Boris Birmaher My name is Boris Birmaher. I’m  one of the Child Psychiatrists who work here at   the University of Pittsburgh Medical Centre.  I specialise in any psychiatric disorders in   children and adults, but my research focus on  bipolar disorder, depression and anxiety. I   was asked to talk today about the challenge  of diagnosis of paediatric bipolar disorder. Then – let’s start, then, as many of you  know, it is true that paediatric bipolar   has been over-diagnosed, and at one time, was  over-diagnosed a lot. However, the opposite is   also true. There are many kids, especially the  younger ones, who have bipolar disorder and the   diagnosis is being missed, or misdiagnosed with  something else, and that you are going to say has   bad implications. And bipolar disorder usually  onset during the adolescent, the larger studies   so far in adults have – retrospectively,  have shown that around 60%, even 70%, of   bipolar disorder onset before age 21, 22, 23, and  there are many studies. There are, like, ten/15   studies showing exactly the same, and actually, a  substantial proportion also onset before age 12. Bipolar disorder, the prevalence is – in children  and adolescents, so bipolar 1 and bipolar 2,   is about one/2%, very similar to the adults.  Now, if we include something called bipolar NOS,   not otherwise specified, that basically,  is subsyndrome of symptoms of bipolar,   then the prevalence is 3% to 4%. Bipolar disorder  significantly affects the psychosocial function   of the child and the family where they  live, and increases risk for suicidality,   substance abuse, legal problems and, of course,  will affect the academic work of the child,   or work if they are beginning to work.  And the problem is that, despite that   it has significant and severe effects  on the child normal development,   the delays in the diagnosis are very common,  and most of the studies have shown that it can   be a five to ten years delay before the right  diagnosis and the right treatment is offered   to the child. Thus, it’s very important to  promptly, and correctly, I’m underlining   the word ‘correctly’, identify the symptoms of  bipolar disorder in children and adolescents. Then, in order to do this, I’m going to review the  problems, because it is exist, bipolar disorder,   but is not easy to diagnosis in many cases.  And I’m going to present the clinical picture,   and I’m going to, also, at the same  time, present the results of our studies,   and there are many studies in the field,  I’m going to concentrate in our studies,   because I don’t have enough time to review all  the literature, but let me start with doing some   examples. This is a real case, of a person  that is painting the different stages, like,   yellow is when this person is very manic,  black is when this person is very depressed,   and the middle is when this person is feeling  with mixed symptoms of mania and depression.   And is very telling, and this is a way  that the person can express the symptoms. Then this kid, for example, he’s  13-years-old when he came to see us,   he – the illness has started earl – much earlier,  and he – the father was the one who did this   graphic, and then, he presented at the beginning  of his life with mild episodes of depression,   and then he had a period of mania. When he was  in his periods of mania, he was running in the   school, being sexually inappropriate,  telling the Janitor how to do his work,   or the Principal how he needs to manage  the school. Very hyperactive, is sleeping,   like, two/three hours at night and  the next day he will do very well. And then, unfortunately, this kid had so many  problems at school that they called the Police,   and then he was expelled from the school. Once  we treat him and explained to the school what   was going on with this kid, he was accepted to  school, and then he did very well, he graduated.   And unfortunately, sometimes he stopped his  medications, specifically he was taking lithium,   and then he began to develop, again, depressions  and manias, and he has for several years this   problem of stopping the lithium and then having  these episodes, until he realised that he cannot   stop it. And now he’s, li – I think he’s 37 now  and he has been doing well with only lithium. Then this case was not difficult to diagnose, and  he’s 13-years-old, but if he was 25, many people   won’t have any problem diagnosing him as bipolar.  