 Dr. Judith Rappaport. Nurses are in the business of nurturing. And we look to our psychologist, scientist, friends to teach us about the population who we will care for. I am so delighted to be able to introduce Dr. Judith Rappaport, who comes to this year's Nobel conference from Washington, DC, where she and her husband live and practice she maintains the privileged life of full-time researchers at the National Institutes of Health. Here, too, they've raised two very successful sons and are going on to be grandparents and improve the race. Born and raised in New York City, she recalls her own childhood as a delightful one where she could go out to play in Central Park and walk home from school with complete safety. Her parents made sure that she was given every opportunity to visit the many museums in New York and even took her to the theater on a monthly basis. So a wonderful beginning. And then she went on to Swarthmore, where she was, again, feeling privileged to attend the many seminars and discussions which propelled her on to the magna cum laude, which she received there, and went on to Harvard Medical School, where she received her degree and did her clinical and research training at Massachusetts Mental Health Center in Boston, Children's Hospital in Washington, DC, and the Laboratory of Psychology at the National Institute of Mental Health in Bethesda, Maryland. Judith learned the Swedish language and found it to be an asset when she was doing her two-year fellowship in Sweden at both the Psychology Institute of Uppsala and the Karolinska Hospital in Stockholm. Since 1984, she has been the chief of the Child Psychiatry Branch, National Institute of Mental Health. In addition to her research at NIMH, she is also professor of psychiatry at George Washington University School of Medicine and clinical professor of psychiatry and pediatrics at Georgetown University Medical School. A past president of the American Psychopathological Association and a fellow of the American Academy of Arts and Sciences, she has served on many editorial boards and national advisory committees, including the Society for Research on Child and Adolescent Psychopathology and the National Anxiety Foundation. Among her numerous honors, awards, and fellowships are the American Psychiatric Association Award for Research and the Presidential Maritorious Executive Rank Award. Dr. Rappapour's research has focused on diagnosis in child psychiatry, of attention deficit hyperactivity disorder, and obsessive compulsive disorder. During the past 10 years, her research group has been studying the clinical phenomenology, neurobiology, and treatment of childhood onset schizophrenia. Judith has authored or co-authored 300 scientific papers over that number and five books, the most well-known being The Boy Who Couldn't Stop Washing. Of this account of the experience and treatment of obsessive compulsive disorder, Oliver Sacks wrote, the dedication and care, the delicate and delicacy and sympathy with which she has entered the patient's lives, give her book an authority and a personal feeling which is deeply moving and impressive. And so today we welcome Dr. Judith Rappapour, who will speak to us about normal and abnormal brain development in childhood and adolescence. Welcome, Judith. Thank you very much, Judy Gardner. And I would like to thank President Johnson, Dr. Robinson, and this entire wonderful community for hosting me and my fellow speakers at what is, I think, really spectacular and unique conference. On a personal note, since Judy has already shared a lot of my autobiographical detail, I would like to say that the two years I had in training in Sweden and Uppsala were among the happiest and warmest memories that my husband and I have and that the sense of the community here and the caring bring back particularly fond memories of those days. I did indeed like it enough to have learned a fair amount of Swedish. Our oldest son is named Eric, spelt with a K. And he complains to this day that no one ever spells it right in this country. And I could give this talk in Swedish, but will not horrify my hosts by doing so. I also want to start my talk by saying that the work I'm going to talk about, which spans the last 12 years or so, is really equally shared by two colleagues, Javier Castellanos and Jay Geed, who have worked with me during all of this time. And what I want to present is really an odyssey, because as we talk about nature and nurture, increasingly we really talk as when we think about behavior as the brain as an organ. And much research is driven by new advances in technology. And I really want to share our adventure, as psychiatrists, the three of us, in looking at development in relation to the growth of the brain, particularly across the ages that are particularly important to all of us, from what most of us have been discussing, from toddlerhood, perhaps, till young adulthood, which had been very understudied. What I want to show you is that it's possible, if you are lucky enough, to be able to work in a research environment that will support very long-term studies, that you can get a unique glimpse at the growth and behavior of an organ that most people don't have the privilege to be able to sit and watch. I'm going to mostly be showing you information about a brain and about brains of children. And what I'm not going to have time to tell you about is that we know a lot about all of these children's families, their childhood, their psychological processes. We know their birthdays. We celebrate them with them. And this will sadly be omitted in the interest of time. From this presentation. But what we set out to do was to see, first of all, what's changing in the brain across these formative years. We wanted to look at developmental trajectories in relation to the measures that we're interested in, whether it's gender, cognitive ability, or psychiatric status. We have some ways of looking. You've heard a lot about TWIN method, and I'll mention that in some studies from Dr. Plowman. And what we realize is if we didn't know about the healthy brain development, we wouldn't be able to have a standard in order to understand the problems, possible problems in that of our patients. I might also add that if you're going to do long-term studies, we felt we were limited to just anatomic measures because it's far more problematic. It's really hard to decide what's the same task in a functional study for a four-year-old and an 18-year-old. And so really what I'm gonna be talking about are just straightforward scans of the brain. This is a picture of an MRI machine, and I'm sure many of you know it. I hope not too many firsthand, but certainly this and the kind