Depression: What (if anything) is it and what are its evolutionary origins

Daniel thank you very much for coming along we look forward to hearing it it's going to be a depression afternoon I think hopefully not depressive so thank you very much and thank you very much everybody I'm I was nervous following Marcus because I thought of potential overlap in our talks and also that he would present it very beautifully and very thoroughly and with many interesting and broad claims and indeed that proved true so thank him for his talk in a way it said both the disadvantaged going second but it gives me a chance to you know to bounce off some of the things you've already said my talk is much more philosophical really I have interested in evolutionary medicine but also in conceptual issues to do with what does it mean to say that someone has a disorder and and how do we know when normality stops and disorder begins which is not by any means a straightforward problem particularly in psychiatry and actually relates to evolutionary problems as well because what does it mean to say something is an adaptation and why does it stop being adaptation and it starts being the adaptation gone wrong or overly expressed or something so these boundary issues arise so this will be a sort of much lighter and more neurotic revisiting of some of the issues that Marcus talked about and I would go further further than Marcus in splitting up depression in the way that I think you'll see so why are we why are we even talking about this well one of the reasons we're talking is that a major source of morbidity in all populations over the world is depressive disorder which is this is just from the non communicable disease collaboration data many many thousands of life years are lost or you know of disability life years have lost a disability through depression which is the blue bars there and that's true in every continent we've ever studied largely because people perhaps aren't hung Huynh together most of those continents and if you add the red bars which are anxiety disorders then you really this is the major source of psychiatric and neurological morbidity so that's why we're all here right because there's a lot of this ill-health that that's coming to our attention and that's not just you know a thing of Europe would nor thha merica that seems to be a thing that's that's very generally true now of course for many diseases dementia would be one that are affecting populations we might say well we wouldn't really expect evolutionary considerations to be very relevant because the force of natural selection declines with age so things start to kill you by the time you're 75 probably have very little evolutionary relevance simply because it's an observation known to or made originally by George C Williams in the 1960s natural selection isn't very good at dealing with things that start to happen after you've done most of your reproduction because in this in effect it can't see them of course we might might do a bit of grandparenting and things like that it might make a difference you know to your inclusive fitness whether you were still in good shape to be a grandparent but at an age when most of the population in over previous centuries would have died natural selection has no force right so we would expect as life expectancies are increased through social and cultural means to uncover a huge reservoir of disease that's just been sitting there but the selection has had no impact on and to some extent though not entirely that may be true of things like dementia but my point here is this is not true of depression right because depression if you look at where the life years are lost and depression is the blue curve on here the onset is about the onset of reproductive capacity you know in a big way and actually the peak the peak loss or the people morbidity is exactly at the time when people are in midlife they're having their kids and so on particularly amongst women so you know if this is a disorder it's a disorder that you've Lucien ought to have something to say about right because it's the time when you're you're sort of achieving your reproductive output so there's an interesting thing with something should be apparently so harmful so prevalent and acting exactly at the time of life you think you know natural selection would have were acted on making people be well even at the cost of of possibly senescent later in life let's remind ourselves what we're talking about these are the dsm-5 diagnostic criteria for a major depression the first two as you know are cardinal criteria that you have to have one of which is either depressed mood I seem to have a circular that the symptom of a depression would be depressed mood but there you go I think you all know what we mean the emotional pain of overwhelming negative effect and had only are the loss of interest in in previously pleasurable activities and then and at least so at least some of the following list I just want to draw attention to some features of this because it will become important right so they're things like changes in appetite or therefore body weight changes in sleep psychomotor changes so either retardation you know you can't drag yourself around or agitation you can't sit still fatigue loss of concentration feelings of worthlessness or excessive guilt and suicidal ideation and for for dsm-5 diagnosis of depression you have to have five of these including one of the firsts to couple of points to make with the possible exception of suicidal ideation I'm not sure about that one all of these are quite normal okay it would be a very strange person indeed who didn't occasionally lose a night's sleep or you know feel they couldn't get themselves out of bed in the morning it will be a very strange person indeed who didn't feel fatigued sometimes it will be a very strange person indeed to have days when they couldn't seem to concentrate or data when they just felt blue for example loss of a spouse or something that Marcus has already mentioned so that's the first thing the second thing is there are bi-directional symptoms in there okay so the sleep changes can either be that you sleep too much or that you don't sleep enough or also that actually that you sleep the right amount but at the wrong time of day so actually it's quite a permissive sort of criterion the appetite or weight changes can be either you eat too much or you eat too little so actually we start to see this is a pretty permissive syndrome you have a set of things all of which with the possible exception of serious ideation are actually normal and some of which you don't even pre specify which direction we go the order for symptom of malaria will be you either be too hot or you'll be too cold right I mean that's really how fever works so already you know we have sense this is quite a curious sort of syndrome given that all of these are perfectly normal what we rely on is something about the number of them and their severity to say that there's a disorder here when you know when you wouldn't you know someone just feeling down for a day you wouldn't yet say that they had a depressive disorder I'll come back to that in more detail but just already kind of flagged up some of the issues we're going to talk about I'm not sure that this slide is perhaps as legible as I would have hoped so I'll talk you through it are there evolutionary explanations for depression well yes there are plenty in out in the literature and yeah so summer sent going over ground that Marcus has