But because he was 13, people began to think – or   people thought that 13-year-old boys cannot have  bipolar, then he was misdiagnosed for a while. There are more complicated cases, and those  are the cases who have a lot of anger,   temper tantrums, and this is where people have  the problem diagnosing bipolar, and most of the   controversy about bipolar. This is one of the  cases, I didn’t – on purpose, I didn’t put the   age of this kid yet, and he consulted because he  has chronic mood and behaviour problems, severe   temper tantrums and ADHD, typical symptoms of  ADHD. And he was diagnosed with this disorder that   is very – I personally think it was premature to  put this in the DSM-5 called “DMDD,” and treated   for years with several medications and therapies,  with partial response, or very temporary response. Now, the kid, and also, his parents, reported  that above and beyond the symptoms of ADHD and the   behaviour problems, he has episodes of depression  that didn’t last the two weeks of DSM-5,   but six to ten days, an episo – and that during  the episodes of depression, he has the typical   symptoms of major depressive disorder, except  for the duration. Then he also have episodes that   lasted only two to five days, when he was angry,  irritable, aggressive again, above and beyond his   regular baseline symptoms, very agitated, silly,  talkative, intrusive, hypersexual behaviour,   sleeping only few hours, and the next day  he’s not tired, ideas of saving the world   and constantly writing notes. Let me show you the  notes quickly. Then, as you can see, he writes up   everything. It’s – but it’s, like, typical  symptoms of flight of ideas, of somebody who   is writing or drawing them. This is another one,  and he sits in the corner, writing hours, writing,   like, all these names and all these things, and  this only happen when he has – only happened when   he has these periods, that lasted to five days,  that he is very agitated, silly and talkative. Now, again, if this child was 20, 21, 22,  not – everyone would say, “I’m – I think he   has bipolar disorder,” but when this came to –  this kid came to see us, he was nine-years-old,   and he has history of these problems for the last  two years. There are many people didn’t believe   that he may have bipolar, and I think – well, he  was treated with stimulants and the ADHD respond,   but then he has recurrences of the ADHD. And  people were thinking that the ADHD medications   didn’t work – any longer work, like, he  developed tolerance, but what happened is,   he was having mood problems, or mood swings,  sometimes depressed, sometimes in the high   mood. He again was diagnosed with bipolar  disorder, plus the ADHD, and he responded,   also, very well to the combination of lithium  and an atypical – I don’t remember which one,   but he responded quite well, and in addition, he  needed the stimulants. He did that for a while.   Unfortunately, he – we lost contact with him,  but he responded well with treatment for a while. Now, these are the cases that are  more difficult to diagnose, and there   is so – there are several of reasons why that  it’s difficult to diagnose bipolar disorder in   children and adolescents. Similar to the adults,  the children also presents with different levels   of mood severity. It can be very mild to very  severe. They’re very – they are several subtypes,   so bipolar disorder 1, 2, with psychosis, not  otherwise specified, cyclothymia, and, also,   the patient can come to the clinic depressed,  manic, hypomanic, mixed, rapid cycling. Then   bipolar disorders in children, but those in  adults have different kind of presentations. Now, as I’m going to show, children tend to  have more mixed presentations and more rapid   changes in their mood, which are, even in adults,  more difficult to diagnose and treat. However,   in addition, in children, there is the effects of  development and symptom expression, for example,   what is grandiosity in a small kid? How you say,  where is the line between pathology and normal   behaviour? It’s difficult, because children  can tell you, “I’m the best football player,”   “I’m the best baseball player,” “I’m the best  singer,” and – or “the best ballerina,” and   this is normal. But when it’s beginning to be  too much and kids are acting upon these ideas,   and sometimes get hurt because they don’t know  the boundary in between what is real or not,   then you begin to be suspicious. Also, the same  I can say elation, or super happy and silly,   when do you draw the line? Then it  makes the diagnosis more difficult. Also, it’s hard for the kid to express  the symptoms, not like an adult, like,   the child is not going to come, “I’m acting  out because I am sad,” or, “Today, I’m elated,   euphoric.” It’s hard, then you will have  to be very patient and lower yourself   at the level of the child in order to  ascertain the symptoms. Another problem   that happening very often is that this  – like the kid that I show you recently,   this kid can have behaviour problems, ADHD, that  make the disorder more difficult to ascertain.   I’m going to show you and give you some tips  regarding the differential diagnosis. Finally,   the Clinician or Investigator interpretation of  the symptoms is – also can affect the diagnosis,   especially Clinicians who don’t believe  that bipolar disorder exists in children. Now, here is a way to help you with the  differential diagnosis. I only included mania,   ADHD, oppositional defiant disorder, ODD and  DMDD. Then in red are the symptoms that I think   can help you more, because if you see the other  symptoms in black, they are all over the place.   