of pictures that you can get from this machine will be nothing new to you. So I wanna start off first with saying just what we learned from a very simple and straightforward technique of assembling large numbers of children and adolescents and asking them to come and simply have MRIs at the NIH and in return for which we would give them a copy of a picture of their brain. They would present these at class, give lectures on it, get credit in science class for it and so on. Now this is a map of the time course of the events that determine what makes up a human brain. And a great deal goes on before after conception in the early weeks and months. And I wanna point out that everything we have to do in the NIH study is just this little period here so that much of the migration of the nerves, the growth of synapses, much of this has happened long before we start our study. But there are very important processes still happening in the time period we study, such as the overgrowth and elimination of synapses, continued myelination of the brain, formation of dendrites, these are connections between the neurons. And so as you can see our hopes were rewarded that we were able to make observations of development even though we were starting for an embryologist very late. This is just a picture to show how lucky we were that colleagues and research nowadays includes larger and larger teams of people with very diverse skills. But I don't wanna dwell on this but this picture is to remind me to say how much we owe a team of collaborating colleagues, mostly engineers at McGill University where they had a automated sequence that allowed us to get automated measures on about 60 or 80 different parts of the brain. And if you have, as we do by now, over 5,000 brain MRI scans, you badly need a fully automated process as you can imagine. And this is just a scheme of what our designs look like. This is age across the bottom here. And these are just individuals. And in fact by now we have many more. This is just to give you the idea that people start at different ages and they come back every two years. And so by now this is a much more crowded slide but therefore we can put these different parts together and study development over a long time course. All right, this is the first picture of data. And you'll see a lot of slides that generally look like this where we plot the volume, what the measure, how many cubic millimeters there are, as opposed to age across the bottom. Here males and females are presented separately. And the first thing you can see which really startled us is how different people's brain size is. This is just the total brain volume and this is what the raw data looks like. As you can see at the same age for the same gender, people differ by almost 40% of brain value. And this is quite remarkable, even people of the same height and weight. And because there's so much noise in that kind of data, you get enormous power by following groups longitudinally and that's why we connected the dots who are the same people who come back so that you can see that you get much less variance and so you can make much more sensitive measures with longitudinal information. And what you see here is here's the slide with the raw data, but what you can see is, and in these slides, red is females and blue is males, there's some difference between the genders. And if you look carefully, you can see that this is really something of a curve. That the brain reaches a peak size overall in the early teens and then slightly declines, but the curves are parallel for males and for females. And that the ventricles slightly increase, that's the middle compartments of the brain where there's just spinal fluid, that the white matter continues to increase. The point being, yes, you can see overall changes in this age range. And one of the first things that we learned that we were impressed by was how different regions of the brain mature at different times. So these arrows here refer to the peak volume, so the frontal lobe and the parietal lobe mature, at least get to their peak volume several years ahead of the temporal lobe. This heterocronicity came back to be important to us later as you'll see. And what we think this reflects, this overgrowth and then decline, actually has been reported before by a wonderful work of Dr. Peter Huttenlocker, who is a neurologist in Chicago, who studied these very rare and unfortunate cases where you could study young people who had died from accidents and look at different ages for synapses in the brain. And what he had shown was not just that, yes, there's an overproduction and then a pruning down in adolescence, but he showed that different regions have different peak maximum points. So this notion of heterocronicity or difference in timing for different regions had been known and we think that's what this is reflecting in our scans. So this is just a summary to say the first thing we learned is different. When you talk about brain maturation, different parts seem to be developing at different speed. That the cerebellum is the latest of all, the so-called little brain, which is in the back of the head. And I'll say more about that later. These are all with absolutely healthy children who have no problems at the time we meet them. And what you can learn now that it's 10 years later, we have a real perspective on the brain growth in children with different disorders, but just with the healthy children looking longitudinally, we see some interesting things. This is just a very short movie to show you the growth that you can see over these years of an important tract of the brain called the corpus callosum, which connects the left and right halves of the brain. Oops, I seem to have gone past it. Let me go, let me go back. Well, I seem to be missing out on this movie. Maybe I can get Pat to come back and help me with this. What the point of this movie is, is that in this case in a subgroup of children who are in a subgroup of children, who are screened, I think it's going now. Thank you so much. Is that it? Yes. What you can see, thank you. As you can see from this movie in a subgroup of children, this part of the brain, the corpus callosum, really changes very dramatically across age. And if you are willing to wait for years at a time, you can make the kind of movies of growth of the brain the way you can watch plants grow in infrared light. This is just showing the dynamic changes and the nose is over here. So unexpectedly, this is showing development in a front to back fashion. That's not true everywhere in the brain. And so this was a puzzle and a surprise to us. This is as it gets later, the changes are moving back there. I'm not gonna