 and these paintings specifically focus on the theme of emotions. So we always try to connect something to this meeting in terms of art or music, fun sometimes. So afterwards you will have a drink over there and you can also enjoy these lovely paintings. Now let's come to the topic of the talk of tonight. And this will be delivered by Peter Jonge. Peter is a professor of psychiatric epidemiology, I said it, at the University of Groningen in the north of the Netherlands. Now Peter is well known for his research on depression and somatic disease. That's where he basically started his research. And he really has some major publications in the field in top-tier journals. So he's really to me one of the people I look to when I want to know something about true science. It's always good to read a publication of his hand. And recently he has broadened his focus on depression in general, negative effects, emotional well-being and other forms of psychopathology at the University in Groningen. Now Peter also has a special affiliation with his university. He used to work with us on a part-time basin for quite some years and to a great pleasure to all of us. So it's a bit of a homecoming for him tonight. Finally I asked him what I also should mention and he wanted me to make sure to you that he's a nice guy. So I said it and I wouldn't say it if he wasn't truly a nice guy. Peter, the floor is yours. Well good evening everyone. Thank you Johan for the kind introduction and you're also a very kind and nice guy. I'm going to talk about emotions and well-being but I know I'm not really an expert on this topic. So I mean most of you will probably know more about this than I myself. So I thought instead I'll do a presentation on money, on nuts, aliens, Bush and Obama because everyone would like that probably. So to start with money I have no conflicts of interest to mention apart from some grants that I got that stimulated my research. But they were not from commercial parties and the grant providers had no role in the design or interpretation of the studies that I present. So first of all something about psychiatric epidemiology. Psychiatric epidemiology is the discipline that deals with the distribution of mental disorders in the population. And when you study the distribution of mental disorders you also look at risk factors of etiology, treatment and it becomes a bit broader. But the essence is the distribution of mental disorders in the population. So as a result I am called every now and then by journalists and they ask me well how many people in the Netherlands are depressed and how many people in the Netherlands have schizophrenia and all questions like this. And I should have an answer to that but basically I don't trust the answer that I have for that. And that is also the core of my presentation because what we have is the DSM. And the DSM came in this world of darkness and helped us to give answers to such questions that journalists may ask. How often do depressions occur in the population? The DSM, the Diagnostic Statistical Manual, was first published in the 50s which was by the way more or less at the same moment that the Dutch nuts factory was starting here pretty close by but that's probably a coincidence. The DSM has led to the first edition where there were 106 different diagnosis. DSM 2, 182, 265, 292, 365. So it is no surprise that DSM 5 which came out in 2014 has even more diagnosis. So it's sometimes a wonder that not everyone is having a mental disorder because we have so many to choose from. But not everyone has a mental disorder in the Netherlands. When you look at the lifetime prevalence of all disorders according to the DSM, 41% of the Dutch people will encounter a mental disorder in his or her life. In the past year this figure is already 23.3%. If you look specifically at mood disorders you will find 19% lifetime and 7.6% in the past year. Which is a lot. When you presented this to a general population people become pretty shocked and journalists also say things like yes but if there are so many disorders, this definition of disorders is that valid, is that good. It's a bit odd that almost the majority of, well not almost, but that 40% of the people will experience a mental disorder. That's a bit nuts in itself. Also there are many people who write that there is a depression epidemic or that we are all becoming more and more crazy in the world. Well if you just look objectively to the figures we have two studies, big epidemiological studies in the Netherlands that looked at this and there you find no indication whatsoever of that number of people who will experience a psychiatric disorder is increasing. It looks like it is decreasing in fact. You can also look at those figures worldwide or at least a bit wider than the Netherlands and you'll find figures like this. High prevalence rates particularly in the USA, 26%. This is past year by the way, 24.5% in Ukraine, very low percentages in Nigeria and Shanghai. These are studies that were done in the framework of the World Mental Health Survey initiative using the same methodology all over the world. And when you look at these figures you try to look at well why are these rates higher in America, Ukraine, Colombia and France and why are they lower in Nigeria and Shanghai? I don't know. And actually no one knows. And actually no one really trusts these figures because when you do those studies a year later you'll just find different prevalence rates. And it's even more, well I'll come to that later. What is good about these studies is those studies are well cited. This study by Demeter Nader et al. is about, cited about 100 times per year and I wish all my publications were so often cited. But I don't know really what we should do with it but therefore of course the introduction where you say mental disorders are very prevalent in the society and then you say Demeter Nader and he has another citation. But it is a bit frustrating that papers like this and most psychiatric epidemiological papers are a bit like this that they're cited so well while they're not really about content stupid figures about well 20%, 10%. Why is that so relevant for us? Well, I will first try to explain where these figures come from and then hopefully conclude with you that these figures are not to be overestimated. This is for example