They’re not so specific, like, hyperactivity  can be in ADHD and mania, irritability happen   in all these disorders, temper tantrums in all  of them, behaviour problems in all of them,   depressive symptoms in all of them. And then  what happen is that sometimes people think that   for bipolar, for mania, are more  severe, but that’s hard to identify. And then one thing that you have to do is see if  the symptoms cluster. One symptom alone doesn’t   make the diagnosis. They need to cluster  and at least for our group, they have to   be episodic. Here, that’s the reason I put this  in italic, in capitals, because, at least for us,   they need to come and go, but they’re – sometimes  in order to see this, you have to follow the child   very carefully. And there’s symptoms that are a  little bit more specific, for example, decreased   needs for sleep. Suppose you see a child who’s  sleeping two/three hours at night, and the next   day they are not tired, and it happens several  nights, they are like that. They – you beg – not   necessarily it’s bipolar, but you begin to be  quite suspicious because who can do this? Any   teenager who doesn’t sleep at night, do you think  they will be able to be awake at school? Rarely. Also, with flight of ideas, the one, for  example, the real flight of ideas, like,   the ones who I present in the case, the kid he  is drawing and writing a lot. Hypersexuality is   also a red flag, especially in kids who have not  been exposed to anything related to sex or abuse   and they have episodic hypersexuality, together  with decreased needs for sleep, euphoria,   talking fast, silliness, and they come in groups.  Then you have to – and they’re episodic. Then you   have to rule out the possibility this kid  has bipolar disorder. Also, psychosis, and   schizophrenia is rare in kids, it happens but it’s  rare, then if you see a kid who is real, and I’m   underlining ‘real’ psychotic, you have to rule out  a mood disorder, both depression and also bipolar. The longitudinal course in a child is very  important, I said before, because, for example,   these are the baseline symptoms, let’s say,  can be normal, but also, can be kids, like,   I show you before with ADHD, oppositional  behaviour, anxiety, etc., and what we are   looking is changes in the symptoms. Like,  the ADHD’s getting worse periodically,   or the kid begin to be – is then anxious or is  depressed, and then begin to have these highs,   that they come and go. The issue is that  sometimes the highs can be two/three/four days,   and similar to the studies in adults, they are  sufficient to diagnose the kid with bipolar.   Because the DSM say fiv – four days for hypomania,  seven days for mania, but then, even in adults,   especially studies in Europe in adults, have  shown that even an adults with two/three days it’s   enough to affect their functioning, and increase  their risk to have the full-blown manic episodes. Now, that’s the reason that we were very  interested to follow these kids and see what   happened. Then, together with the University  of Los Angeles and the university in – Brown   in Rhode Island, and our university here in  Pittsburgh, 20 years ago, we developed this study,   “The Course and Outcome of Bipolar Disorder in  Youth,” and I would refer this – to this study   as “COBY.” And we decide to study to see what  happened with kids who are being referred to our   clinics with diagnosis of bipolar disorder. And  then we were able to recruit 438 kids, on average,   we follow them for 20 years, and the retention  has been, like, in this moment, like, 70%. We just – we finish the study, but the retention  was quite good across sites. We were able to have   153 kids with bipolar not otherwise specify, which  we were very interested to see what happened with   them over time, kids with bipolar 2, and most of  the sample were kids who fulfilled DSM-IV, when   we started the study, for bipolar 1. The BP-NOS,  the kids who we have our own criteria, because the   one of the DSM is very vague, and we didn’t want  very weak diagnosed kids with bipolar disorder.  Now, the mean age of these kids at the last  assessment was 26-years-old, some of them are 31,   32 in these days. And when we follow them  over time, 80% of the subject have at least   one recurrence, 60% experience two or more  recurrences, and most of the recurrences,   like, 75%, were depression, very similar to  what is described in adults. And they have,   on average, during all these follow-up, three  recovery periods, and they last on a – in a   median of 2.8 years, ranging from two months,  and some people 19 years doing very well. There was continuity when – because this is  a longitudinal study, so long – we can look   at continuity, when the kids were diagnosed  very early in life with bipolar disorder,   what happened with them over time.  