take time for the rest of the movie, it shows slight gender differences and so on. But it's to give you a sense of the potential of this wonderful non-invasive technique and what if you're patient, what you can learn with time. So what we've learned so far was that prospective studies are much more sensitive, that clearly developmental trajectories are going to differ for different brain regions and that we think these relate to overproduction and pruning of synapses in the cortex. And more about that later. Another thing that we learned because we were interested in using this wonderfully powerful twin technique to look at brain development and that some things were just startlingly genetic. We could tell who were the twins and who were fraternal twins as opposed to identical twins just by looking, for example, at the shape of the corpus callosum. There's an arrow here showing the MZ twins, whereas this white matter tract, the one we showed you the movie of developing, looks very different for dizygotic twins. That's not true of all brain parts. And in fact, if you compare volumes between the mono and dizygotic twins, to our surprise, that last developing piece of the brain, the cerebellum, appears to have the smallest amount of heritability, in the sense it has the least difference between the mono and the dizygotic twin for volume. These are plots of comparing one twin against the other and therefore you can see the correlation for monozygotic versus the correlation for dizygotic twins. So already we were very intrigued. It appears that there may be one part of the brain that is, quote, less heritable and develops later on in life. And that of course gets you thinking immediately about what are the kinds of later skills and input that one might need that this part of the brain may subsume. We're in the process of trying to pin this down further. There are many parts of the cerebellum as there are subparts to all of these areas that I'm naming so simply. So this is the beginning of work, not the end of it. Clearly though, in terms of just the raw structure, it looks as if at least in a gross sense that genetic factors are gonna be generally very strong, although there'll be exceptions, that there are these really big regional differences in terms of development and how heredity things are, and that we're very interested in the cerebellum as an input possibly to environmental influences. One of the questions we asked in terms of this curve, you know this overproduction of synapses and the pruning down is supposed to, in some general sense at least it was always my model and that of many cognitive neuroscientists, that this overproduction and pruning down that late adolescence somehow gives you a leaner, meaner thinking machine by the time you're finished at that point that perhaps is more appropriate for the tasks you're going to do at a certain age. In terms of brain volume, when we measure IQ versus total brain size or the size of any part of the brain, the back, the front, the white matter, the gray matter, we get a low correlation between, in this particular case, a correlation of 0.35, but highly significant between the brain volume and the size. But then we thought, well, what about this overproduction? If you're much more gifted, are you more likely to prune more synapses, grow more synapses? And the answer is nothing that we can find. These are these developmental curves for the, we just divided our large group in half and this holds up now with double the size of subjects. If you look at these curves for whether it's the highest IQ or the lowest, this isn't a wide range because these were all children in the healthy range, but there isn't any hint that there's a different shape, there's a different level, because I've told you you have slightly different brain, slightly, but significantly different brain volume, but in the sense of whatever this production is, we can't seem to relate it to some obvious and simplistic measures of function. Similarly, the frontal lobe which subserves many aspects of attention, I'm not gonna show you this, but whether you're very attentive or very inattentive, these curves look just the same. So there's a lot that seems like structurally given here. This is just showing the same thing with white matter, even cerebellum. Another thing that you can do with these is to see what's changing during this age range and I don't, you can take the white matter and each voxel, which are these little units that scans are made out of, you can correlate them with qualities of your subjects and what we notice was with age what seems to change the most during our child adolescent age range is a particular part of the white matter which Eric Candell, you're not quizzed on these talks, but if you remember Dr. Candell's talk, he did talk about a very important white matter track that connects Broca's area and Wernicke's area and that when severed can make people not connect and not have meaningful discourse and this seems to be developing across our age range very strongly because that's what changes the most with age. Now I throw this just to show you the power and the wealth of what this kind of approach can do. Well now that I've given you a background in normal brain development, let me talk a little bit about the clinical groups that we have in parallel been applying this to and what's been so wonderful has been the opportunity to have yoked groups, very much the way you would yoke, cattle and field. We have groups that were matched for age and sex that were scanned along the same time so that we could always compare their development over the last years. These projects all started in 1989 and are really reaching fruition about now and so I first wanna talk about studies with hyperactive children, attention deficit hyperactivity disorder as it is called. Most of you at least know the label, you define it in terms of the way the definition is currently operationalized by at least six types of evidence of inattention and or hyperactivity and fidgetyness that starts at least before age seven. You can only make this diagnosis in competent professional hands if there's really clinically significant impairment in several settings and usually the people that in as I say in conservative hands, it is still a fairly common disorder being anywhere from three to 5% depending on how the surveys are made. And one of the things we were asking just simply, this is a very controversial syndrome as many psychiatric syndromes are, there's no diagnostic tests you do for it. And we were wondering if you would see any brain abnormalities in this disorder and what the developmental trajectories are. And we wanted to know also if