already touched upon but let me just review them before I take a slightly more critical perspective on them we've got a lot of evolutionary explanations out there in the literature actually perhaps a fan starting for no particular reason one side of the circle one key explanation is to do with sickness behavior so we know that animals that are suffering from infectious disease or just are in a poor state otherwise one of the things they do is they want to take themselves away and hide under a bush they don't want to do anything they don't want to eat they don't want to you know exert themselves in any ways and there's plenty of straightforward adaptations that what's going on there is that really the only useful thing you can do in that circumstance is let your immune system get on with the job so don't expose yourself to anything else don't expose your kin to whatever infection you've got don't try and do anything with it at a time when your your energy just needs to be saved for for mounting an immune response and I'm given the the enormous growth in information that shows that depression in many people has a big immune component you can see the link with sickness behavior so one of the leading evolutionary explanations for depression is to do with sickness behavior and as Marcus has already talked about inflammation and immune system activity there are also variants of that hypothesis out there in the literature that just more generally this is about saving energy right so for example if you live in a world where you're very powerless so this is not just to do with infectious disease but more generally there may be recent times when you need to save energy right it's it's a the weather conditions are poor for hunting what can I do what I can't go hunting but I could save energy the conditions for acting are unpropitious in whatever way I'm powerless in a situation or my somatic status you know not not up to the job or simply I live in an auto correlated world and the last time I did something you know that went badly so rather than repeat it I wait until prevailing conditions improved so there's a lot of hypotheses going back decades about the idea that this is a phenotype which is about sort of saving energy and and and so on related to that I'm just so I'm flipping around the circle a bit the one that says managing risk taking adverse circumstances so a related idea that I've written about and others have too is that you is that animals and including human animals are constantly having to make decisions about whether to seek new opportunities and goals or you know stick with what they've got whether to avoid potentially dangerous situations or whether those are exactly the situation's they should be going into and you need kind of mechanisms that manage that they're taking information about what state am I in and what stage is the environment in that will tell me is that you know is this a good time to be doing something risky so by risky there's a in evolutionary biology risk has a particular technical meaning which is to do with the fact you don't know how things will turn out they might turn out very well or they might turn out quite badly and there may be times when it's just not it doesn't make sense circumstance is not an opera from propitious for you to take on new actions and so maybe what depression is doing is saying you know this is not a good time to act you don't have the social support or you don't have the the energy or you know the situation is too oppressive so you know you need to keep your powder dry and that would explain the pessimism the anhedonia that the people aren't willingness to go and go do things related to that of course at the core of depression is emotional pain it's an aversive state and it's certainly been argued in the literature that this is a state that firstly needs to be aversive because it's about learning okay so when your partner leaves you it feels awful and one of the things that teaches you is to be nicer to your partners in future or whatever right order to navigate that particular maybe choose different ones might be also view another another approach but you know things being valence has an important function in learning theory because we avoid that which is aversive by definition so the negative affective valence of some of these symptoms may be to do with their role in learning and therefore hopefully behaving more adaptively in future how related to that there's a specific hypothesis called the social navigation hypothesis of depression that kind of says you know when things go badly wrong you need to ruminate on why that is you need to develop this mindset which kind of endlessly goes back through what you could have done differently what know what what things you might have been responsible for in the fact that situation turned out the way it did and that rumination you know which is obviously we see as a pathological symptom of depression but could be argued to be a kind of useful thing from that point of view and similarly depression is to do you know causes people to disengage from sets of goals and Randy's nasi earlier in his career wrote a lot about the idea that when people are depressed what they do is disengage from a set of goals and there might be times in life people's lives or that's actually a good thing to do so just finally and without prolonging there's too much longer there's a bunch of hypotheses of depression which are about signaling so one of the things and particularly in the case of postpartum depression this has been argued is that depressed people are signaling to their caregivers I need more investment here I need more help here I can't do this look at me I'm in pain um you know I'm crying I'm unable to sort of complete this you know you need to come and help and of course the lack of social support is a major epidemiological risk factor for depression so that's an interesting sort of hypotheses hypothesis so I just wanted to go through all that to give you a flavor of the idea that there's a bunch of ideas about out there about how depression could be adaptive and this is this literature this evolutionary literature I think varies in its intent from from some of it which is quite sort of in a sense quite anti anti medical anti medical model which is sort of saying that stuff that you're calling you know a dysfunction is actually just the normal functioning of adaptations that are doing their job right so that's quite a radical critique of the sort of medical approaches to depression through to people who I think many many of us in this room would would be better fitness description who fully accept that this is a medical disorder in at least some cases but we're just trying to understand what kind of disorder it is and what normal functions it arises from right so where you are on the kind of over medicalization question it seems to me orthogonal to whether your interest in these adaptive ideas you could believe interested in these EV active ideas as a kind of source of where the maladaptive over expression of these defenses comes from okay so we've got a load of ideas about evolutionary functions of depression I think there are a lot of issues with them so the first one is that there are a lot of them right how do they all fit together which one's right we don't know I mean the troubles they all have a kind of prima facie a plausibility about them all yeah yeah saving energy you're right depressed people lose to save energy oh yeah well you know signaling they do