 Good morning. Thank you. And thank you very much for including our work in your series. Because the kind of things that we are doing now really is a combination of the fields of epidemiology and genetics. I need the pointer, I think. Let's see where that went. Okay, thank you very much. Yeah. So this morning I'm going to do three things. I have three aims. The first is to summarize what we've learned about the epidemiology of mood disorders and specifically bipolar disorder in general population surveys now across the world. Second, to summarize what we've learned about the genetic epidemiology and to provide a summary of the advances in the application of the tools of molecular genetics to bipolar disorder. And finally, to describe the initial results of our large-scale family study of mood spectrum disorders that we have been conducting at the NIMH Intramural Program for the last decade. So to start with the impact of mood disorders in the context of neuropsychiatric disorders and chronic diseases that are non-communicable or non-infectious, we see here a slide from the World Health Organization that shows us the percent of life years lost to disability or premature mortality that's due to this range of the 10 most disabling diseases across the lifespan. And here we see in this slide that neuropsychiatric disorders, and this is a combination of disorders within the domains of neurology and psychiatry, explained 28% of the disability-adjusted life years across the world. And then we can go around this chart to show the impact of cardiovascular disease, 22%, cancer, 11%, non-communicable diseases and so forth. What this tells us is within the realm of neuropsychiatric disorders, we see that the major psychiatric disorders, schizophrenia, bipolar disorder, dementia, this would include Alzheimer's dementia, substance abuse disorders, epilepsy and so forth, are quite prevalent and now are surpassing the impact of the diseases that we've studied across our lives and thought were the most important culprits because these conditions are chronic, but they begin very early in life and are very disabling. So it's not that it's a relative impact statement that our field deserves more support than yours because as I'm going to show you later on today, the overlap across these fields is extremely important and it's going to be timely as we learn more about the interrelated processes of these conditions. So I wanted to show you the DSM-5 criteria. They're the new criteria that were initially released about a year ago today and I'll talk a little bit more about that today because it turns out that it's central to our studies of treatment and etiology and particular genetics. This has been a major challenge in our field that we have diseases without biological laboratory markers and they depend on clinical phenomenology and our designation is to whether the people actually meet criteria for these conditions. So we have several different subgroups of mood disorders. Bipolar one disorder is the traditional manic depressive disorder that we know about from popular press. It's become increasingly advertised in the population because of number of people who suffer from this condition that really does not mimic what we always thought were the manifestations of bipolar disorder as we expand our boundaries and understanding of these conditions. But for our criteria for bipolar disorder requires that they have either an elevated or irritable mood for more than one week and then four of the seven symptoms of mania. So it's the Chinese menu approach where we count the number of symptoms, does it last for a particular duration. We don't include recurrence as one of the criteria but from our studies and those of many others across the world recurrence turns out to be one of the most important criteria for these conditions because if people only meet these once they may be very impaired in fact leading to suicide or major life outcomes but the recurrence is really what leads to disability and they're the people who we have the most difficulty in our treatment sector. The second major disorder in the mood spectrum is bipolar two disorder which is a disorder that is lower in terms of its impact and severity. It's defined as having a hypomanic episode so that would be the same symptoms that we see for mania but less severe, less impairing and they last less than the full week of people with mania. It also requires impairment plus what's interesting here is require that the people have had a major depressive episode whereas for bipolar one disorder they do not have to have met criteria for a major depressive episode. And finally we have what's well known to all of us major depression which is defined somewhat arbitrarily as four of nine symptoms of depression, sleep difficulty, concentration problems, of course depressed mood and so forth. They have to have a duration of two weeks or more. Again that's arbitrary as we go into the population we can see can last varying lengths of time and that there has to be impairment that leads to some kind of difficulty in life functioning. Now the difficulty when we apply these criteria in the general population to people who are not coming to us for treatment has raised a number of questions about where these appropriate thresholds and boundaries should lie. So one of the major contributions of epidemiology in the last I would say two decades as to been to teach us what are the manifestations of these conditions in people who we don't see in the clinical settings. This slide shows the countries now that have participated in the large World Mental Health Initiative it's sponsored by the World Health Organization and Harvard University Ronald Kessler was the primary architect of this large study where we collected data across the world in a comparable fashion that we could actually compare rates of not only bipolar disorder but the full spectrum of mental disorders and physical disorders across the world. What we see here from these studies is that bipolar disorder the lifetime prevalence is about two percent higher than we originally believed but there's very high variability across the world. So in some of the African countries and East Asian countries we have very low rates whereas in the United States and Australia and some of the European countries we have the greatest rates that approximate two to three percent across the lifetime. These are a lot higher estimates than when I entered the field when we originally believed that manic depressive or bipolar illness was extremely rare condition. This study this summarizes the results from the first eleven countries and we published this in 2011. Now we've had a number of countries subsequently join the World Health Initiative and so the rates are becoming we have greater representation of countries across the world so that we can then start to look at differences by culture, ethnicity and most importantly we can