these had anything to do with clinical outcomes. Now based on years of observing hyperactive children and of seeing what sort of tasks they do well and they do badly and cognitive tests, there are various schemes for what the abnormality is. And the most popular schemes have been the relationship between the cortex and the basal ganglia, the quadate and putamen, they sort of literally lower nerve knots inside the brain. And these were areas that we were particularly interested in measuring. And what we found, and this is with 100 hyperactive children and 100 controls, more so, what we found is that both for males and for females, there's a slightly but significantly lower volume for the hyperactive children. And this is about the same degree from boys and for girls, although the male brain is larger and that's why these two are different from those two. And we also found some abnormality in the quadate volume, the part that was hypothesized, both for males and for females. The findings aren't absolutely identical, but they're more the same than that. And one part of the cerebellum, a cerebellar vermis we actually found was very significantly smaller, both for males and for females. We were particularly interested in that part because of our interest in regional heritability. So this is just summarizing what I've just told you. I wanna emphasize that none of these are strong enough findings to be diagnostically useful. So whatever takeaway message there is, I want you to see this as an area of interesting and I hope important research, but if nobody you know who has a hyperactive child with one very rare exception, which I'll talk about in a minute, should go and get an MRI scan, that's absolutely certain. In terms of this diagram, therefore we certainly would include parts of the cerebellum, but basically it supported the notion what we had all along. Now as I've made it clear, we're doing these long term studies and one of the most striking things about hyperactivity in children is that about 50% by the time they're aged 18 to 20 are do much better and about the other half keep on having continued problems. So we said well what's happening to these measures as they get older? And this really represents because of how noisy the raw data is, these are curves formed by almost 500 scans. And this is the curve for the healthy volunteers in yellow and the green is the curve for the ADHD. And statistically this doesn't change, it's what we call a fixed abnormality, it's non-progressive. And of course one of the question that everybody asks us is what about medication effects could this be because they're taking a drug? But we were fortunate enough to be able to obtain children who are severely hyperactive, who have never had medication. And the curve is really the same, if anything. It looks perhaps worse in some years, but it's not really significantly different. So it's not a medication effect. And then of course you wanna know, well does either an abnormality at the beginning or the end have any prediction to how these children do? Because we followed all these children to find out, well are you in the half that got better or the half that got worse? And certainly these overall volumes don't seem predictive. So far they don't tell you that. But since we're still in the middle of these analyses, there's a hint that perhaps with respect to the basal ganglia, there may be some prediction for the girls. But basically no, this doesn't seem to do it. So in terms of what Dr. Maccabee was saying earlier, yes, this is one of the most convincing and by far the largest study that says yes, there's really a quotes subtle brain abnormality in hyperactive children. But on the other hand, it doesn't seem as if that's gonna have much to do with whether you're in the half that gets better or not. At least as we measure it now. Now one of the things that you've seen before in several people's presentation is that the twin measures give you an idea of how heritable a disorder is. And of all the child psychiatric disorders, maybe of all the disorders in psychiatry adults included, it looks like ADHD as we call hyperactivity these days is among the most heritable. And these are just various twin studies that make this point where the blue is monozygotic and the orange is dizygotic. The heritability looks like around 70% or higher. And we decided to try to look at environmental effects by searching, and this was one of the hardest searches we ever did for identical twins who were discordant for hyperactivity. That is where one twin was just fine and the other had ADHD. And as you can see from these statistics, this is incredibly hard to do. And what's more, it's probably, the heritability as we found out is probably even higher than this because some local twin studies try to send us discordant twins. But if you're doing very large twin studies, it's extremely hard to have clinical details on the non-hyperactive twins so-called. And we found that we got 100 referrals from people and from families of twins. We found out that by the time we boiled down this group, we had, and actually met them, it turned out that only about 15 pairs really met our definition of truly discordant where one had ADHD and the other didn't. Sometimes the so-called normal twin was really on a stimulant and the people who were sent us the twin study didn't know that. And other times they had a different disorder. And so we ended up doing what we thought was a very small study. But it was an important one because we're obliged to send all of our MRIs to a clinician. And so I can tell you that of the 400, just single-ten ADHD subjects that we had, 400 different MRI scans have been read as clinically normal. But when we started to do these discordant monopsychotic twins, we already had two just frank lesions in the affected twin. This is one where there probably is an old stroke that children can have very minor strokes even before they're born or around the time of birth. This is not common, but one of these was in this group of this very rare pair of 15-year-old twins, discordant for ADHD. So in other words, there was some abnormality and something special. If you have identical twins, you should get an MRI if one has ADHD and the other doesn't. A second known abnormality had a very big area here called caven septum polysidum, which is very unusually large, and when it's that large, does go with various behavioral abnormalities. So we found this, but even in the others for which there wasn't any lesion, these two, the affected twin, had a smaller frontal lobe, smaller caudate volume. Interestingly, they didn't differ in the cerebellum, which may be the quote's less genetic part of this