signal you know and so so which is it right so in a sense if everything we think of as the evolutionary explanation seems a good idea as a scientist that means we're on the wrong track okay science is about finding the things in the world that are definitely not right to leave that small set that might actually be right so if everything we can come up with seems a pretty good idea then I think you know we've got an epistemological problem the other thing is to my knowledge there is really have us take a strong view into a depression is an adaptation then you are committed to the claim that people who experience bad life events and develop depression have higher fitness or better outcomes than people who experience those same life events and don't have them because the definition of an adaptation is something that improves survival and reproduction relative to not having it and I just don't think we have any such evidence for the capacity for depression right I don't know of any study that says you know what people who get to breathe you know depressed after they're bereaved it's horrible at the time but 20 years later they've got over it much better I just don't know any of any evidence of that kind and on the contrary all of the evidence on depression is people who have bad life events and become depressed have a worse long-term follow-up than people have live bad life events but are resilient to them you know relatively resilient because they'll have you know a transient reaction get you know having an episode that we meet sort of diagnostic criteria for depression is just really a predictor of a you'll have other episodes in the future and be your physical health will be worse in the future and is actually a risk factor for all kinds of things like de predamond sure so this doesn't really fit with the idea of it yeah it's nasty at the time but a bit like fever it's an adaptation and actually the people who haven't got it a really you know badly off so that's a problem right if you're going to take a very strong adaptation history of depression I think the evidence just isn't there to support them how can we save evolutionary ideas about depression well I think there's a few ways we can do it so a distinction has already been made in the earlier discussion is to distinguish between low mood and depression and/or markus call it depression and clinical depression but however you want to make that cut that particular pine and the idea goes here that you have a set of adaptations where the adaptations for low mood and these are normal right these are things that all normally developing human beings have and they're they're adaptive it's you know if your loved one would die and you wouldn't feel sad about it you're you know you're you're not well-equipped to navigate the social shoals of being an effective human and and you know that that would be a bad thing if you have high levels of it of inflammation you don't get sickness behavior yeah then that would be bad so so what people have said well these adaptive explanations may apply to low mood but then sometimes the low mood system goes wrong and what you see in clinical depression is that the low mood system has gone wrong why might it go wrong well there are genetic mutations might make people there are various kinds of neurobiological bowel functions and their developmental malfunction perhaps you know you're simply developing in a an unfavorable environment and then of course let's not forget there's just chance right this is a complex systems and they have to develop biologically in real time in a kind of sequential way and sometimes they're just going to you know end up not working the way that they should so on that view there's this category of stuff low mood that's species-typical universal and adaptive and there's this other category of stuff depression which happens when that system goes wrong now I would find it I mean it's interesting the conversation we had about about hunter-gatherers and you know whether they have depression and so on if this view is right I'd be very surprised I'm very skeptical at the claim that there's no depression among hunter-gatherers doesn't make any sense to you right I mean those those populations have genetic mutations they have chance developmental insults you know every human organ we've ever studied has gone wrong in some proportion of the population so the idea that there you know as long as you're living in hunter-gatherer environment your mood system couldn't go wrong seems to me crazy of course the prevalence might be different it might be much lower and you know that that could be a thing but you know these are recurrent causes there might be something about the modern environment that makes that pathology more likely to occur and that would be interesting but the idea that there wouldn't be any among hunter-gatherers makes no sense to me biologically okay so that's fine so I think that says that sort of symptom on the face of it gets us out of this bind and says okay depression is not an adaptation we all agree that but the capacity for low mood is and the rest of my talk I'm going to sort of get a bit more critical even about that view in in a couple of different ways the first thing is that though I common sense tells us that this distinction between pathological depression and ordinary low mood exists there's actually no empirical evidence that allows us to identify where it falls okay so if you take a standard depressive symptom checklist this is the GHQ the general health questionnaire and you just give it to the British population in a way what you'd like to see is a bimodal distribution right there's this sort of group of people who've got a few symptoms because something Bad's happened in their life and there's this other group of people that got pathological depression and they do about five to ten percent of the population and they would have a different distribution but there's an absolutely continuous distribution okay and there's actually to my knowledge no principal basis for putting the boundary anywhere in particular so much so that as you know over the successive DSM's the boundary has changed so all of the we know we know that all of the individual symptoms of depression suicidal ideation may be a difficult one but grant me that occur transiently they're going quite a lot of the population this thing about saying well if you've got five symptoms you're depressed you know and therefore you've not there's a particular empirical basis to that it's just a kind of rule of thumb and in fact if you want to do things like well you know how could we validate our particular you know our particular diagnostic cut points when we attempt to do that it goes very badly generally so for example if you if you if you did if you identify depression using much looser criteria of just subclinical negative effect that's just a good predictor of long-term outcomes such as physical health risk of developing dementia and other things like that as it is to have met the DSM four or five criteria so there's really nothing about our current criteria that you know allows us to say yeah definitely there's something about the people who go over five or whatever so although common sense tells me that must be right there are people who whose mood systems are working just fine but have had a set of adverse experience and there other people's mood system has gone wrong due to some chemical imbalance or something the phenotypic dose data don't tell you where that is