And then, hmmm, there was continuity,   family history of mania predicted continuity of  adult bipolar 1 and 2. Half didn’t have mania   or hypomania in adulthood, but they continued  to have periods of subthreshold mania. Now,   remember, this is a naturalistic study,  then almost all of these kids are getting   treatment. Then there is an effect of  treatment that it will affect the outcome,   but they – some of these people, similar  to adults, will continue to have episodes. Now, because we are using some rating scales, that  we were able to follow these kids very closely,   we were able to calculate which percentage of  the follow-up, the kid – these participants,   were in euthymia, meaning, without mood symptoms,  and depression, syndromal symptoms, subsyndromal   symptoms, etc. And then, if you see here, and  this is a 16 year alan – 16 years analysis,   55% of the time the people were euthymic. However,  the rest of the time they have subsyndromal,   meaning also subclinical symptoms,  subsyndromal, and most of the time they   are subsyndromal symptoms, which they are also  super important because they affect the child   and as I’m going to show you, they increase  the risk for suicidality and substance abuse. Now, most of the symptoms were mixed, rapid  cycling and depression, and we were able to   compare the symptoms of the children who came  with diagnosis only of bipolar 1 with adults   who were participate in a different study, a  follow-up study in adults, also with bipolar   1. And then because both studies use the same  rating skills in something called ‘the life’,   we were able to compare the two studies,  and this is what we find. The children and   adolescents with bipolar 1 are in red and the  adults are in dark blue. And then, this – the   only significant finding is that the children and  adolescents with bipolar 1 have less time well,   they have much more mixed and rapid cycling  symptoms, and they have more polarity changes,   meaning more fluctuations in the mood. The rest  of the symptoms, and this is major depressive   episodes, mania episodes, subsyndromal symptoms,  there were no difference. And this is important   because then, if the kids have more mixed and  rapid symptoms and more polarity changes, one,   it’s more difficult to diagnose them, and  second, they tend not to respond well to   the mood stabilisers like the adults, and this is  what the randomised controlled trials are showing. Now, because they – when you follow-up people,  especially with bipolar, they are – the outcome   and what happen over time, is very heterogenous,  it’s not – they may be subgroup of people – we   decided to do an an – a special analysis, so see  if there is subgroup. And then what I’m going to   show is here in the Y axis, is the percentage of  time that they are euthymic, meaning without mood   symptoms, and here, the months of follow-up,  that go from six months ‘til 96 months. Then,   bear with me. There was a subgroup that was  predominantly euthymic, it’s, like, 24%. These   people are doing well, and actually, we think  this group, there is a subgroup of people who   do well even without medication, suggesting that  maybe some people, the bipolar disappear, which is   the same what happen with epilepsy in children.  Sometimes, when people have epilepsy, and when   you follow, when they’re adult, the epilepsy is  gone, the same can happen with Tourette’s, etc. There was, also, unfortunately, a group,  like 20%, who were ill most of the time,   a group that was in the middle, and an  interesting group that started poorly,   and we call “ill with improved course,”  also, like, 20%. They start as bad as the   other group and around three years,  they began to improve. And quickly,   because we – this is prospective, we know  what happen even before they began to improve,   and the two main factors that were associated  with improvement was that the parents who have   substance abuse began to improve, and, also, the  socioeconomical status of the parents improve.   Which is very interesting because if the parents  improve, then also help the kids improve, too. Now, what happen with the children with  bipolar not otherwise specified? This is   the kids who came with subthreshold manic  symptoms. At least in COBY, in our study,   the main reason that these kids were given the  diagnosis of NOS was because they didn’t have   the four days for hypomania and the seven days  for mania, but besides this, most of these kids   have almost all the symptoms. Now, these  children were very similar to the children   and adolescents with bipolar 1, in they  have similar poor social functioning,   present of the same – prevalence of the comorbid  disorders, the same risk for substance abuse and   suicidal a – behaviours, same family history  of mania. And as I show here, like, so far,   about 50% converted into bipolar 1 or 2, similar,  like, almost 30% bipolar 1, 26, bipolar 2, and   the main predictor was family history of bipolar.  