disorder from our formulation. At any rate, so we do think phenocopies do occur, even though there's a very strong heritability, and they're however probably pretty rare in ordinary environments, and probably some aspect of this circuit of the basal ganglia seems to be particularly vulnerable to damage. Again, this like all of our other studies is pretty much still ongoing, and we are in the process of looking at the white matter tracks for this discordant pair sample to see where the wiring is wrong for the others for whom we don't have frank lesions. This is, as I say, just the beginning in a way because functional studies of selected groups I think will be particularly important. So this is the fact that we were interested in huge numbers in longitudinal studies that made us choose anatomy, but I think they'll be very interesting functional studies to localize which parts of the brain are really working in the most deviant way, and I think more functional correlates will be important for the future. In the last third of this talk, what I would like to do is talk about another group that has been particularly perplexing to us, and that is cases that shouldn't happen, which is like many psychiatric diseases, childhood schizophrenia really starts in typically late adolescence, early adulthood, and we had collected a sample and been assembling a sample over many years of children who have a disease that, quote, shouldn't happen because they became schizophrenic with an onset before their 13th birthday. We had a number of research and clinical reasons for wanting to look at this very ill group, but one of the advantages of studying a group in adolescence is you can look at the brain development in a group at that age. This is just epidemiologic data that happens to be from the UK but could have been from any of several other countries of the rates of admission with a diagnosis of schizophrenia. And as you can see, when you get down below the age of 14, and this is age at admission as opposed to rate, and once you get down below the age of 14, it becomes extremely rare. And even we, in studying this, and by now we have closer to 1500 referrals and have screened many children, and actually just a little bit over 70 children have participated in our very complex study to date. And I'm not gonna, of course, begin to tell you about the many other aspects of the study and the families and the relatives of these children. But what I can say is that in schizophrenia in general, what people see is that you have a smaller brain volume, larger ventricles, some medial temporal lobe structures, particularly the hippocampus may be smaller, and that in adult years, this isn't seen as progressive. When we looked at the childhood data, this was a striking that they have a smaller brain volume, even more so, and increased lateral ventricles, and it seems as if this seems somewhat more dramatic. Now, in terms of what forms schizophrenia or any disease, as you know, it's going to be a combination of certainly genes. There's evidence that pre- or pre-natal complications are important, and there's an important theory that synaptic pruning may have something to do with this. It's a selective decrease in the gray matter volume, what you see in schizophrenia, and this has been a hypothesized in adults as being a very important process. So we set about seeing what's happening during adolescence now that we know so much about the normal development and we know about the fixed lesion in the hyperactive children, and we set out to compare the schizophrenic children. And what you see here, the first thing we noticed was and the red are connected dots for schizophrenic children in some of our early studies, and the yellow are normal controls. Again, the dots connected because these are children who returned every two years, and the blue are children who were sent to us by clinicians and families who thought they were schizophrenic, but turned out to have another kind of chronic disorder which was a very severe kind of chronic depression, some becoming bipolar, but none had the very severe disorder of schizophrenia. And what we found is unlike what's seen in adults was a clear increase during these years. This is a plot of the volume of the brain ventricles, and this is baseline and onto-year re-scan. And then we realized that this is really major and this is not the usual plot, this is just the percent change between the ages of about 13 and 18. And we see unlike the ADHD and unlike certainly the healthy children which are in yellow, what we have here is a very impressive loss of tissue in the gray matter compared to the controls. And in contrast, this group of children in blue that are a mixture of children with some psychotic symptoms but basically long-term chronic depression and mood disorders, they're not showing this at all, even though they have the same medications as the schizophrenic children and the same cognitive level. And this is much more striking in adolescence than it is in adulthood. This is a slide of an effect size which is a statistical term to say how big an effect is the difference in a study. In our case, the difference in change between the healthy controls and the schizophrenics. And that's our study up here, the effect size, whereas a few other adult studies had shown a little bit of an effect but really not very much, whether it's for cortical gray loss or ventricular size in changing. During these adolescence, these children are not developing normally their neurologic exam, which improves with age for healthy children, as shown by the yellow dots. You may have fewer neurologic signs as you complete adolescence, but our patients, they don't change across this age range. They're learning less in school, understandably, because of their illness. And what you see, this is now the more familiar way of just volume plotted against age. You see this progressive loss, particularly in the frontal and temporal lobes. So this is a kind of summary of where we are now that we're finding that the longitudinal study is very important. We've talked about the strong genetic input in ADHD, but which appears to be a fixed early lesion that doesn't change. And now we see the schizophrenic children where particularly for loss of gray matter are very much changing across this age. One other technique of how we look at this has been very important to us, and that's been this another group at UCLA that has another way of showing the cortical change much more graphically in pictures that you recognize as actually what brains look like. And if you can subtract, because I told you that the controls come back every two years also, so we can subtract patients from