and of course the people with very high symptom scores they're probably recommend represent a mixture of these things some people just had terribly have terrible lives really very bad life events recently and some people yes maybe have a you know vulnerability or something but you there's a real point of rarity between those those things in fact what we know about developing human systems like the moon system is that they're very variable right they're variable across individuals all human systems are the same some people have more active immune systems than others and you'd expect some people that have more active low mood systems than others you can't sort of find obviously from there the people is gone wrong now philosophically what we what we think we're doing with these kind of things and with the things like dsm-5 of course we don't literally believe that yeah your mood system is working well until day 13 after your spouse leaves you but it goes wrong on day 14 though a strict into you know strict interpretation of what the DSM says would say on day 13 you're just a normal person your low moods isn't working find a vault you know it hasn't got better yet you know you developed a pathology now we all accept that that's sort of arbitrary we don't literally believe that I suppose what we believe is that these arbitrary cutoff probabilistically tell us about some unobserved state of the brain either being in pathology or or not but I'm just saying that there's no great neurobiological evidence or any kind of evidence that we've got those those boundaries in in the right place at the moment and more than that excuse me different people who meet the criteria meet them in very different ways so as I said for a dsm-5 diagnosis you have to have five of these symptoms now given that there are about eight or nine symptoms there it follows actually that and especially given that three of the symptoms you can have opposite phenotypes on them and still qualify right so we can have person a who has very blue moods as putting on weight sleeps all the time is slowed down and feel suicidal a person B who's anhedonic losing weight can't sleep sped up and and can't concentrate most of people get the same diagnosis right in fact they have no symptoms at all overlapping so this is a real problem and for the reasons I'm going to come on and say also notice that many things on this list are on lists for other disorders too right so that's a further problem so I've kind of come on to why these things are problematic I think they're problematic clinically but they're also problematic from an evolutionary point of view which I will come back to our promise this is just actually mark has already mentioned these studies so I and and also I feel the led legibility is not very high so I won't go into them in detail this is just some famous studies that show different adverse life events produce different patterns of symptoms actually the data when you really start delving into them a much less impressive than the abstract seems to make it sound like they're going to be because actually they're quite a lot of symptoms that are common to all of the adverse life events but there are some statistical differences so example after romantic loss people routinely lose weight and there are others there after other kinds of adverse events that make people routinely gain weight so there are these subtype differences but here's what I wanted to talk about because it's relevant both clinically and and to the Volusia nari stuff which I'll come back to what are we doing when we say that someone has depression we're we're alluding to what we call in psychology a latent variable that's something you can't directly observe but which you infer on the basis of observable sanctums signs and symptoms so if someone comes to you with anhedonia fatigue increased appetite in and suicidality you infer this underlying thing called their depression okay and that's sort of fine in a way you need classification frameworks and this is the one you use but let's be careful of reifying that because then we say the depressions what we want to treat but the present doesn't exist you haven't observed it what you've actually observed is anhedonia fatigue kappa titan and suicide hours but you've inferred this latent variable which you then ascribed properties to and this is significant because as I've already told you two different patients might come with entirely different set of symptoms in each case you infer the same underlying latent variable because you've got this kind of many possible mappings to the latent variable of different probabilistic signs and symptoms and what you then end up doing is in your sort of the universe of things you might consider for treatment is the universe of things that have been considered relevant for that latent variable depression not the universe of things that's been considered of found to be effective for that particular set of symptoms which might be an had earlier and fatigue for one patient they might be suicidality and insomnia for a different patient so already that's telling you you know by saying oh these are both cases of depression I know how to treat depression mask the fact that in fact you've got two different problems in front of you we do this latent variable thing all the time in psychology right just think about something like extraversion right we see people doing various behaviors we infer this thing called extraversion oh my god I've got this thing are this person's an extrovert so now I know I understand why they do the things they do but of course the extraversion doesn't explain the behaviors because you didn't know about the extraversion apart from from the behavior so the entirely circular to claim that they explain the behaviors similarly why is this person an had only can fatigue because they've got depression how do I know they've got depression well because they're anhedonic and fatigue there's no explanatory purchase in that latent variable useful heuristically though it might well be there's a new there's an approach that's in clinical psychology is very much you know sort of the current thing in the field called the sim tomek's approach or the network based approach now I don't know to what it's and that's percolated into psychology as are saying clinic to psychiatry and clinical psychologists the thing du jour and I'll talk a bit about that and there's a big debate about whether this is really just the emperor's new clothes or there's actually anything you know substantively different from what we all did anyway in this symptom X or the network approach the basic idea of the sim terminus approach is you abandon the latent variable of depression altogether except possibly as a kind of you know useful kind of catch-all but you don't assume that there is a real thing inside the person called their depression what you say is what symptoms do they have so someone has a fatigue you're not as early as someone else might have had some different symptoms and what you understand is those are the phenomena you seek to explain and those are the phenomena you seek to treat without arguing that they're signed in signs and symptoms of some underlying thing you haven't yet observed and crucially in the sim tomek's approach you allow the idea that symptoms have causal consequences for each other so for example you might say well the aunt had only is actually caused by the fatigue and the suicidality is caused by the anhedonia right or whatever in which case