look at treatment patterns across the world. I think the most important finding here is that the majority of people with bipolar illness in many countries do not receive treatment for bipolar illness in the mental health sector. Many of these people are diagnosed with major depressive episodes but the mania and the hypomania goes unrecognized. So this slide shows, this is a slide that is showing the National Comorbidity Adolescent Survey which was the first nationally representative survey of youth in the United States where we have representative samples across the country and we interviewed kids from the age of 13 to 18 to inquire about the full range of symptoms of mental disorders and again we also looked at a range of risk factors, correlates and most important we collected data on the full range of medical conditions and developmental conditions so that we could start to look at patterns of comorbidity between mood disorders, diabetes, cancer, allergies, asthma and these kinds of conditions because it turns out that these are going to be the most important contributors to disability and impact of these diseases. So what we found in this study was that the rates of bipolar illness were just manic episodes at this point, not full blown bipolar illness that would lead them to necessarily treatment but nevertheless they met criteria and suffered from impairment in some way. At ages 13 to 14, a little greater than 1.5% of youth met criteria for a manic history of a manic episode plus depression and across the adolescent development we saw a dramatic increase to the point that at age 17 to 18 the rates approximate those that we see in adults across the world. Originally when I was at Western Psychiatric Institute in Pittsburgh when we were studying bipolar illness in clinical samples and did not have data from general population surveys it was believed that bipolar illness did not occur in adolescents certainly not in children and as we go out into the population we see a number of these children suffering from the symptoms of bipolar illness that were not recognized. These youth were identified as having behavior problems, substance problems and a whole host of consequences of the agitation and the overactivity, difficulty thinking and so forth that are the core features of bipolar disorder. Now the other point I would like to make here, these are cross-sectional rates. So when we quote a rate of 4.5% at age 18 and this is just at the point when youth leave home they go to university or they go to military or into the job market. This is the time that is the peak period of onset for not only this condition but also for some of the other more serious mental disorders such as psychosis, substance abuse, eating disorders and so forth so this period of life turns out to be a critical period for us to see the emergence of these conditions and if we follow them and start to identify and detect these disorders at age 38 or so often times they have so many consequences of these conditions it's very difficult to retreat and prevent subsequent manifestations of these conditions. Now because these are cross-sectional and we cannot and for which proportion of kids are going to go on to manifest lifelong patterns of these conditions but we can couple this with a number of longitudinal studies, prospective longitudinal studies and indeed a substantial proportion of these youths who we see at age 17, 18 meet criteria for bipolar disorder do not go on to manifest bipolar disorder in early adulthood so some of these manifestations of many are time limited and it appears that some of them will mature out of some of these symptoms as they progress into early adulthood but then there's another substantial proportion probably 80% who go on to develop the consequences of bipolar illness as they move into young adult life so this slide shows a quote from the director of our institute Dr. Tom Insel who has been one of the leaders in admonishing clinicians and researchers to think about mental disorders beyond the diagnostic criteria that I showed you earlier where we describe the conditions as categorically yes and no the person either has bipolar illness or they don't they meet criteria for depression or not because when we look at these conditions and we start to understand the symptom manifestations and their biology, these do not behave the way our diagnostic criteria have emerged in psychiatry where we set up arbitrary thresholds for the number of symptoms, duration and so forth and those have been based primarily on clinical samples where when people come for treatment there's no question that they manifest five of nine symptoms they have it for two weeks or more, they have it a lot more than this so it's the group in the middle, it's those people who have some symptoms but they may not have severe impairment in their lives they may not last as long, they may be recurrent but only last for a few days it's those groups of people who it's not clear do they truly have a disorder or not and for healthcare and insurance purposes we have to code people according to the different conditions that they have but as we move into the population it's very clear that we're dealing with underlying dimensions of emotion, of activity, of concentration, of people using substances what is dependence what is not and so forth and we really need to be thinking about the underlying dimensional manifestations of the core processes and features underlying these conditions so from a research perspective and the kind of work I'll be telling you about today that's the kind of things that we've been studying to say well where do we, at what point does something become a disorder? I think it's very parallel to the debate in the field of hypertension where we've seen the thresholds for what is elevated systolic blood pressure, diastolic blood pressure at what point does that become hypertension and over the years that's changed dramatically so what is now seen as hypertension would not have been classified in the World Health Organization statistics as hypertension 20 years ago so if we're dealing with these thresholds it's very important that we have follow-up data that say at what point does this predict the person is going to progress to develop consequences of this condition so in the psychiatry now we really move to the drawing board to try to apply the dimensional measures of these phenomena so that we can understand more about their risk factors the underlying genetic and environmental factors that lead to the manifestations of these conditions so as I noted most of our evidence about treatment and about phenomenology in psychiatry has been based on clinical samples people who come to us because they've had serious manifestations of major mental disorders such as schizophrenia, bipolar disorder and so forth but what's happened in the last decade is there's been a dramatic increase in the gap between the manifestations