their yoke controls at age 12, 14, 16, 18, and so on. And what's clear is that the schizophrenics are losing matter over time, a great deal more than are the healthy controls. But what was striking to us was that even where they were losing this, there was a real wave of change. When you take the young patients and subtract them from their controls, the difference is much more in the back of the brain. Five years later, it's all over. And so across adolescents, oh, and I should say that this is not happening with this contrast group either. So across adolescents, this is happening also on the medial side of the brain, you're seeing this progressive loss. And let's see if I get this right this time. This should be showing our movie. You can make a movie of this if you're willing to wait 12 years to do it. And this is the movie of the progressive change of what's happening in adolescence between our patient group, which is very different from what would happen with hyperactivity between the ages of about 13 and 18. But this is a time limited window in adolescence during which in normals, you're getting this overproduction of synapses and loss. What you're getting in the schizophrenics appears to be mostly just a decline. An interesting point here, and I hope I don't lose you on this, is that at least in studies in adults whereby comparing mono and dizygotic twins, they can see which parts of the brain disorder are more heritable. Remember, I've told you that our patients go from back to front. The most genetically determined aspects seem to be the certain parts of the frontal and the temporal lobe. Whereas what our patients do is start from a totally different part of the brain. We have a wild speculation that perhaps it was an environmental trigger elsewhere in the brain that sets it off. And yet, you end up with particular deficits in a region that appear to be most heritable. One question we're always asked is, well, do your children end up with dementia and neurologic disease because that's not characteristic of schizophrenia. And the answer seemed to us when we began, this rate of loss couldn't possibly continue. And we do indeed find out, this is statistically significant, though it may not look so, that this rate of loss as they get past age 20 does decline. And as our cases get older, we look like the adult cases. So it looks as if adolescence is a time limited window in which these very dramatic changes are occurring. And just to make things more complicated and to leave you, because I'm about getting toward the end of the enormous amount of data that I think I've perhaps been overwhelming you with, I want to leave you with a puzzle about this information because on the one hand, we have our more severe patients within our group of quite ill children have more loss. But on the other hand, when we look at our children who in our hands with the adjustment of medication and advice to the family and protracted hospitalization at the NIH, our group also improve over the years. We find that the ones with the more loss also are showing, I'm sure you wouldn't have voted this way if I asked you to raise your hands but are finding more clinical improvement. So what does this mean? Is it possible? Certainly we have evidence that genes are important in this. Synaptic pruning may be excessive in adolescence, but one question is what does any of this mean? It's clear that we can say that schizophrenia is a disease of abnormal brain development. And the late changes that we're showing here in adolescence, they can be a reflection of late maturation due to an early lesion, as some would suppose. I think that it's probably a later action of genes or some trigger that act later in development, even though there were clearly earlier impairments. And a third possibility, and these are not mutually exclusive at all, is that it may be a reaction to if you have an illness and your synapses aren't working right, perhaps this is a case where it is a plastic response to an illness, and it's trying to in fact get rid of the bad parts to help you improve better. I think all of these are possible. And so I leave you more with questions than with answers. It's clear there are strong genetic effects. It's clear there are many avenues for which you're going to have environmental effects, which may be our most important and most positive avenues for treatment, that you can see so-called fixed lesions, such as ADHD, or progressive lesions, such as with schizophrenia, that I don't think we can blame drugs for this, even though that's a big complication in psychiatric research, because the stimulant drugs didn't matter with the ADHD children's developmental abnormality. And for the schizophrenics, we had this contrast group of children who have chronic and very severe mood disorders, and they're not showing this at all, and many of them are on all of the same drugs as our patients. And I think that these are really just going to begin to let us know something about the candidate genes that give us clues for looking at in terms of what underlies these abnormalities. I just want to close by mentioning that in addition to Javier Castellanos and Jay Geed, who are really co-workers on all of these, that the Montreal Group and UCLA and geneticist colleagues are very important in these studies. Thank you very much. Like to invite our panel, reassemble here. We'll take questions from the audience again, of course. The nice thing about the NIA is you can work on a study for 12 years, and it takes that long to get it straight, and even then we mostly don't get it straight. I've been asked to announce that those of you who will be in search of coffee should look in the forum. That's the building to our left here. Since it's raining outside, coffee will be available inside. It will be in the forum, be up the stairs and into the next building. Oh, one other thing, too, this evening, we have our banquet starting at 6.30 and our talk is scheduled for eight o'clock. If you have a ticket for the banquet, you're welcome to come to the banquet hall in the marketplace. If you do not have a banquet ticket, you can still hear the talk at an overflow seating in Alumni Hall, which is in our Student Union building upstairs, and there'll be signs pointing you to that way. Okay, do any of the panelists have any questions or observations? Dr. Kagan. That was really lovely, and those are the kinds of data that we need that will build us toward understanding. I just have one question. If you remember the slide you presented of IQ against volume, and you presented the scatter plot, it's always been my experience that when you take abstract concepts like that, that if you get a correlation of 