depression which is one of the most studied mental disorders in the literature. The DSM says well if you want to diagnose a depression you should have at least one of the two core symptoms depressed mood or diminished interest and in total at least five of those symptoms and when you look at this list of symptoms then you can conclude and ask several things. First of all you can calculate that there are 227 profiles that fulfill DSM criteria. You can ask yourself why is there a distinction between core symptoms and other symptoms? Why is there some score calculated and why is the cutoff of this some score at five and not at four or six? Do we want to use a cutoff score at all or should we just use the some score as a continuous variable? Why is a symptom like rumination not in this because you don't see so many people who are depressed and who are not ruminating? And how can one single disease, depression, produce such contradicting symptoms like weight gain or weight loss in hypersomnia to much sleep to little sleep? How is it possible that one single disorder can explain the presence of such symptoms? And that can be a very long answer to that but the short answer to that is this one. There is a book of abbreviations and first of all there is an abbreviation of those abbreviations which is called Alboa, a whole bunch of abbreviations but in this list there are all kinds of abbreviations that describe group processes and this is one of the favorite ones for me, Boksek, a bunch of guys sitting around guessing and basically that's what happened with the DSM. There are just a group of very qualified guys who would sit around the table and they would just start, well maybe we should have a disorder like depression and we call that depression if someone has this and this symptom and then someone says well yeah but then you should have rumination in it as well and then the other says well now let's not do rumination because we call that generalized anxiety disorder and it is a process where specialized people, authorities in the field come together and they decide it's not an empirical matter. It was not like there is something like depression and we just have to operationalize it. No, it was also a bit of an invention and you can look at it from different sides but I think what DSM basically did was changed the world from something like this to maybe something like this or maybe something like that. We can make diagnosis in a reliable way because we know what we mean with depression but we don't really know if it's valid, if it really represents something real. Before the DSM it was more like this. We also didn't know it was valid or not what psychiatrists and psychologists did but at least we knew that it was unreliable. Now we know that it's reliable to a certain extent but the question of course is is something that is operationalized in this way, something of interest, something that is representative of something real? Well, you can look at this negatively and then you have the pessimist induction which was described by Larry Lawden and he said, well, in the past there have been many mistakes before. So psychiatrists have described the term drapetomania which was a disease that occurred in slaves and that was that slaves all of a sudden had the strange fantasy that they wanted to be free. This was described seriously in 1851 and of course there was also a therapy for that which was you should whip the devil out of these people. There was one and if it didn't help then you should cut their big toes that they at least in some kind of preventative psychiatry that people would not run away even if they had this strange disease. Another one is homosexuality which was in DSM from 1952 to 1974 and then abandoned. What this says is not so much that DSM is worthless but that it's a cultural phenomenon that it represents something that is going on in the big world where also very ugly things happen. This was from my inaugural speech here a couple of years ago. I wish it was something like this that we could make a picture of someone put it on an x-ray and decide whether or not this was a depression or not but it still doesn't work in psychiatry like this and it's different in that aspect from many other somatic diseases. There are also somatic diseases where it's as fake as it is in psychiatry but in psychiatry just many things are vague. On the other hand when you Google when you look on Internet and you type in DSM and don't exist or something like this you will get pictures like this which to me actually signaled that it does exist and of course when you look around in the world there are many people who have very serious mental disorders and if you only look at DSM from such a cultural pessimistic point of view you might miss the real cases that are being detected people who have to collect all kinds of stuff in their house people who have to wash their hands 100 times a day people who are so depressed that they cannot do anything anymore without any reason people who have very strange ideas and bizarre representations of the reality these are real problems and DSM does try to categorize those problems so when you look at it from a philosophical point of view you could distinguish from a realist position where you say well things like depression mania, schizophrenia are just very real things like we have water molecules or whatever and DSM tries to describe those real things as good as possible on the other hand you could look at things from a constructivist way which says that basically all mental disorders are inventions are just cultural expressions of society and they are not real and of course which was argued recently also by Ken Kentler the position in between those two is probably the most fruitful which is a pragmatist position in which you for example also say well temperature is probably something real and I can measure it probably pretty well with a thermometer although I know there are all kinds of assumptions behind those assessments still I think that it's reasonable enough to make such assessments myself slightly even more on this side but to just say that it's a constructivist approach that is also well not very satisfactory and also has a tendency to underestimate all the suffering which is also related to mental disorders the take-home message this far is at least