophrenia. And we develop schizophrenia, or an individual develops schizophrenia, if they have a large number of these factors that are above some threshold value here. But if you're down here, you might carry a lot of those factors, but you won't, you don't have enough of them to, to develop the disorder. So what Falconer reasoned here is that, in what he called the multifactorial threshold model, is that if you want to understand the inheritability, or how strong genetic factors here, what you need to do is estimate the heritability of this underlying liability. And you, we can do that with quantitative methods that are a little bit more advanced than what we want to get into here. If you do that, what you get is, this is what the predicted risk to relatives is, if the heritability of that liability for schizophrenia is 80%. It looks almost identical to the actual empirical risk, the observed risk figures. If you look at the estimates of the heritability of the, strictly speaking the liability of schizophrenia, they generally cluster between 60 and 80%. Schizophrenia appears to be a pretty strongly heritable trait. Is it 80% heritable? 60% heritable? Given what we talked about in the third unit, that's hard to say. But it does seem to be strongly heritable. Actually, strong heritability is another feature of neurodevelopmental disorders. ADHD, autism, intellectual disability, learning disabilities, these are all neurodevelopment disorders. We don't have time to go through all of these in this course. They all tend to be highly heritable. Schizophrenia is a neurodevelopment disorder. It has a feature like a neurodevelopment disorder. Because it has a strong heritability. We don't, we really can't say precisely how heritable it is. But it seems to be 60, 70, 80% heritable. That figure, even though we can't precisely estimate it, that it's strong will actually play a role when we try to begin to understand if we can identify those genetic factors, because we begin, we can begin to, to try to answer the question is, are we accounting in the factors we identify, for the 60, 70, 80%. This time we talked about twin and adoption studies telling us something about the genetics of schizophrenia. Next time, we'll try to talk about how they also tell us something about the environment of schizophrenia. [BLANK_AUDIO]