0.3, it's because 10% are showing it, and in that scatter plot, it's just those nine kids in the lower left. If you took out those nine, you wouldn't have much of a relationship. It's the ones with the very small volume. There have been about, no, I certainly agree with your observation of our very scattered data, but I can say that there have been over a dozen studies with very large numbers of adults, and they all get, ours correlation was 0.35, they all get between about 0.35 and about 0.40, and so there have been dozens of studies that look all exactly the same between volume and IQ. That's right, but suppose if we had all those studies, it was, they're always due to that 10% down at the low end. In other words, let's assume that that's honest, that is the correlation in nature, 0.3 to 0.4, but let's take all these studies and look at all the scatter plots. It could be that it's those 10% way down here, if you got rid of them, then there'd be no correlation. Is that a linear relation, in other words? Well, I think that's the nature of correlations in general. Of course we eliminate what are considered outliers, but from a statistical point of view, we wouldn't have any way of deciding that we should throw out one particular group than the other, because it's a fairly, it looked like the distribution in our group was rather normal, and so it would be hard to rationalize throwing out one piece rather than the other, and I think the distributions in the adult studies that I've read looked about the same. Do we have any other observations? I have a question, I mean, just why is the cerebellum take so long to develop? We don't know, in the twins, we started to look at certain obvious questions and were wrong on everything, the cerebellum's known for motor control, and so we gave exercise questionnaires to all the twins, and we thought maybe we'd have differences between ones who were couch potatoes as opposed to ones that were jocks, and we had a surprising number of monozygotic twins for whom that was the same, and the answer was not a hint, and we also looked to see if some were more rhythmic, rhythmic, you know, were some tapped dancers and others had two left feet, and the answer was forget about it, so we don't have a clue. I have a question here. How many different types of brain lesions are found in children with ADHD? I wanna stress that this was an extraordinarily rare event that we had to comb the United States, first of all, for identical twins who were discordant. As far as just the typical hyperactive child, we didn't find any in hundreds and hundreds of scans. We also found no chromosomal abnormalities. We did this as comparison for other groups, but there's no medical tests that we would recommend worth doing. We haven't found any in our ordinary hyperactive children. So questions that could you elaborate on what role prefrontal lobe dysfunction plays in ADHD, or is research focusing more on the basal ganglia? Right. Well, that's a very good question because it reminds me to stress more than maybe the circuit I showed did stress that people think in terms of circuitry and that the frontal lobe basal ganglia loops seem important in the few functional studies that have been done, the connectivity between them. And I don't think that any of the studies I know would particularly point to one or the other so far that they've even been enough to address that. I understand the corpus callosum is smaller in the male. What is its significance in male-female differences? I would have to pass on that. I don't know the answer. Anybody else wanna take a guess at that one? Eleanor's heard it too. I have a question here that this is slightly different. All I'm waiting for some more questions come here. So would everyone on the panel please share with us if they are parents themselves and if their children ever influenced by their research or if their children ever influenced their research or area of interest? Yes, I'm a parent. I have two boys and it's kind of an interesting question because one thing I've noticed about people's study attachment is they often don't have children. Disproportionately. Have you noticed that? So yes, I do have two children and one is adopted. I inherited him from a wife who died who had been by a former marriage. So you do experience the differences between genetically related children and those who aren't. So I guess they did influence me. I influenced my, one of my sons, who I really wanted to go into psychology by trying to get him to go into psychology and as a result, he's a journalist. My children were control subjects for some of our studies but they said that my husband and I worked too hard and they wanted to do something else. I have one daughter and I was too much of a psychologist. When she was two years old, we had returned from the Cincinnati Zoo and in those days, the gear shift was on the wheel and I put it in gear and the car was in the dry. She wanted to stay in the car so we went, my wife and I went in the house and two minutes later we heard an enormous noise. And as she obviously had released it and fell many, many feet, fortunately she was unhurt. So then I made an error. I was gonna be a good child psychologist. Should we talk about it or not? And I made the mistake of deciding, no, we'll not talk about it. She was unhurt and we didn't, nothing was mentioned for 25 years. And then one summer afternoon, I wasn't thinking of it and I said to her, we're having ice cream. I said, what's your earliest memory? And she shocked me by saying, daddy, you should have, we should have talked about it because I often thought about it and I never knew whether it was fantasy or reality. So, and now I'm gonna give you the insight about anybody who knows children understand this. I said, why did you release the break? I remember, she's two years old. She said, because you always took the car to the end of the blacktop of the driveway and this time you didn't. And I was trying, in other words, the child too is so sensitive to rules that she was trying to move the car to the edge of the blacktop. I have three adopted children. The eldest was adopted when she was 10 and the two littlest one, when they were each seven months old, they're very different from each other as I think one's natural children would be as well. But the issue of adoption came up. Your story is so interesting to me, Jerry, because my little daughter, when she was four, was walking along with me and one of her little friends and she said to her friend, do you know why I wear glasses? And her friend said, no. She said, well, it's because I can't see out of my near side. And then she said, and that's why you adopted me, isn't it, mommy? I