to be that any classification system but you can say the same about all psychological constructs in which we are interested in they should have the form of a falsifiable hypothesis and should not be considered as something that is necessarily true what I mean with that is probably everything that we are doing in psychiatry and psychology is work in progress is not any construct that we have worked with this far still needs to be elaborated and this is certainly true for DSM but it's true for many things that I encounter in my work well I'll just take a little bit to turn okay then we go to the more scientific part of this lecture one way to look at the validity of DSM is to look at it from a bit of a different perspective and what I want to do with you is to do a kind of thought experiment that there is a big bomb here at this moment all people are being killed an alien comes down with his ship and he wants to know what mental disorders are or at least how people think about mental disorders and of course he cannot ask anyone of you because you're all dead so the only thing he can do is look at the literature look at the handbooks and of course what he finds first is the DSM-5 and what does he find when he reads the DSM-5 he finds that mental disorders are apparently categorical things you can either have it or you can not have it the diagnosis of depression is either positive or negative mental disorders are characterized by complaints and symptoms and not by positive effect or well-being when you look at the DSM you will hardly find those words you will find something about disability but not so much about the positive side of things so mental disorders are at least something bad and they're based on symptoms the direction of dysfunction is often irrelevant like depression can lead to little sleep or too much sleep or to psychomotor retardation slowing down or being jumpy also he finds that depression and anxiety are probably independent entities because they're being described in completely different chapters mood disorders, anxiety disorders and five, the presence of mental disorders is largely defined on inter-individual differences so it does mean that you will get a mental disorder if you behave differently than someone else and not so much if you behave differently compared to yourself in another time these five findings I would like to challenge in the remainder of this talk and I'll do that with data from a study that we have recently done which is called How Nuts Are a Dutch which is a crowdsourcing study in which 13,000 participants assessed themselves and there's a cross-sectional part where 13,000 people participated but also a longitudinal part where more than 400 people assessed themselves for at least 60 to 90 times in a row during one month and so there are more than 25,000 assessments of people involved we have very recently published the baseline paper of this study and I want to stress that these data are publicly available if people are interested in using our data please contact me by now we have already 38 approved research proposals and I hope that this will be more because many people spend a lot of time in assessing themselves for us and for science this is the geographic distribution the density of the population of the Netherlands this is how our sample is distributed in the Netherlands and you can see that big cities like Groningen are even visible but also Tilburg is here and all cities that pop up here are really representative geographically it's a very well representative study of course like you can imagine in those kind of studies that women are over-representative and people with a higher education but we can't wait for such things and we do that in the results that we publish so why did people participate in this study? it was called a crowdsourcing study which means that people want to learn something by participating in this study what they got was they got something back as soon as they are filled in a questionnaire they would get back their own results and how they would compare to other people they would get a personality profile when they did participate in the longitudinal study they would do that with a mobile telephone and they would get back from us their pattern of emotions during the month but also how their mood and how their well-being was during specific activities and also they would get a personal etiological model or a representation of their own emotional life and I will not present this in detail but by now we have made an automated way of making such networks so that for every participant we have done this for more than 400 times we could make a personal etiological model which may or may not be helpful for a person to understand how they are feeling and we have also did a few publications where we described the method that we applied but I was tempted to go in detail on this but some other people will do that during this meeting also what I would like to do with you is to go to these five points that the alien found about the DSM and I want to refute them mental disorders are not categorical they are largely continuous all disorders that are in the DSM occur in milder levels in the general population and what you see typically with all symptoms is that there is a kind of right skewed distribution that it is not that one group of person has no depressive symptoms and one group of people has all the depressive symptoms but it's all a bit in between within individuals however disorders can sometimes fall from the sky but that is not really what DSM is describing for example when you... this is just very obvious if you administer the panas the positive and negative effect scale to a population you will find that most people are in this quadrant with a lot of positive effect not so many people are in this but still a significant amount but most importantly it's gradual it's not that there is one group here and one group there but it's shifting gradually through this whole field and if you know the level of negative effect of a person you still don't know much about the positive effect because there is still a lot of variation in one level of the other even if you administer a questionnaire the CAPE which measures the most bizarre symptoms in psychiatry, psychotic symptoms you will find that many people report symptoms