it would follow the prediction would follow the if you treated the fatigue the Antonia and the suicidality would go away but what you're dealing with here is a case of fatigue is a case where the clinical imperative is to treat the fatigue not to you know infer a latent variable and then apply the treatments previously have been found effective for that particular latent variable it's an important and subtle change because what it says actually is I mean it goes further down Marcus's road of saying there's one thing there's 12 subtypes it says we don't have to have any discrete and countable number of sub times what we have is people with problems and we need to find out what problems they've got and in what combinations right and then we need you know we'd to say well what's the evidence based on treatments they've been effective for people with those problems in those combinations okay and I just I'm slightly conscious of time but there's plenty of evidence of example the individual esm symptoms have individual individually different risk factors again the evidence the data is not as nice as it sounds like they ought to be when you start looking in detail what does this intermix approach to give you well what people do is that they take data on actual symptoms without regard for diagnostic categories and they make these networks of them and so what a network like this shows this is from the National comorbidities survey is this is the symptoms of generalized anxiety disorder and major depressive disorder in many thousands of patients and the size of the node is the frequency of that symptom or the prevalence of that symptom and the length the strength of the edge you know that how bold it is is how commonly those symptoms Co vary and you can't probably see but to the left hand end is the symptoms of major depression the right hand ends are the symptoms of generalized anxiety disorder so these were really discrete disorders what you'd hope was that there'd be a lot of bold edges between among the symptoms on the left side and among the symptoms of the right side and rather few edges right it's going between the two clusters because the only people in whom that symptoms on the two sides would covary will be people who had both disorders actually what you find this isn't true at all that the symptoms of generalized anxiety disorder are often found with people to get the diagnosis of major depression and vice-versa so much so that we've known for ages that there's comorbidity between these two things and actually the inter-rater reliability on whether in the dsm-5 field trials the integrative reliability on whether someone had generalized anxiety disorder or depression was very poor it was it was such that it'll be considered unacceptable in a psychology publication so these things all go together right but you know they go together more often than the idea that there are discrete disorders suggest but on the other hand they don't go together often enough to say there's just one disorder but because people don't have them all actually or then have them all to the same extent so we need this intermediate position that says actually every individual has a different network of symptoms and in different individuals there might be different causal pathways from one symptom to another so I think I possibly need to just slightly fast forward to make sure I can get in everything I wanted to say so what the kind of network I showed you on the previous slide these are cross-sectional from many patients but in a way the idea of the symptom it's approaching the long term the sort of ideal of it is that you would develop causal networks of symptoms and end up in an individual place patient in which the edge is represented the strength of the causal influence in that patient from one symptom to another how you do that's difficult but we can come back to that so these might be different like for the patient on the left on the right and the way you would want to treat things depends on your understanding of what that what the Keystone symbols these are symptoms in that community are and the ones that have kind of causal knock on the other things let's come back to our evolutionary perspective a bit I think this sort of shifts the emphasis rather from saying okay we need to explain this thing called depression or saying like oh we don't explain that we need to explain the thing called low mood what are the actual entities we want explanations for well on this approach they're just the symptoms themselves why would humans have a capacity for fatigue that's not very difficult why would humans have a capacity for anhedonia well it's not very difficult to think of cases why that might be true why would humans have a capacity for suicidality that is a bit more challenging like that admit why would people sometimes not be able to sleep well it's very easy to think of kind of adaptive situations about that we know that animals don't sleep so well for example when they're in high predation environments why do people sometimes want to sleep more well you sleep well when you recover from an illness so the individual symptoms are not very hard to explain I think the other thing we need to mount an explanation for from an evolutionary perspective is the linkages between symptoms right because although I've said every one of these networks will be different for different patients they will be statistically at the population level more common linkages than others so for example sleep problems and fatigue relief no surprise you know but they're really well linked but if you take that seriously and say okay well you know if in fact I'm going to be sort of literal about the about the symptoms are being presented with then treat sleep problems okay similarly peoples and I think this sort of rather changes your view from saying there's this kind of thing for their depression which is hidden in there somewhere if I can only find it to say this is someone who's got a whole list of problems but at the head of that list is that they can't sleep and actually treat their insomnia and you make a strong prediction that a lot of downstream things would kind of what kind of change so understanding from an evolutionary point of view which are the important causal linkages and I think there's one from inflammation to sickness behavior there's one from sleep problems to fatigue I mean understanding what these individual links are it's enormously helpful in terms of thinking about therapeutic strategies and it's more helpful in my view than trying to think of evolutionary explanations for diagnostic categories that we ourselves as researchers are kind of sticking on to what order in effect networks of phenomenon how much time have I got can you show me you about 10 minutes perfect okay I'll go relatively fast through this last bit let me just briefly come back to so on this view where there are just symptoms and edges you know do we need this notion that there's normal low mood and and maladaptive depression I think yes probably we kind of still do end up with some sort of notion of that distinction perhaps it's rather different I think the view that's being pushed by this symptom expedience are complex dynamical systems okay so there might be for example feedback loops sleep causes fatigue and you know sleep problems causes fatigue and fatigue causes and had onea but anhedonia causes sleep problems right so you can get these kind of perfect storms so what you get is a dynamical system where the