of bipolar illness in the population and those who ever come to mental health care so we're not improving the proportion of people who get specialty treatment it's diverging and getting less likely that somebody with one of these major illnesses will be seen by a mental health specialist we have some statistics on that from our studies and here we show that of those with bipolar illness only 50% have any service contact at all where they have been diagnosed with bipolar illness and only about 30% have been seen in a specialty mental health sector so that doesn't even imply that they've seen a psychiatrist for the manifestations of bipolar illness and as we move to inpatient settings this really becomes about 5-10% of people yet most of what we know about this condition is based on the people who have been inpatient settings or have been in outpatient settings where they've participated in research programs so as this has occurred over time the big population surveys have shown that there's a large proportion of people out there who are not recognized with these conditions because if they appear in treatment it's because they're appearing in other sectors they may be identifying the education sector of course the criminal justice sector or often times internal medicine they may manifest with major depression but there is no capacity to do a full clinical work up to detect hidden bipolarity for example so if we go to clinical samples for example people who have been treated for major depression about half of those individuals will suffer from underlying manic or hypomanic symptoms and disorder as we move to the population it's a little bit lower but nevertheless much of the time the symptoms of mania and hypomania go unrecognized because for some people the symptoms of mania are very adaptive they have increased energy, increased activity they speak quickly, their thoughts go faster they say that they may be more creative but it sounds like this can be very positive but unfortunately people who suffer from a real manic episode really suffer from severe agitation because their thoughts are going so fast they're so active that they become agitated often times they're diagnosed as having an anxiety disorder particularly after they've been treated with an antidepressant it's like they have an achesia, reaction and so often times this could be that we have triggered mania in people in whom there is hidden bipolarity and without a full evaluation and some experience in detecting what is agitation, what are the secondary side effects that we don't need to worry about their progressing to a full manic episode I think it's really important that we may be missing such people if they never get to a mental health professional or we don't have consultation with people who can help us make that determination and again without biomarkers it's very difficult to do that and I think only through follow-up are we going to be able to really solve this issue so this is a recent paper that just was published on the Swedish registry that investigated people across Sweden with a prospective longitudinal follow-up of people in Sweden who were treated with major depression and what they showed there was a three-fold increase in the triggering of mania when people were treated with antidepressant monotherapy so among those people with major depression who had underlying bipolarity that was perhaps not recognized triggered manic episodes in a substantial portion of those people they then looked at people who were recognized as having bipolar who were also treated with a mood stabilizer and indeed the use of antidepressants led to a diminution in major depression over the following year now if we don't follow patients if we don't follow them along like they do in many countries where there are such registries or we can actually obtain such data we may never know of people who never come back to our setting that we've actually triggered an episode of mania and that these people start to get into difficulties with substance abuse which is a major consequence of bipolar disorder it's one of the major drivers of the development of alcohol dependence and other substance use conditions so these are very important observations we don't have data in the United States on this but again speaks to the importance of the comprehensive evaluation and the recognition of underlying mania and hypomania in individuals who come for treatment for either depression or anxiety the next I think the next major observation that was made primarily in population studies that was not well recognized in clinical studies was the extent to which people with any of the major mental disorders I could use schizophrenia as an example I could use an eating disorder almost any of the major conditions in our field we see that people do not only suffer from that disorder so even though we code them as a bipolar one, bipolar two psychotic, not otherwise specified and so forth the majority of people in the population suffer from three or more mental disorders across their lifetime so often times they first manifest anxiety disorders in childhood they may develop behavior problems conduct problems, attention deficit disorder and then in adulthood they may have a manic episode and then go on to have repeated depressive episodes over time so we're almost never dealing with somebody who just has one of these conditions and then when we move to physical conditions and that would be on any of the major chronic diseases we don't deal with acute infectious diseases here but now there is new information about the dramatic overlap between infectious diseases and the subsequent development of psychoses and major depression in adulthood and now that's a very important observation that is now being followed in fact neurodevelopmental disorders in children are related to infectious diseases that may occur during the prenatal era and with these large registries we can start to look at how these disorders are related over time what we don't know is what is causal and what isn't so if people have repeated infections is that a cause of depression or is it that people are just more susceptible to a range of disorders but there's substantial work on this in big population surveys particularly in Scandinavia and Europe where I think we're really learning a lot about these kinds of associations of infectious diseases that may help us understand their etiology as well so with people who have comorbid mental and physical disorders and most commonly we'll see diabetes and depression with bipolar illness we tend to see differential increased risk of cardiovascular disease for many years I've studied migraine and we rarely see people with migraine from one of these conditions over time it looks like it may be hyper autonomic arousal that leads to anxiety, depression, migraine a whole host of gastrointestinal conditions so there tends to be overlap in these conditions