that qualify as indicators for schizophrenia and psychosis people are not what they seem to be well most people say at least sometimes or often or almost always beliefs in telepathic communication almost half of the people believe in that being a very special, unusual person well I think here you can see that many people in our sample are from academia destined to be someone important 20% says yes conspiracies we see against you 100% says sometimes there is all the gradual distribution and even like hearing voices when alone 5% of the population does experience that what it tells me is that mental disorders even the most extreme ones are continuous and this is probably difficult to follow but I don't want to discuss all the details of that but we did a latent class analysis on all CAPE symptoms and what you find is all kinds of classes where only one group would qualify as relatively mental fit and that is reflecting about 10% of our sample and you will find I think this is the academic group of another 10% of the grandiose there are all kinds of names you can give to it but there are not so many people who have none of these symptoms that is basically the message of it and also that the classes that we find differ either in positive symptoms or in negative symptoms and it's not like only positive symptoms matter yes the second point was that DSM is full with complaints with symptoms and there is very little about well-being and in my opinion this is really not good and this is really a failure of DSM to account for a lot of variation that occurs in society for example what we find when we administer the DAS that is a measure to assess depression, anxiety and stress and you can qualify as normal, mild, moderate or severe and severe is like 5% of the population and what we did here was we looked at levels of happiness for those people of course C is people with severe depression and anxiety and stress on average have slightly lower levels of happiness but the most striking is that there is a huge variation in happiness in those people and it's not like if you have those symptoms that you cannot be happy anymore you can still be simultaneously happy and stressed and anxious and stressed and depressed sorry I meant happy anyway I think that if you would do something with this variation in happiness within such a category you could probably improve systems like DSM tremendously and this is also relevant because the kind of interventions that we do in psychiatry now are all meant to reduce symptoms and not so much to promote happiness and positive effects and I think there is a whole gap in the psychiatric literature we could fill with very relevant research the third point I wanted to make was the direction dysfunction which is often irrelevant in DSM 5 and for that I will my position is that this regard of compound symptoms hampers scientific progress and I will try to elaborate on that Mr. Zimmerman and colleagues described several years ago already that there are three types of symptoms mentioned in the DSM single criteria like low mood compound symptoms like in two forms like the compound opposite you could when you're having either weight gain or weight loss it would count as a depression symptom and compound related that is if you will feel worthless or guilty it would count as a symptom I am very much interested in this symptom because it's hard for me to imagine a disease that would give me both weight gain and weight loss when you go and talk to a cardiologist and you would say well I'm having this myocardial infarction and it produces either pain in my chest or a very pleasant feeling in my chest it would be a very strange situation and that is basically what DSM did here I discussed this with one of the people who were involved in DSM he said well people can have weight gain and weight loss at the same time but I don't know how you do that I just go in one direction personally anyway there are three of these kind of symptoms in depression weight gain or weight loss insomnia or hypersomnia psychomotor agitation or retardation I recently this paper was published by Femke Lamers who did a class analysis on all depression items and what she found was three different classes but basically those classes differed on appetite whether or not you had decreased or increased appetite whether or not you had weight gain or weight loss whether or not you had insomnia or hypersomnia the rest of those symptoms didn't really differ so what that analysis basically captured was the compound opposite symptoms of DSM which is interesting because what she did next was she described two kinds of depression melancholic depression which is just normal depression metabolic depression is depression that will result in increased appetite weight gain etc she compared those classes on several variables well obviously people who have increased weight have higher waist circumference BMI etc but more interesting is that they also have higher levels of inflammation while the more traditional melancholic depression have more cortisol dysregulations what she concluded from that is probably depression captures two different phenotypes and when you pile them up together you will get no correlations between etiological variables and when you disentangle them in more empirically interesting phenotypes you can make that visible to me it just says that DSM is on the wrong track by including compound opposite symptoms in their diagnostic system the fourth one was that depression and anxiety are described in two completely different chapters as if they were completely different disorders we all know that the reality is that there is comorbidity of more than 50% between major depressive episodes and generalized anxiety disorder for example so what we did was we have this data set in Groningen lifelines where we have a general population sample in which we administered the psychiatric interview which assesses all symptoms of anxiety and depression and interestingly for a large subgroup of those people there are no skips if you do a normal psychiatric interview and you don't have the core symptoms of depression you will go out of that part because we are so used to that depression consists of the core symptoms you cannot have a depression without low mood or lack of interest that you go out of the section immediately