thought, what? And I couldn't. Then I remembered that when she was two, I had taken her to the eye doctor and said, I've noticed that my little girl, when she looks at a picture book, his eyes right down on the book. And he said, oh, she's extremely near sight and he looked at me and at that time I wasn't wearing glasses. He said, I see this doesn't come from you. Is your husband very near-sighted? I said, well, she's adopted, so that's not relevant. Something was going around in her mind for two years until she was four. Oh, great. I'm childless. That's wrong. So you said, that's relationships. Oh, thank you all. Well, here's, we'll allow you to speculate a little bit here. Are the decreases in neuronal activity in the temporal lobe responsible for hallucinatory characteristics seen in schizophrenia? There have been some correlations that do relate specific symptoms in schizophrenia to changes and some have done for a particular part of a particular gyrus in the temporal lobe, particularly. There have been several studies that did that. In our particular study, we didn't find such a correlation, but they have been reported. Another one said, do premature babies make up the brain growth out of the uterus at the same rate they would have grown in utero? That's getting to be a big area. I know that in our hyperactive, in our healthy children, we didn't have children who were premature, but in the ADHD sample we did, and those with the smaller brain size were overrepresented with children who had been premature in that sample. I believe there are studies that suggests that within non-hyperactive children that there's a slight difference that stays long term, suggesting that in the very last weeks of gestation, there may be some extra growth factors that are important. Question, are there any studies on brain volume or brain matter in criminals? Not that I'm aware of using these newer techniques. Well, here's one. Do you believe that ADHD is being overdiagnosed and overtreated? If so, what might be the consequences and how could the diagnosis be made more specifically? I think that's a very important question, and there's different information that goes both ways. I know of certainly many stories where people feel that stimulants are being used in children who are relatively mild. When you look at some population studies, some epidemiologic studies, they tend to suggest that many hyperactive children, if not most, in fact, aren't being treated at all. So it's hard to know what's growing recognition and treatment as opposed to really overdiagnosis. I do think, though, that it's worrisome that today the HMOs will only pay physicians to prescribe medication and not to do other things like counseling. And so there may well be pressure on young doctors working for HMOs to feel like they should do something more. But I really wanna stress that the epidemiologic studies, I now do not suggest that there's a lot of other children that are very significant getting treated that shouldn't be. Well, here's a question again on ADHD. Why is it that stimulants are successful in reducing symptoms in ADHD? Stimulants are fascinating and really complex drugs because what they do is they focus on task behavior and they do it for everybody in general, whether they're hyperactive or not. They were used in the war for pilots and people who had long-term, tedious tasks. And hyperactive children, when they need to do reading or math, move less on stimulants. But if they're in gym class, like football or basketball, they move more. So it's more to do with focusing attention to the task and it's a very complex thing. As far as the biochemistry goes, dopamine is thought to be very important, but that's not set specifically pinned down. This question here, would you please discuss the heritability of obsessive-compulsive disorder? There've been very few twin studies and it appears to have some heritability. What's always fascinating to me about obsessive-compulsive disorder is the way, even when two identical twins have the disorder, the way the symptoms can be totally different. And I don't think anyone really understands symptom choice. I'll tell one anecdote, since I've given so much formal data, which is that we had identical twins in a study from a family that was not neat and not religious. And one of the twins washed all day and the other prayed all day. And each one thought the other was very peculiar. That's wonderful. That's wonderful. That's wonderful. Do you think childhood onset schizophrenia is related to viral infections, or here we have another one, allergies, nutrition? As far as the viral theories in adult schizophrenia, there are some very convincing studies that viral infection at a particular time of pregnancies, these done with very large numbers at a time of a flu epidemic. That may slightly increase the risk for schizophrenia in women pregnant at those times. In our very small sample, we don't think that's a risk factor by the crude measure of all their birth records because we looked at the birth records for all of our very early onset patients and they don't differ from that of their well-symptom siblings or of population norms. So for our cohort, I think there's some other mysterious and hopefully very interesting and to be discovered reason why they got sick so young. But for adults, I think most people think that that's been shown. Here's our follow-up question on this. Could you briefly review the literature on treating early psychotic symptoms in children with anti-psychotic medications with the idea of preventing the full-blown disorder? What do you think of such a treatment program? That's an area of very active research as to whether it's worthwhile taking very low doses and treating it. The very low doses of some of these anti-psychotics do appear safe and they appear quite effective. And so I don't know if you'd know what the rate are. Because schizophrenia isn't that common, you'd have to have very, very large numbers to be able to get a group that is not schizophrenic that you're sure are gonna become schizophrenic. Most of the people who would be in such an early study probably have mood disorders and the low dose of drugs would certainly work for them. So I think it's technically, in theory, it's an idea worth doing. But technically, I think it's very hard because you need such huge numbers of people, most of whom who will have something else. And whether that many people should take a drug for so long, I have something of a bias against it but the jury's not in. Any other questions from the panel here? In that case, speaking of stimulants, it's probably time for coffee at this point and so we'll adjourn until three o'clock.