Then here in red are the ones who didn’t convert,   and blue the ones who convert who had family  history of bipolar disorder. Then, again, many   of these kids have a lot of problems, comorbid  disorders, poor families, poor functioning,   etc., and several of them, almost half of  the sample, converted then bipolar 1 and 2. Now, the other things that we find,  as I said, many mood recurrences,   mainly depression, a lot of suicidal ideation  and attempts, substance abuse, comorbid disorder,   they have also metabolic syndrome, not only  because the use of medications. Even taken   into account the effects of the medication,  they tend to get obese. And they were exposed   to a lot of negative events, they have a lot of  sexual risk behaviour, there were increased risks   for sexual and physical abuse, many of them have  hospitalisation, and a lot of family conflicts. Regarding suicidal attempts, at intake, almost  40%, they already have suicidal attempts. You   can see if you add the part that is black  and grey and light grey, this is, like,   30%, and this is people with moderate, severe  and extreme suicidal attempts. That’s a lot,   and actually, bipolar disorder is one of  the psychiatric illness that increases   the most the risk for suicidality. And when  we continue to follow these kids over time,   and we add the 37%, is – half of  the sample have suicidal attempts,   and then they didn’t occur immediately, which  give us a window of opportunity for prevention. And then the same with substance  abuse, when they came to the study,   16% have substance abuse, and then, during  the follow, another 32%, then in – almost 50%,   and the new onsets were after three years.  Again, given, like, a period of time that   we can work to prevent substance abuse, because,  as you know, once they develop substance abuse,   the treatment and the prognosis gets more  complicated. Then if we can do the prevention,   it help with the prognosis of these children.  But here in United States, the most common was   use of marijuana, followed by alcohol, but 76%  of the kids who develop substance use disorder   were using two or more substances, and they  were no difference among bipolar 1, 2 and NOS. Now, the predictions of recurrence or  worse outcome was early age of onset,   which has been proved by many people, even an  adult studies, low socioeconomical status is   bad for everything, including bipolar  disorder, history of severe symptoms,   subthreshold symptoms as I told you before.  Pay attention to the subthreshold symptoms,   they are not, like, something to ignore.  People with more previous recurrence will have   more – high risk for mood recurrences. Exposure to  abuse, sexual or physical, is bad for everything,   too, but here also, in bipolar, increase the  risk that they have recurrences and do poorly. Comorbid disorders are suspected, and  family history of mania, and I said before,   family history also substance abuse, increase  the chance that the kids will do worse. Now,   all these things are very important for treatment,  because you then can treat the comorbid disorders.   you cannot treat the family history, but you can  tre – help the parents to improve both the mood   disorders and substance abuse, then improve the  mood disorders and the severity, it will help. Now, everything which I showed so far is for  the group as a whole, and it’s very informative,   and when you read papers, both in adults and  also, in child psychiatry, not only about bipolar,   about everything, when they present the average  for the group, it’s important, but it’s not   talking about the individual. Then we are very  interested in the person. If somebody walk into   your office we want – we know about the group as  a whole, but what about this person specifically?   And this is what is being called now  ‘personalised’, or ‘individualised medicine’,   and there’s not so much in psychiatry, but because  we have a large group, follow-up for 20 years, we   decided to do the same what they do in medicine,  and with something called the ‘risk calculator’.  And then or – and, basically, the risk  calculator is a mathematical model,   that you enter all the risk factors and then  yield a score, which tell you if a person is   at risk or not. Let me give you an example. For a  myocardial infarction, a heart attack, you can put   in the calculator the age, the sex, the race,  the severity of the blood pressure, the diet,   family history of the person, that does exercise  or not, put this here, and they give you a score.   Then the score can be you have a 5% to get a heart  attack in the next five or ten years, or you can   have a risk of 50% run and do something. And then  it’s very informative to do prevention, and, also,   for education to the per – the patient and the  family, and the same is being done for cancer,   many other illnesses in medicine. And I’m  sure if you ask your Physicians, they – I’m   sure that maybe when you go for your physical,  they are doing they are doing this for you. Then we did the same in – to predict recurrences.  