what we did was we did no skips we gave 16 internalizing symptoms to the people we administered those in addition we assessed the quality of life with the rent 36 as physical, social and work functioning because we wanted to do to see if there are specific subtypes of people with those 16 internalizing symptoms that would diverge as much as possible on disability and what we found was the following and again I don't want to really bore you with the details but show the general message which is one big group of people with no or hardly any symptoms some people who had some symptoms like sleeping problems and fatigue and those were people who had somatic disease basically there are some people who had anxiety symptoms the symptoms on the right side of the line are anxiety and on the left side there are depression symptoms there is one group of specific anxiety one group of people with anxiety and a little bit of depression one group of people with extreme anxiety and some depression and one group of extreme anxiety and some more depression but we didn't find a subgroup of people with only depression which to me was very puzzling and even by the way in this severe group 30% would qualify for depression or GHD GAD this also reminded me directly to some things that Eric Turner and later on Anelika Roest who actually comes from here and probably many of you will know her found when they evaluated the effects of antidepressants which are called antidepressants but I think they should revise their name into anti-everything medication because they work equally well for MDD for generalized anxiety disorder for panic for social anxiety for post-traumatic stress disorder for obsessive compulsive disorder there is no need to market them for depression only because they work equally well or equally not so well for all disorders in internalizing spectrum and what I meant with the last part that they don't work so well is that an effect size of 0.3, 0.4 is really disappointing because this was what Cohen meant once with effect sizes and he said well you know an effect size of 2 what we have for statins 2 or 3 that is interesting but 0.3 is really very small and when you compare antidepressants with active placebo that means pills where you also get some stomach problems from or things like that then the effect size is even getting less than it's about 0.2 so we are really talking about very small effect sizes which is probably related to the fact that we have very messy phenotypes. Another way to look at the effect is by the way explained variance an effect size of 0.3 translates to about 2% explained variance with antidepressants you can actively treat about 2% of the depression burden which you would not have treated when you have given a placebo also another way to look at it is when you give people either placebo or antidepressants and you look retrospectively at their distribution in only 12% you would be able to detect who has got the antidepressant and who has got the placebo so I'm sorry to disappoint you but those are really very small effects and it's even worse if you look at the heterogeneity this is for example when you visualize myrtesepine versus placebo which just has an average effect size you see that some people respond and some people just don't and it's the same in placebo and in Ramaron and when you combine them together you will find an effect size of 0.35 but when you would show this to an individual like what do you want I wouldn't be so certain if they would choose an antidepressant so does depression exist as an independent entity well of course first of all I want to make clear I don't want to say that depression doesn't exist or those complaints don't exist I know very well that many people suffer from symptoms that can be classified as a major depressive disorder but what I mean to say is that 30 years after introducing this concept we still know very little about its etiology the treatments that work for depression work for many things equally well are not so very well and even when we look at a big population samples depression was not identified as a disorder on its own but only as something secondary to anxiety disorders and particularly to generalized anxiety disorder there are hardly any people who are depressed and not ruminating which is the core symptom of GAD the last one that the alien found was that the presence of mental disorders is largely defined on inter-individual differences and I want to challenge that assumption that is made in DSM as well because I think that is often much more important to look at inter-individual variations than to look at inter-individual differences we are probably all aware that most of our current statistical methods like analysis of variance correlation coefficients regression analysis are based on the work from these gentlemen who said basically that the unit of interest in our research is not so much the individual but the race which is an interesting statement they both came from agricultural research and they wanted to look on the effect of specific things on the growth of letters or potatoes or things like that but these techniques were then later used in psychology and in social sciences in general but also in medicine and there the assumption that one potato is a bit the same as the other potato and this letter is the same as that letter doesn't really hold for people because people are so very different from each other at least that's what we think because we are one of the people and probably we are a bit biased by seeing that heterogeneity but still we are not primarily interested in what is in the race but we are actually very much interested in what happens in individuals and those people were actually also interested in eugenetics and stuff how you can make the race better by not letting some people reproduce and things like you don't want to be involved with but basically what they are interested in is describing a race and in psychotherapy in treatment of individuals you are always interested in what is going on within this individual and I think that is where we have to make a big correction to the kind of methods that we are using those people for example are the average Dutch people if you compare all photographs of people from the Netherlands you would get pictures like this the average