over time but what we see in terms of the impact to the public health is we see increased mortality there's several studies now showing that somebody with cardiovascular disease severe cardiovascular disease who also suffers from depression has a 5-fold increased risk of mortality compared to people either people with cardiovascular disease alone or depression alone so people in the Netherlands particularly and in Switzerland where I'm involved in several studies that are combined internal medicine and psychiatry are following people over time to look at the course and the interrelationships of these disorders over time but what we do know is increased mortality increased healthcare costs poor treatment response and perhaps that goes back to Alvin Feinstein's original definition of comorbidity when I was at Yale that's when he first presented this term of comorbidity and he was admonishing us when we're analyzing clinical trial data that we can't just look at the index disorder or we may have invalid inferences because a person with diabetes plus depression may have very differential treatment response to somebody who just has diabetes or depression alone and without characterizing comorbid conditions whether they're causal or not we are going to misinterpret the studies that we are doing systematic clinical trials so finally for the first part of my talk I wanted to summarize the importance of comorbidity for all of the mental disorders and its impact on healthcare the Robert Wood Johnson Foundation has developed the Synthesis Project where they're now investigating treatment recommendations for people who have comorbid disorders because what they found in terms of healthcare utilization and cost was people with comorbid disorders actually preoccupied the healthcare system so those with comorbid disorders were responsible for 80% of the healthcare costs and it may be that the secondary disorder was not recognized so that the person suffered had more severe manifestations of the second disorder or they didn't respond to treatment for the first but nevertheless it's these mental medical or physical comorbidity that really is one of the largest challenges we have in our healthcare system Okay now to move on to genetic epidemiology I'd like to summarize what we've learned about genetic epidemiology of bipolar illness in particular I know that my colleague Dr. Francis McMahon was here last week and I thought he was going to present a summary of bipolar illness because he's a world's expert in that area but his lecture on the ethics of genetic testing is very important it's a question that we have dealt with for many years long before we had genetic markers I've been studying families of people my entire career and people want to know what is the risk to my offspring and if we do identify genetic markers the more we know, the more we can predict risk and help people to prevent the manifestations of these diseases and treat people early if children do start to manifest these conditions or we can start to recognize them and let families know and form them of the risk of this disorder in their children so we've known a lot about the genetic epidemiology of psychiatric disorders this summarized what we've learned over 30 years and here this is a recent review that we did and here we see that the risk ratio the risk ratio of bipolar illness in the relatives of people with bipolar illness compared to those of controls this is absolutely critical that you have controls it's about 7 to 10 fold so if somebody comes someone with bipolar illness comes for treatment and wants to know the increase what is the risk we have a 10 fold increase now that is if the risk is about 1 to 2% the good news is it's only about 1 to 2% of the population only about 10% of the relatives will develop the disease so it's not in any way it's not like the autistoma dominant diseases such as Huntington's disease where we see about 50% risk so the risk is somewhat lower but nevertheless it's increased among relatives for major depression it's not as high it's about 2 to 3 fold increase in the relatives but major depression is extremely common as I'll show you up to 40% of people in the population will suffer from a major depressive episode at one or another point in their lives it's quite common but what is less common is severe major depression that's recurrent across the lifespan so we can adjust these estimates based on now a growing body of evidence from controlled family studies where we can really predict to some extent what the risk is at certain ages we have gender specific risk and even subtype specific risk of disorder based on evidence from familial and twin studies now the heritability of these conditions particularly bipolar illness has been known to be very high this is the proportion of variance that is attributable to genetic factors so across studies about 60% to 95% of the variance in some studies can be attributed to genetic risk factors based on our evidence from twin studies and large scale family studies where we look at the approximation of Mendelian patterns of transmission for major depression it's much lower and heritability is less than 50% and what this means these aren't telling us how many genes they're not telling us when they act or how they act but they just tell us there are genes underlying this condition and for bipolar illness it appears that there may be some major genes but these genes may be we can't detect them due to heterogeneity of this condition in the general population of samples that we've accumulated in the recent decade so this is a recent slide from Dr. McMahon and here we summarize the findings of the most recent GWAS studies those are the case control studies investigating systematically looking at genetic markers in cases of people with bipolar disorder compared to controls we now have 12,000 cases in these meta-analyses but unfortunately these are all the variants that have been associated with bipolar disorder in these studies but the only two that have been replicated particularly this one is the calcium channel marker on chromosome 12 and what's interesting about this and this has now been fairly well replicated so of all the markers this is really the only one that we have at hand that has been replicated in these large studies now what's interesting about this marker is it's also been implicated for a number of other psychiatric disorders particularly schizophrenia some of the studies of migraine accumulating studies GWAS studies also implicate a different form of the calcium channel gene and some of the epilepsy and so forth so this variant may be related in general to some of the brain diseases rather than being specific for bipolar illness but if I or anybody else who was studying bipolar disorder 20 years ago had been asked in the era of molecular genetics for which disorders are we most likely to find genetic variants we would have all predicted