The – we entered the risk factors known in the   literature that predict recurrences, and the risk  calculator was not bad. Then let me explain this,   and the Y axis is the area under the curve,  and then you can see here it goes from .65,   from – to .95, basically, it’s the  accuracy. If you have an accuracy of 95%,   it’s great. If you have an accuracy of  65, it’s not so great. And the – here in   the X axis is the months of predictor, and  then in – here in red is mania/hypomania,   this one in the middle is any polarity, and here  in the bottom is major depression. The accuracies,   in general, is 80%, which is very good,  actually, for mania/hypomania. It’s almost 90%,   for three years to five years, and  then this is very good, because,   for example, for heart attacks, it’s, like, 75%  accuracy. Then I think it’s working quite well. And then it’s available here in this website,  you can play with it, but when you have a risk   calculator, you don’t use it until is going to  be externally validated. What do I mean by that?   You need to have a different sample, hopefully by  different Investigators, and if they can reproduce   the findings, if they can reproduce the findings,  you begin to believe from this more, and actually,   this is what happened. I don’t – you remember  the study that I told you that we were able to   compare the longitudinal outcome between the COBY  sample and adult. Using the same sample of adults,   we calculate the accuracy of the risk calculator  and, actually, it was very similar. And COBY,   as I show you there, the accuracy was, like, 80%,  and, also, in the sample of adults, it was, like,   almost 80%, too. Then it was validated, and  then, it works for children and adolescents and   for adults. And I think this is an instrument,  you can see it in the website, the questions   to – you have to answer a questionnaire in order  to calculate the risk, and the questions are very   easy to ascertain. There is a small instrument  and then – which is also available, and then   it’s more or less the same questions that you ask  when you follow a patient with bipolar disorder. Now, you can also ask, “Does family history  help to clarify the diagnosis?” Because you   have the symptoms, suppose you have the  symptoms, they cluster, they’re episodic,   and you said, “Is family history helpful?” And  in order to answer this, we have another study   called “BIOS,” the “Pittsburgh Bipolar Offspring  Study,” and then, this is a study that also,   we started 20 years ago, and we have been  following these kids all the time with a 80%   retention. And what we did is, we compare  children of parents with bipolar disorder,   bipolar 1 and 2, with a random sample in the  community of parents without bipolar disorder,   and, also, the families don’t have bipolar  disorder, and, also, the children. And then   the control sample were normal, or they can have  psychiatric disorders, but not bipolar disorder. And I don’t have time to explain and show all  the results, but the – relevant for this talk   is the following. Then in the top, you have the  comparison of the children of the bipolar parents   in green and in orange, the controls, and the  top is bipolar 1, 2 and NOS, and the bottom is   only bipolar 1 and 2. Let’s go to the top graph.  Then in the axis – in the X axis is the age, and   then the Y axis is percent with bipolar spectrum  disorder. Then if you see then we – so far it’s,   like, 23% or 25% of the children with  bipolar, with – children of – I’m sorry,   children of parents with bipolar 1 and 2, so  far, 25% develop bipolar, and the controls is 3%,   which is very similar to all the community  studies. This is what you expect. But – and   begun to be significantly different, more or  less at age nine/ten, it begins to separate. But this is a very important clinical,  to answer the question I made before,   it’s 25% of the kids, it means 75% not. Then  not because the child has a behaviour problem,   so any emotional problems, and there is  family history of bipolar, it doesn’t   mean that because there is family served by  polar disorder, they have bipolar disorder.   The same if you look at bipolar 1 or 2, it’s,  like, 13% versus 1.5 is the orange line. Then,   again, then we were very interested and said,  “It’s only 25%, but can we really go and – a   kid who enters into the office, can we say who  of these 25% we can say really develop bipolar   disorder?” And then we did an analysis of  the symptoms and this is what we found. And I’m going to – you are going to see that I  divided the analysis, and orange is parents who   the bipolar disorder started late in life,  and late is relative, I’m talking about 21,   22, 23 years old, and parents who started their  illness very early in life, they – below age 21,   22, then when they were teenagers. As I said  before, early onset bipolar is a more severe   illness, so, more – yes, the prognosis is worse.  