male, the average female these are the average Americans, the male and the female the white I must say because it was the first question that I got once during when I show this why are they white probably the black people were not at home at that moment but I don't know this is what we actually do very much this is the course of depression from a sample of 267 people that were followed for 3 years on the 9 DSM depression symptoms and you can see this is a kind of spaghetti plot where all kinds of courses of depression are possible what we do with such data is we make a summary of those and we say depression starts with an extreme period and then there is a period of recovery and you have residual symptoms and that is the formal way of expressing what is going on depression is an episodic phenomenon with residual symptoms and this is the reality depression is different for everyone the course is different for everyone and we make summaries like this which are completely not useful for individuals anymore the alternative that we need is already described by Raymond Kettle in the 50s he said well instead of looking at two-dimensional data where we only look at persons and variables we should add a third dimension a temporal dimension and we should have a lot of temporal assessments of people in order to be able to capture fluctuations in things for example when you look at something easy like there is a relationship between depression and activity people who are more depressed become less active there is a negative correlation between those two that doesn't mean that that is true for all the individuals in one individual the difference can be the correlation can be this way and the other this way but when you summarize them you will get this kind of line what we wanted to do was to see if this is really happening in the kind of data that we have for example the most famous example that we had a couple of years ago was this one that when you take secretaries and you would correlate the number of typos that they make with the number of words you will find a negative correlation the more typos one makes the less minutes per word so when you would see that figure you would conclude well okay so I just need to make the people work faster type more words per minute then they will make less typos that's an obvious wrong conclusion because the reality is that within individuals the correlation is the other way around what I wanted to know is whether this kind of phenomenon that we actually reach the wrong conclusion when we look at data this way if that really happens with the kind of data that we have and it does well this is one example where we find the relation between physical activity and positive effect is positive on a population level but within individuals it's almost nonexistent we could do that because we had both the large sample of cross-sectional people and the prospective data also we found there is huge heterogeneity for example if you look at cortisol and negative effect you will find a very small correlation like mostly in the literature but when you look at individuals you would see that fluctuations in negative effects and cortisol are either positively or negatively correlated so everything we say on a population level doesn't really hold for individuals the same with physical activity and effect we find that with negative effect particularly for some people there would be a negative association with depression for some people there would be a positive association also for example between somatic symptoms and positive and negative effect this is even more striking in the sense that when you look at the population you will find that somatic symptoms are correlated with negative effect which is logical if you have a lot of somatic symptoms you will experience generally more negative effect but when you look at inter-individual fluctuations you will find a correlation of 0.1 within people and positive effect is then much more important so if you want to reduce somatic symptoms within people then it makes much more sense to address the positive effect and not the negative effect so here you could really make the wrong conclusion when you would look only on population level statistics while when you look at inter-individual statistics you would reach completely different or opposite conclusion this is a picture again from Ken Cantler who looked in this huge twin sibling database on the etiology of depression which is also done on a group level statistic this model explains about 20% of the etiology of depression so in reality it should be much more complicated even than this in individual this is completely meaningless this is only for maybe for policy makers or insurance people I don't know but it's not for clinicians who are interested in treating a person because basically what you do I'm sorry this is in Dutch I didn't have time to translate this in English but also second of all I borrowed this from my girlfriend Marijs Zardusma who made this picture suppose you have four different people with depression and one would have depression because she moves she doesn't have enough activities one because he has a very bad sleeping pattern one because he has very little social context and one because she has memory problems the etiology for all these individuals could be very simple but still what we do is we combine those four people in a non-existing person and we make a complex etiological model like this one while for one person he could become depressed because his wife died a couple of weeks ago and we make complex pictures like this that is really not the way we should proceed it's the same like when you have two presidents we make an average president of that which is also pretty meaningless in reality we are all different we all wear different colors of sweaters and we have different colors and hairstyles and whatever we often don't need to make averages of things that we don't want to average also when you look at intra-individual variations of mood you can come to very interesting new findings like we recently did when we looked at the individual courses of affect in people and what we found that for some people depression would behave like this some minor fluctuations around a specific set point while for other people it would really go like a lot of depressive