bipolar illness I believe because bipolar illness tends to be characterized by vertical transmission of families very high risk in relatives many of the families we saw really do look like they have a major gene because we see it across multiple generations we see it both vertically and horizontally when we study the pedigrees of people with bipolar illness now ironically we would have probably said schizophrenia was going to be less tractable to the tools of molecular genetics now what we have are probably yesterday people at the Brode who were summarizing the findings from the large collaborative studies they now believe they can explain 20% of the variants with this series of markers that have been identified and replicated for schizophrenia now the last point for the genetics summary that I'm providing today is this slide I just wanted to show you where the field is going now and some of the conundrums that we have those of us who are studying families studying the genetics of these conditions there was a paper that received a lot of attention because in this very large Swedish registry that I was referring to earlier in the antidepressant treatment study when they linked all the pedigrees of people in the registries they found that in the families of people with bipolar disorder they also tended to have increased risk of schizophrenia major depression autism, a number of other neuropsychiatric disorders so they suggested that there might be common risk factors for these conditions that lead to disparate manifestations over time there's heterogeneity of the underlying genes for these conditions now that's the opposite what many of us who have done systematic family studies for many years have actually found and what we see is there's specificity of mood disorders and psychotic disorders such that if we study families of people with schizophrenia we do not see increases in bipolar illness in the relatives so we don't tend to see this what we call cross aggregation of disorders which would imply that the underlying risk factors whether genetic or environmental are shared or common across these conditions but the truth is somewhere in between because we both have these based on these shared molecular markers as I just described the calcium channel genes and some of the other genes that may convey increased risk of neuropsychiatric disorders perhaps one could posit that in response to environmental exposures some kind of environmental hit that some people are more likely to develop neuropsychiatric disorders but it's very clear that we have both common genetic factors and environmental factors underline a range of these diseases when we look at environmental factors it becomes even less specific and now we're showing that some of the factors that we thought were unique to autism or schizophrenia are actually shared across autism and schizophrenia and indeed some of those factors such as season of birth migration a number of these non-specific environmental factors early infections perinatal factors and so forth actually are related to elevated risk of a whole host of conditions so at this point we have no specific environmental factors that predict the risk of bipolar illness but we do have some shared factors that will increase risk across all these conditions so what we have is both specificity where we do think we have some specific factors underline these conditions and we have non-specific genetic and environmental factors that must combine to lead to the risk of these conditions so in the last part of my talk I wanted to talk about the study we've been doing across the street and we're literally across the street we're in the Porter Neuroscience Center that people here say oh you mean the building with all the lights it's directly across from suburban hospital it is a laboratory building we've now done a large extension there's a very exciting energy now because the goal of the Porter Neuroscience Center is to bring together people who are studying brain from all kinds of perspectives from NHGRI the National Institute of Neurological Disease and Stroke NIMH, that's our group and others in the building we now have people from National Institute of Aging the I Institute National Institute of Child Health and Development and so forth so it's really exciting time to bring together some of the advances in the basic science that could be very important in helping us to understand why we have this overlap in mental and physical disorders and how it might lead to specific conditions so we started this study I was at Yale University for 20 years and I came to the NIH because at that time I predicted that we would really need very extensive studies of families and those of you in the genetics field know that over the last decade we haven't really been studying families that was what we used to do when we didn't we didn't have molecular biology to actually look at genes so we had to infer patterns of transmission such as those that you were taught in your high school biology and kids are now learning this much earlier by looking at patterns of expression in families so we can look at sex linked diseases hemophilia and these kinds of conditions where we just looked at the patterns of expression in families the traditional population genetics approaches but what we wanted to do our goal was to look at how not only how these disorders ran in families but we wanted to look at the underlying biologic processes and we wanted to work across the domains of medicine we wanted to work with neurology particularly neurology child health other fields so that we could get a comprehensive picture of the manifestation not only of mental disorders but also physical disorders so by coming to the NIH we could then devote effort and particularly methodologic development to characterize families from the local community so we did a very large sampling of the local community of the Bethesda Washington the whole Washington DC Virginia area and we've now collected about 500 families we have about 100 more in process so that we have a lot of evidence that can be used to address a lot of questions secondary to those of our primary interest by doing careful characterization of people at the NIH clinical center so let me just tell you some initial findings from that before I stop so what we do is we start with these very large population studies and our observations in those studies turned out to be extremely important in reconceptualizing how we think about bipolar disorder so the three findings in general population samples were first that bipolar disorder begins a lot earlier than we believed you start to see it manifest in adolescence secondly in adolescence bipolar illness really manifests with a change in activity and not mood so bipolar illness was not characterized by the person necessarily having high mood being energetic as if they were on cocaine rather what you see is a dramatic increase