And the X axis is children with no symptoms,   children with anxiety/depression, children  with anxiety/depression and affective   lability or irritability, and children who have  anxiety/depression, affective lability and begin   to have some manic symptoms, not manic disorder.  A little bit of manic symptoms that come and go. Then let me share – show you what we found.  This is the one without symptoms, still,   if you have parents with early onset bipolar, it  can go up to 10%, if the children have anxiety and   depression only, anxiety, depression and affective  lability or irritability, and if the children   have all the symptoms. Let’s concentrate in the  last one, then look at the blue rectangle here,   is the parents who have early onset bipolar  disorder and the children have all these symptoms,   it’s almost 50%. 50% of the children may  develop bipolar disorder. It’s very high,   and then those are the children that  you have to pay a lot of attention. It has a lot of clinical implications. I don’t  have time to go into this, but for example,   if you see a child like this, who enter in the  clinic and they have severe anxiety symptoms,   and they don’t respond to cognitive behaviour  therapy for anxiety, do you give them the SSRIs?   Yes or not? This is children that you have to  make a decision if you are going to do the SSRIs,   and they’re not responding to anything, then  you will have to do it. But very careful,   you explain this to the child and to the parents,  you teach them the red flags for a manic symptoms,   and if they develop the manic symptoms, you  stop the SSRIs, and then you start treatment   for bipolar disorder. Perhaps when the bipolar  disorder is under control, you can come back   to the SSRIs. Anyway, I don’t have time to  go in this. If you want, you can email me,   or try to communicate with me and I will  be more than glad to answer your questions.  Now, remember I told you about the risk calculator  to predict recurrences? Because we were also very   interested about a single kid. This what I show  you is for the group, as a whole, then we decide   to do one for a particular child. Then we produce  a risk calculator, and this is the risk calculator   that we did. It’s also available at the same  website, but you have to be careful because it   has to be validated, and the accuracy is 73%. The  next slide I’m going to present something genetic   and we are going to publish – and there is a  paper in press that we have now, and then you can   look for it, then if we add the genetic to this  risk calculator, the accuracy increase to 80%. Now, this is what I mentioned before, is  there a biological test to help to clarify   the diagnosis? The an – short answer, no. Then  there are no x-rays, no imaging, blood tests,   anything that help, and in the past, we  were very, oh, maybe neuroimaging, not yet,   and maybe the genetic studies, not yet. But there  is something now, something called ‘polygenic risk   score’, then before I show our results, let me  explain what is the polygenic risk score. Like,   for example, most of the illnesses and many  traits, like height, are polygenic, meaning   that they are genetic, height is very genetic,  but it’s determined by many genes. For example,   for height, it is more than 1,000 genes. Then if  you take only one gene, the effect is very small,   but if you add them, if you add them,  the strength of the effect increases. And again, I don’t have enough time to explain  everything about polygenic risk score, but then,   you add the effect of all these genes and then,  you do some calculation and give you a score,   which talks about your risk to develop the  illness from the genetic point of view.   Using the Pittsburgh Bipolar Offspring Study,  because it’s a large sample, and we have children   of parents with bipolar, we have children of  control, and we collected DNA in all the sample,   and the sample is – like, in total, the sample  with parents and everything, is around 1,300   to 400 people. Then we were able to do the  calculations, and by the way, when you do the   polygenic risk score, you don’t need super huge  samples, like 50,000 people, 100,000 people, etc. And what we did with the polygenic risk score,  we can call – we were able to calculate the   polygenic risk score for bipolar disorder. Using  the data and genetic data, something called “GWAS   for Bipolar Disorder,” we also calculate the  polygenic risk score for depression, ADHD,   and schizophrenia, and those are the results. Let  me orient you to this graph. This is the parents,   not the children. Here is the risk score  and those are the scores, 0.2 and -2,   and in red is the parents with bipolar disorder  and blue is the parents without bipolar disorder.   