causality is actually flowing around and unfortunately coming back and eating its own tail sometimes and what we know about complex dynamical systems is even relic ones with relatively small number of nodes have multiple stable states at least if certain things are sure about them master says they might normally be homeostatic that you know you push on one symptom the whole system kind of creeps over there because of the causal linkages among symptoms but because you know of negative feedback loops it does eventually sort of push yourself back and yeah so most people most people's depression is self-limiting actually for on the individual episode basis but similarly in it when something bad happens in your life you have multiple kind of affective reactions to that but it is you know just tend to be self-limiting but something we know about complex dynamical systems is they do get sometimes pushed into different states right the state where you could imagine if you had particularly strong jolt to one symptom that could set up a positive feed-forward loop that would keep that system for a long time in a state where it you know where it's not designed to be so I think you end up with this notion of pathological depression but as a kind of a case where a complex dynamical system has been perturbed for whatever reason in such a way that it's gone into an alternative stable state and then you need to understand what are the sin they're gonna sort of help me get back to homeostasis I'm just in interest of time I'm not gonna sort of dwell on this but I've written about mood and actually the idea that the various kind of perturbations of the mood system we talked about could be these alternative as it were these alternative states of a complex dynamical system I think it's not just depression and generalized anxiety disorder which are respectively the you know the state where reward systems and pleasure systems are turned down and the relate the the state where and systems for detecting threats are turned up but also we need to talk about things like dysphoric mania which is where people actually the reward systems are absolutely charging away and people are looking for stuff to do but the same time they feel awful bit because their punishment threat detection systems are also very active as it and also euphoric mania where people people that charging around seeking rewards we can think of these as these sort of alternative temporarily stable states of this complex dynamical system in view of time I won't be able to talk much about this but there's some really interesting work that's using within patient longitudinal data to try understand these systems and it turns out that in for example in remitted depressive patients you can begin to predict so what they do in this study says they give people a bleeper and they give they pull them you know frequently at random intervals to say how you're feeling right now are you feeling tired right now are you feeling happier right now are you feeling worried right now and so they treat each of these worried nurse and sadness and tiredness and sleepiness and so on I treat each of these as a kind of node and build the network of how those things predict each other over time and you end up with a lot of data for every individual person that you can actually predict when patients are going to remit because the oscillations and particularly the couplings in these networks start to change so whereas you know in sort of a healthy state we might all have these fluctuations but they they return relatively quickly and also a fluctuation on one of the axes say sadness doesn't have big implications from a fluctuation on the other one as you're going towards an episode those linkages start to be start to be stronger so the prediction you know from one symptom or one effective rating to another start to be stronger and the autocorrelation starts to be bigger so that when it gets perturbed in one direction it starts to be slower to come back and then eventually you know I suppose X I bother say in the in the disorder in the episode itself it's going there and staying there I have a look at that paper it's very interesting stuff about really what's going on the individual at the individual level so I've got one more bit after this sort of interim summary but I'll do it quickly so maybe if this view is right the entities in need of eeveelution explanation is not depression or perhaps not even low mood but the individual components fatigue vigilance etc I don't think those are very difficult to explain but also the causal linkages between symptoms and also between things we don't consider some symptoms of depression and depression so for example inflammation chronic stress all these things which we don't consider symptoms of depression I think we need to have a completely open mind about whether the you know that there's the sort of causes need to be sort on the psychological level you know at all I think you know the mind and the body have no natural boundary between them and I think infectious disease as a cause of depression I think that autonomic nervous system changes I think changes in physical activity changes in diet we have to consider all of these things as possibilities of things that perturb that system so we need to understand that causal linkages between symptoms and also between non symptoms ie causes and symptoms quite a lot of these symptoms have animal counterparts and it individually well studied when people say I've got a mouse model of depression they almost never have what they've got is a mouse model of one symptom of depression and so actually you know we already know quite a lot about them I do believe that must much much of what we label as depression and anxiety represents the normal functioning of this network in the face of environments that let's face it are often pretty adverse and ultimately that is you know homeostatic thing people who on average experience worse environments will experience more depressive symptoms but it doesn't mean they'll have to press depression episodes as such but I do believe it is a homeostatic network in most cases but disorder whatever we take that to mean does arise for all kinds of reasons possibly including that the contemporary environment does not contain certain regularly occurring features of this kind of environments we've more typically lived in the insured a return a homeostatic return to this often most frequent stable state physical activity I think is a great one right so no one was really built for a world where it will be possible to not be physically active for long long periods of time and it may be that having to do that I mean to get up and do that by because you needed to was a way of of kind of interrupting one stable state of that network and returning it okay final point for me what's the benefit of the evolutionary perspective on all this I just would like to sort of state though I can't elaborate I don't think evolution helps you very necessarily very much for thinking about medicine but I think it helps you a lot with thinking about public health okay and most of the people right if you wanted to reduce the prevalence or the incidence of depression you have to understand that most of the people who are going to become depressed next year and they come from the population that's healthy this year so treating the people who are already depressed is never going to do it for you right you've got to stop them turning up and and the work and so a lot of the