in activity and they're moving around a lot they're not sleeping it's quite objectively observed rather than relying on their internal representation of the state then when we went back to look at adults across the world because we have international data we found the increased activity seems to be the core feature of bipolar disorder and not a change in mood necessarily so that was the second observation and third that the manifestations of mania tend to be disaggregated from those of depression so mania and depression which was believed to be manic depression with one end we have depression the other end we have mania therefore the bipolar what we're going to show you in our family study is it is not a single underlying dimension at all we have major depression and we have mania and they're independent the etiology appears to be independent yet people with mania also tend to have depression so in terms of genetics in terms of treatment before people start to manifest this kind of switch that we see in people who have the disorder over time we can define mania as the core target of our studies and that has led us as I'll show you in a moment I'll show you a lot more translational work on the biology of activity in circadian rhythm and that it's a very exciting time for us because a lot is known about that so after we observed these kinds of things in the population we then apply the questions to our family studies here and then finally within our families we are examining intermediate phenotypes, biologic processes that may underlie by polar illness and a range of other disorders so I'll quickly summarize that for you and show you our major findings so we rely on the family study because we believe that if we study diseases within families and this is true across medicine you can assume that there's less heterogeneity within families than between families so if we're dealing with disorders about orphan mutations where one family for instance for hypertension you may see a number of distinct mutations in individual families but according to Richard Lifton at Yale they affect the kidney function so this is one of the ways that people look at these conditions, is it a specific mutation that affects all people with this condition or could we have individual mutations within families that accumulate and affect a particular system so we've been thinking that with the family study there's more homogeneity within families than between families where we believe heterogeneity is going to be one of the major challenges to our understanding of the diseases so we want to know what runs in the family did this guy here, did the dad who's looking at his kid look like this when he was in the nursery as well so our study just very quickly we collected a local community cohort we got a marketing survey almost 10 years ago we recruited people, we did not recruit them for bipolar illness we were interested in health and behavior because we're also interested in a range of other disorders I work with a lot of experts we're interested in migraine, we're interested in epilepsy and so forth so what we did was look at health and behavior we brought people in to the NIH interviewed their families and we were very interested particularly in mood disorders bipolar so we supplemented our sample to over-represent mood disorders from people who came to the NIH or volunteered for our studies so I'll just go ahead the major point I wanted to make here is that we evaluated mental and physical comorbidity in our studies we had to develop a lot of instrumentation because it just didn't exist so we systematically evaluate all the sleep disorders we've developed a tool to evaluate all of the sleep disorders in our population samples we developed a tool to collect data on headache syndromes and make differential diagnosis so we have a lot of tools for a lot of the more common medical disorders that are syndromic and just require us to collect symptom cluster data so the study is about 500 people there's a total of 3,500 relatives approximately we've interviewed about 1200 relatives of the 500 pro bands and we brought in to the clinical center about 500 people at this time so major finding is as follows so when we looked at mania we didn't look at bipolar disorder we looked at mania what we found was there was an 8 fold increase in mania in the relatives of people with mania likewise we found about a 2 fold increase in depression in the relatives of people with major depression now remember the summary of the genetic epidemiology this exactly replicates other studies in the field but the most important question was the cross aggregation because if depression and mania are underlying dimension what we would expect in families is that relatives of people with mania should have an increased risk of depression conversely relatives of people with depression should have an increased risk of mania and we did not find cross aggregation so what we found was there was not a significant link between mania and a pro band and depression in the relatives conversely there was no increase in mania rates of mania and the relatives with major depression this suggested independent transmission in families such that mania and depression probably do not result from the common underlying familial factors and this has been replicated in genetic studies as well now this study was a replication of our study but I just wanted to make the point that we also saw in this study that investigated schizophrenia and all the subtypes of psychosis we found the same thing psychosis breeds true in families so it's not schizoaffective type 1 schizoaffective type 2 all these subgroups what we see is there's a main effect of psychosis and psychosis and even the subtypes breed true in families so what we have here are three dimensions we have mania, psychosis and major depression and virtually none of the genetic studies have really conducted their phenotyping this way so we look at people with schizophrenia we look at people with depression we look at people with bipolar illness but nobody's actually looked at mania alone now very interestingly we started to look at the children of the pro bands in our study here in this region and we already start to see in people under 30 if we just look here the offspring of people with bipolar disorder 33% of them have already manifested spectrum bipolar disorder before the age of 30 compared to 12% of controls so we're already starting to see these manifest in adolescents and young adults okay so what we believe we have here is a deconstruction of the traditional concept of bipolar disorder in that we have mania, we have depression and we really if we're looking at the underlying manifestations we really ought to be looking at what leads to mania what leads to manic episodes what leads somebody to go into a period of activation and if anybody's experienced mania I understand there's many psychiatrist here when