And there were no difference in depression. There  was a little bit difference with schizophrenia,   but when you control for confounders, for many  other variables, the difference disappear. And   no difference in ADHD, but there was  a significant difference for bipolar,   bipolar polygenic risk score, and  actually, they separate quite nicely,   and it’s specific, because you don’t  find for polygenic risk scores. And this is the parents, let me show you the chi  – the offspring, the children of these parents,   it’s exactly the same. If you see, if I come  back to the parents and this is the children,   parents, children, and it’s almost like a copy  of each other, and the only significant is the   bipolar polygenic risk score. Can you use this  clinically? No, not yet, if the effects are small,   and then – but I think maybe in the future, when  there would be more genetic – big genetic studies,   and then see more genes. When we did this  study, we used, like, 64 genes rel – like,   associated with bipolar. Recently, there  was one study with 260 genes. We have   to redo all these analysis with the 260,  maybe this will be even more significant,   but we will have to wait. Then, unfortunately,  we don’t have any genetic studies that can – or   biological studies that can help us with the  diagnosis. Not yet, hopefully in the future. Then, in summary, bipolar disorder usually onset  during the adolescence, but as I said before,   can be manifested during the childhood. The  correct diagnoses require good clinical knowledge,   and super important, close and comprehensive  longitudinal assessment, and these days,   you can use apps to track the mood. We tell the  adolescents, “Track your mood in the app,” and we   tell the parents to track about their children and  the children about themself and is very easy with   all these applications that are for free. There  are many in the market. I usually tell the kids   to try until they find the one that they like.  You can track day-by-day the mood of these kids. The correct diagno – and, also, I’m sorry, I  forgot, and these days it is very easy to take   your phone and do a movie of the child. Then  if you think the child is in a manic episode,   make a movie and bring it to the Clinician  who is treating the child. Because sometimes   parents and children say, “Oh, I have a  manic episode,” then you see the movie   and it was not a manic episode, it was, like,  a temper outburst, or sometimes you see it,   and definitely you see the manic episode, that  you’re not going to see it in your office. The correct diagnosis is the most important factor  when selecting the most appropriate treatment,   and thus, preventing complications, like  suicidality and substance abuse, and the   correct diagnosis include the comorbid disorders.  Because if you don’t treat the comorbid disorders,   don’t expect that the child is going to improve,  and this is common in medicine. If you have   diabetes and hypertension, you don’t treat the  diabetes only, because then the person will do   poorly because the cardiovascular illness. You  treat both, and here, you have to treat both too. The use of risk calculators may be help to  identify who is at risk individually and   who is at risk to have more recurrence,  and hopefully, maybe one day we will use   these tools in order to follow the patients  more carefully. Family history of bipolar,   particularly early onset bipolar helps, but  not always means that the symptoms that the   child present represent bipolar disorder. Those  risk – the results of the bipolar polygenic risk   scores seems to be promising, but as I said  before, we cannot use them yet. And finally,   you have, also, to take into account  that many of the studies, including ours,   are mainly – they were mainly done with white  populations. then sometimes you cannot generalise   to other ethnic groups or cultures. But for  example, for the polygenic risk score, a study   that is going to be published very soon, they also  include Black populations and Asian populations,   and the results look quite similar, but we need  more studies including minority populations. I think with this I will finish. I want  to say thank you to the families that have   been participating. It’s not easy to do all these  rating scales, and like, for years, for 20 years,   like in the BIOS study, we follow the families  every other year, and in the COBY study,   on average, was yearly, and the co-operation was  wonderful. Then we owe to the help of all these   families and these families know that  the results can also help other people. Again, it – I’m presenting this, but  behind the scenes, like the movies,   there are hundreds of other people working, and  many hours to gather the data, maintain the data,   analyse, interpret and publish the data. All  the staff that work in the clinic help a lot,   and they were funding our studies, mainly the  National Institute of Mental Health, but also,   the Alicia Koplowitz Foundation of Spain. We pay  all the genetic studies with this Foundation,   and, also, the Fine Foundation in Pittsburgh.  This is our group in Pittsburgh, and again,   all of us were making contributions  to everything what I present. Okay, thank you. Again, if you have any  questions or anything, I’m available. Bye, bye.