things that affect the states of this system are already well-known poverty social isolation physical inactivity pollution Marcus mentioned food insecurity poor working conditions and lack of job control shift work things that mess with your sleep right so these are not narrowly medical issues these are public health issues and they and they spin out into public health policy and when I think one of the things for me that the evolutionary approach buys you it's not so much it says you know which bit of the brain you need to look at is it says well what what's a decent environment for a human being to live in so if you really believed about reducing suffering and ill health you know let's spend a lot of our effort on trying to make those environments and then hopefully we'll have to spend a bit less on you know on sort of curing the problems there's shame of that I know that's controversial and maybe a topic for another day but I think my cue has come to stop thank you about the public health rather than the simplistic treatment of one individual so III don't think that's actually that controversial I think the college itself has looked at a lot of things George might help it thank you very much very interesting just a couple of questions one simple why do you say suicidal thoughts are not normal and secondly would you say that your approach is quite typical of psychology in the sense that we were talking earlier today during dr. reigns talked about different levels of meaning and sociability in attending group analysis and one of the ideas there is taking the non-problem seriously so thank you for that I when I said that suicidal thoughts are not normal actually don't know statistically what their distribution is in the population in this sort of so-called healthy population how frequent are they and so on I simply don't know much about that I suppose I suppose I merely meant that from an evolutionary from a sort of lay you know unconsidered evolutionary point of view things like fatigue are very easy to understand but things like suicidality you're a little more puzzling on the face of it so that's all that's all I meant I'm not saying you couldn't show that in fact they were more broadly distributed and and and also had you know some sort of reliable generator that could be understood but I should stress this wasn't particularly my approach I just put put it together through this talk but the network approach is very much in the psychological tradition of Demento of viewing difference dimensionally rather than categorically but I think goes beyond that with this idea of networks of symptoms psychologists love to classify psychiatrists as being you know very discreet kind of everything's in a box way which in practice always turns out to be unfair to the psychiatrist but so yes this is more of a psychological approach most psychiatrists I talk to say look I have no problem with any of this but we need some labels we need to communicate with each other and that's why we have these kind of so it's really about not abolishing the labels but just remembering that the labels are conveniences thank you sorry Thank You Daniel I take your point that you you warned us not to be drawn into being over adaptiveness yes but what would you say about the cognitive symptoms of depression memory loss and lack of concentration hmm I mean it's really hard to argue a resource saving feature of those you know symptoms yeah they're quite common and actually even when we treats all of the other symptoms we can't really treat those problems and they cause more and more problems so it's clearly not adaptive so what would you say in terms of evolutionary well what I would say is that there are constraints right organisms have finite finite energetic resources so I think that often what's happening in depression is that that simply the burden of other stuff that's going on the tracks resources from from the brain's ability to to do so I'm not there's no there's no positive adaptive function of those things but so for example we know amongst ultra marathon runners that their cognitive abilities at the ends of these long runs are really really down and there's no there's no positive adaptive reason it's just that they're so the feet has so sapped them that there just isn't any energy left to allocate so I guess I would go down that kind of route of a sort of by-product thing now the interesting thing about that might be you'd sort of deal with the other symptoms and it's a while for those things to come back I do believe that they can come back though that's I guess that's an empirical so I think some of the symptoms rather than saying oh you know that there's an adaptive function for this it's just like you can't manage to do certain things whilst other stuff is going on I guess thank you sir self gentleman behind you hello my name's charlie shot at them and working age adults psychiatrist thanks very much interesting talk I just wanted to make one comment because you were talking about symptoms and symptom existe where exactly which study was but it was based on the star D interventions and they did look at symptoms and then they used a machine learning algorithm basically to subdivide the patients who obviously had treatment resistant depression and then very much data-driven it wasn't Theory driven but he then was able to say actually this particular symptom cluster this particular intervention might work so I think there have been some yeah borås into that and there was even moves to then even primary care having a sort of working algorithm that physicians could actually put start putting symptoms in without having much theoretical background or you know sort of competency without comprehension kind of yeah very much thank you so if I if I if I seem to come across as saying everyone has these very simplistic you know discreet diagnostic boxes I apologize because I you know I I'm just saying I think that's the route we need to go down actually yeah yeah yeah yeah no go got it you're on my list Marcus thank you very much Daniel for a fascinating talk I just wanted what wanted your comment about reproductive success and depression yes it seemed to me from from a large study by power at all I think 2013 is that JAMA or something somewhere like that I think in China where depression stood out as the only major psychiatric disorder that did not involve decline a reduced reproductive success men had 90 percent of of reproductive success of the general population and women a hundred percent and that's very interesting whereas all the other conditions I mean schizophrenia was the worse yeah we're we're you know which was like men had 20 percent or something where women 40 percent and all the others were slightly higher but depression almost no reduction in reproductive success comments that's very interesting I don't I didn't know that particular fact I knew in schizophrenia that they've been these kinds of studies and it's all it's always been very low I didn't know that fact about depression so that's very very interesting I suppose that typically the the course of depression is much more remitting than the schizophrenia and much more sadiq until you you can see how how that could come about but I I mean what what that I don't onion but let's not make too much of it what it possibly says is that people who end up having that diagnostic label may be pretty normal you know they may you know just just have a run of bad events as it were therefore we know that its course as much you