you see somebody with mania I mean it's pretty remarkable you see somebody moving along and then goes into an episode like this but I think what was most interesting to me was when I went back and started to look at traditional descriptions of mania in the early German Swiss literature Mendo, who was a neurologist and psychiatrist in 1855 said that this is a process of the brain where we have a speeding of all processes and the mood disorders do not have to be a core feature so this has now let us okay this has now let us to say well what is it about mania what is transmitted in families what leads to this activation so what we did was to look when we started this studies and I think what's unique about this family study so we call it next generation family studies was we brought people to the clinical center of the NIH and we examined the biologic processes that had been purported to be associated with bipolar disorder so what we did was bring people in and we studied all of the potential biologic features of this condition now many of them as you know look at the lower intensities under like a whole range of disorders infectious multiple sclerosis and so forth but at the time this was the only finding in imaging that discriminated people with bipolar disorder from controls we also looked at sensory perception particularly affection because we see changes in perception over time people with mania have increased sensory sensory thresholds actually we see changes in neurocognitive functioning so we've conducted that we see temperament changes is it some kind of temperamental factor and the two things that have been most exciting to us have been the reactivity where we actually challenge the system using the tilt table we do autonomic nervous system testing and we look at heart rate variability over 24 hours with halter monitors and then the startle reflex which is a well characterized brain pathway that is supposedly associated with anxiety traits and fears but this is the this is the group of measures that have most discriminated bipolar disorder in our preliminary analyses and what we see here we've looked at I just want to do this last slide we have used a novel approach to mobile technology where we use active watches I don't know how many of you have these Fitbit things that everybody's wearing but we use activity monitors on people and they wear them for two weeks so we can document what their activity is like and not only what it not is it averaged across the day how much average we look at variability we look at timing and so forth but what was unique about our study is we coupled it with electronic diaries so as when you do a halter monitor and you fill out a checklist of what you're doing at the time what we do with the electronic diaries is we can measure energy psychic energy, mood how much they're eating and so forth and so on so we can link together sleep, eating energy, mood and so forth so we're looking at these homeostatic systems and their regulation over two weeks and then we'll bring people back in to look at stability over time so these are the findings and this is my final set of findings before I summarize for you so what we've found from these analyses is that bipolar illness is associated with greater variability in sleep, activity and energy than other affected subgroups and controls when you look at them across two weeks when they are euthymic they're not in episodes and we've also looked at the effect of antidepressants have a huge effect on this they have much more variability much less stability of their circadian patterns second we see a circadian shift in activity towards later in the day and it's really pronounced in these people compared to people from the general population and even compared to people with major depression who do not manifest any of these changes and finally what we see is if you look at all the systems we look at sleep, we look at energy psychic energy, mood and activity and we see that these homeostatic systems in people with bipolar are much more interdependent such that if somebody who has bipolar disorder loses sleep you perturb one of the systems you see much greater shifts in the others so there's something about these homeostatic regulation that the whole system goes awry if you perturb one or the other system so it's suggesting two things that we should be looking at circadian rhythms in people with bipolar disorder as a primary disturbance that we should be understanding patterns of activity we don't know that's the only window it could also be the sleep it could be the regulation of how they eat and their mood and so forth and we were trying to put together these systems to kind of identify what the primary driver is at this point it looks like energy psychic energy is probably the driver and if somebody increases energy they have more activity and they have alleviation of depressed mood so it looks like we may want to increase help people to increase their energy if we look at our data rather than focus on depressed mood which may not even be the core feature of this condition so in summary from what I've told you today bipolar disorder exists on a spectrum it has major impact on morbidity and mortality so morbidity of both other mental disorders and physical disorder is pervasive and should be investigated the onset of mania tends to occur in adolescence and by the age of 30 about 80% of people will have manifest their first manic episode bipolar disorder is strongly heritable but the identification of specific markers is elusive and then in my last slide I think what does this convey to people studying other diseases because we're also looking at Parkinson's disease in our people's sleep disorders we have a lot of findings of people have other diseases but by dissecting the key components and their biologic underpinnings we may understand more about what the genes do and how environmental factors have impact on the underline biology and the population samples if we are in specialty psychiatric settings we really don't see the manifestations of this condition that may be typical in the population and then finally these objective measures I think are extremely important in understanding the manifestations of these diseases over time that I think really improve upon our ability to collect data when we see people monthly about how they have been doing across many domains so in the last slides I just wanted to cite the number of people who are my collaborators and you couldn't do this anywhere else than the NIH or the NIH intramural program or extramural program because we have required experts in all of these different domains to help us so these are our NIH collaborators we have lots of outside collaborators as well who have been helping and informing this work and finally my staff thank you very much yes right basically three questions one, can we discriminate bipolar 1 and bipolar 2 disorder so let me look at