symptoms recovery then recovery and what we did was we calculated by modality coefficient and if the coefficient that we borrowed somewhere from the statistics literature and we can actually qualify people as having this kind of gradual depression and this kind of sudden depression changes and our colleague Marika Vickers who I saw just sitting there and trying to hide is doing very interesting stuff on these kind of transitions in the context of dynamical models and probably these are the kind of depressions that we are really interested in and these are the kind of depressions that are actually very normal at least that would be my hypothesis also if there is time well it's 9 o'clock already very shortly mentioned that there are also methods very easy well not very easy but relatively easy methods to disentangle the heterogeneity of such three-dimensional data like we have the principal component analysis for two-dimensional data you have multi-mode PCA for three-dimensional data in which you can easily look what are the most important components of data that describe differences between people on symptoms and course and their interactions well I would like to finish with this and thank you all for your attention questions it is 9 o'clock but I would be happy to address them but otherwise people are free to leave of course there is a microphone but maybe you can now I can hear you say that it doesn't exist depression on itself it doesn't exist so I wonder what is real yeah well I carefully tried not to say depression does not exist of course there are many people with symptoms that are now classified as depression and I'm not saying that depression does not exist but it is not present as a single on a population level it is not visible as a single entity that's what I mean to say with that and so it may well be that there are interventions that help a person with a low mood or with a depression but the way psychiatry has organized the diagnostic system will actually hamper scientific progress because if you would classify things in the wrong way the kind of studies that we do is we compare healthy controls to people with depression and when I say well those depression that are selected for those groups are not really the kind of things that you see in the population then you have a problem because then you don't know how to generalize your findings to what is actually going on so if someone has a new intervention whatever intervention which is helpful it is very good of course and I would be very interested to learn more about it but I wonder if this is really evaluated on the right basis that would be my question okay well so we have the same question what is the role of the pharmaceutical industry in this because how more diseases the more drugs you can sell I think well there are many people who have a very specific opinion about that some people argue that even many people who were in the DSM committees had some ties with pharmaceutical companies what I basically see is that the scientists that I know are really trying to do as good as possible their own research but they are tied to this consensus based system of DSM I don't see many people who actually try to manipulate research findings to make pharmaceutical companies earn a lot of money but of course when a scientific pharmaceutical company funds research from a person I can imagine that there will be some kind of bias in those results so I really think it's a very bad idea that pharmaceutical companies fund research on the treatment of the drugs that they develop themselves I think personally that the state the government should organize such kind of research that you know 100% sure that this is unbiased and that there is not some kind of hidden financial interest in it so I don't know if this is really an answer but I do share your worries about that but what I see from scientists is that they really do their best but I don't think the system is very good in this well thank you for your presentation it was very enlightening I think but as a researcher the question is if we cannot work with averages which I mean was explained very well from you whether we left to work with for instance your own example on the efficiency of antidepressants that's based on averages too the size effect and I completely agree with your approach that we look at the person but to have statistical powers we need numbers so what are we going to do? Those are very smart questions I think one is we can study the optimal way of classifying psychiatric symptoms and by incorporating positive and negative symptoms for example the kind of techniques that we are developing where I work which have been developed already they need something like 60 or 90 assessments within a person so instead of getting your study power from the number of people you can also get them from a number of assessments and number of fluctuations within a person and you can if you have like I'll just say 100 people and 50 assessments you have in fact 50,000 points of data where you can do your statistics on but then at least you can explicitly look at where your heterogeneity lies and then you can see for example if you would make individual models for all your 50 people or 100 people and then you say hey for 20 of those I find the same kind of model so let's put them together and that is probably a meaningful subgroup and then you are a bit further on I think in defining interesting phenotypes Okay, Peter I would like to thank you once more for this very interesting and thought provoking presentation I have two small presents here for you I am trying to multitask now which is very hard for me as a man but first of all these are Dutch strobe paths but because you are being Dutch you will be well acquainted with it I don't know where you feel depressed or not or underweight or overweight but I also have I will get weight loss or weight gain from this well with this you will definitely get weight gain these are chocolates the upper layer is with liquid so you have to eat them at once or your fingers get a bit sticky it was a very thought provoking thought very high level and I would like to ask for a warm round of applause thank you very much and then finally we have drinks and if you still haven't had enough nuts there are nuts outside so be our guest thank you very much