that first we have analyzed those data with our colleagues in the Netherlands and we found that bipolar 2 disorder does not seem to be on the same spectrum as bipolar 1 bipolar 2 disorder looks like they are more akin to depression but that's completely circular because to have bipolar 2 disorder you have to have had a major depressive episode so if we disaggregate the hypomania from that it's probably not different it's just of a lower magnitude the second would be the treatment is the treatment different for bipolar 1 and 2 the answer is yes or not for 1 and 2 but what we are now defining is the core feature of mania so it's not the depressed mood that's driving these conditions it's a lack of energy and fatigue which was described long ago by the group in England and at Columbia they differentially showed that monoline oxidizer inhibitors were more effective in people with bipolar 2 disorder and in fact they found that the tricyclics at the time they did not have SSRIs where people were characterized by the response to tricyclics so yeah there probably is heterogeneity there we do show in our study that there is familial aggregation of the atypical subtype now that leads to your third question which is the biologic measures of the inflammatory immune system so we have looked at all of those we are analyzing the data now and it does look like CRP is elevated in people with atypical depression so to the set that somebody with mania has atypical depression it looks like it is related to CRP inflammatory processes that our colleagues in Switzerland with very large studies now are looking at actual immune markers and there are several papers on this that I could share with you so yes looking at those perhaps it depends on how you treat also the inflammatory and the mental disorder but it really leads to specificity that is a terrible thing because epidemiology is defined by being broken down by sex and age my daughter who is down in the audience finished her training she had this little sweatshirt that says epidemiologists broken down by sex and age and I didn't mention that in this talk but basically we don't see sex differences in transmission so it is not related that we have differential transmission in men and women bipolar disorder is more equal in men and women than major depression which by far is more common in women an age I have already described I think it is a very important issue that what do we call bipolar illness without biologic markers you could say a kid has temper tantrums has bipolar illness sometimes the treatments I think where this emerged is children have aggressivity or many of these kids have major neurodevelopmental disorders that we don't talk about many of those children respond to things like the mood stabilizers and so forth so that led people in a circular fashion to say that is bipolar illness but we in our high risk studies the swiss high risk study that now has 500 children do not see bipolar illness emerge until adolescence so we don't see mania we may see anxiety but we don't see these neurodevelopmental disorders in fact these people the kids tend to have the characterizes having more intact neurologic functioning they do better on neuropsychological functioning so it looks like it is not an accident of some kind of insult so I agree with you it is what do you call this but these are the best studies to identify when it really begins okay yeah so that's been a great interest of ours for many many years and we've also very carefully looked at families of people with migraine in this sample because we are interested in the overlap and I think with migraine we're dealing with even more heterogeneity than we see for major depression we published a paper on the Zurich cohort study a follow up of 30 years of people with migraine and what you see just like depression we only see that about 5% of people continue to manifest migraine across their lifetime it's highly common in women who may have migraine for one year of their life and then they don't go on you never see them in treatment so there's also major differences in the subtypes so the kind of migraine that is characterized more by neurologic symptoms that's the subgroup that we wrote the first paper showing that later on they're at increased risk for stroke so long after they've seized having episodes of migraine they continue to be at increased risk for stroke in the absence of the usual suspect risk factors for migraine so we were recommending treating people with low dose aspirin even after they cease having migraine and by the way aspirin is one of the best treatments for migraine so I think when we look at the heterogeneity then we can understand more about its relationship to the bipolar illness but a very interesting question I think what happens is when you see people in clinics and I was in a clinic that was a bipolar clinic we saw probably a thousand people over a few years you tend to see the people with mania also develop depression so we're not sure exactly why that's related they're highly comorbid in fact it's almost impossible to see somebody with severe bipolar illness who has not also had depression so we're not saying that they don't have both because they tend to have both particularly those who suffer from severe manifestations but what this shows is that they are two independent pathways to get to the depression and bipolar it's not one disease so my mentor from Switzerland is coming here in early July and he was the first one to disentangle mania from bipolar from unipolar illness it's Jules Ochs from Switzerland who's now 87 and he says that in treatment settings you'll almost never see recurrent mania but in the population particularly if they don't have severe manifestations and they have the positive benefits of having mania you see a lot of people out there who have unipolar mania but many of those people may not come for treatment so I think that's going to be the issue I think in clinical settings we have to look at people with mania probably have depression yes now I think the one thing we got into the DSM was that people should examine changes in activity as one of the core features of mania and lots of has been written about this yes so there are a lot of people out there who probably I mean they may suffer they have family difficulties they may have a lot of problems but they don't get into treatment because many people self-medicate that's the other major point prospective studies by far show this is one of the most causal mechanisms we see in our field that people with bipolar illness mania self-medicate with alcohol so if you follow them over time they may have been having mania that you will never distinguish from their own pharmacology you will never be able to go back and retrospectively determine that they were suffering from mania but they treat themselves during the manic phase because they're agitated alcohol really diminishes that oh good