know much different from that of a psychotic disorder but that's very interesting thank you for that that's like to add to that that we've spoken about this I'm sorry yes over here about it being a sort of normal experience maybe in extremis but that doesn't mean we shouldn't try and alleviate the suffering no that's because people people get quite cross with evolutionary psychiatrists are saying it's normal therefore you don't have to do anything that's not what we're saying no no I thank you for that and that you reminded me I didn't sort of say that at any point but I mean this how always some symptom occurs near to me and it's not something I've personally written about it's just what I've been reading about recently my understanding of the center mcstuffins is not deconstructive of the idea that people have real medical problems it merely says you treat the symptom profile that turns up and so if people are presenting you know with a more anxious agitated presentation you know you need to you need to listen to that and I mean you could go further with it and then say well why would that be what's been happening to you that this is how you find you know but but no absolutely not like the you know the alleviation of suffering still remains the goal here I think if you want to understand there are any fear of animals adaptive study in the environment where that adaptation was evolved so if we study modern humans we live so different environment that we are not able to understand that at any rate was adopt even it just makes it very difficult to understand and if you want to do explain the person should be able to oppress only humans but in all animals so we would need a general theory if it fits empirical evidence not only in humans but in also other animals I tried to do it it's very difficult but let's hope that some day people find but only looking symptoms won't help because because the data collected among modern humans don't tell much about what happens billions of years ago and we those adaptation with emergency absolutely of course you're right and I suppose if you if you hope we might develop a cause of understanding of how these things are linked to each other that presumably reflects the evolved system I appreciate that doesn't tell you how it associated with reproductive success in the past but I I mean I believe that you can develop a causal understanding of how different psychological phenomena relate in modern humans just because they're living in a weird environment and stop your doing that necessarily when you can't understand is what the natural history would have been and that's where I'm together studies a valuable but few difficult to do and this on yes it is much of what's been said about is about the individuals fitness resilience and susceptibility and how that relates to the environment but I wondered if there was anything to be said about the group Fitness susceptibility and resilience and what might be advantageous for the individual might not be for the group and vice versa well thank you for that I mean that's you sort of let off a rocket there you may not have realized but the whole idea that groups are things are fitness is very controversial in evolutionary biology and I think the general consensus is they are media under certain very restrictive conditions mainly whether people are but there's very little migration between groups and and therefore what and what ends up is the individuals within the within groups tend to they're not very related to each other and not closely related to individuals and other groups the general thought is that groups throughout human evolution have been very fluid and with lots of fishing and fusion and that doesn't lend itself to a situation where the notion of group fitness really has an interpretable meaning so I mean I think to talk about the fitness of a of a colony of bees or wasps is perfectly fine and it makes a lot of sense right because you know they're there lot is bound up together but for a highly mobile fission fusion species like humans the idea that there is a thing like group fitness is very controversial amongst eeveelution apologists I have my own prejudices on the subject and markers may have different views than me but I'll leave that there ultimately reproductive success is you know it's individuals that reproduce you know a society doesn't have a baby an individual does so so the notion of Fitness at the group level it's unclear what that means it would be clear in a bee or a wasp because there are queens and they're you know that's a different matter but there are those sorts of reproduction and other things that influence reproductive success and ideas can also reproduce it absolutely if you look about you know cultural evolution then that's a different matter also what I said doesn't in any way imply that societies aren't important to our well-being right that the property of the society we live in is fantastically important to the well-being of every one of us is just that the technical notion of fitness which is a notion of the evolutionary biology doesn't have it it doesn't have an interpretation when groups are of the nature of human groups lastly for copy yeah thanks for that I really really love that all I was and I think you've said something you weren't you're just really well articulated something as you say most of us here know that this is a category these are words and we're sort of using them as shortcuts to describe phenomena and that can be very useful in I've really liked the idea of making that boundary when someone's locked in sort of self-perpetuating dynamic and network of systems so I see how it's very useful for treatment one problem I'm having all the time especially with my students at the moment who come and ask me things about well what does this mean if you know there's sort of evolutionary interpretations that you went over the moral clock it was a philosophical talk right hmm so in medicine that boundary we can decide where that boundary should be but there's sort of freewill connotations and philosophical implications of this a huge right and now whilst it's great that were increasing awareness about mental health problems people are being like oh there's something wrong with my brain hmm and therefore this is excusable when really you know the distinction between someone who's in jail and and due to depression or someone who's in jail because they're a bit Moody it's an arbitrary one how do we balance yeah when when students are asking me or when people are trying to communicate this how do we balance the benefits of the medical benefits of categorize and getting people to seek treatment and then this misinterpretation that you've got a pathology which is due to neuroanatomy but all other personality characteristics are you know completely independent and you know I mean it's a can of worms but I'd love to well I think if you say these are kind of worms I think it's very interesting and I think the difference between you know the other species Marcus is talking about and humans is that humans have an awareness of the availability of a diagnostic category and I think that in itself changes the phenomenon phenomena I mean for someone who's as biologically oriented as me this is a dangerous concession to make but it's true that people when you make available and an interpretive framework for the kinds of adverse experiences people are having people have different experiences that then fit that fit they are available frameworks of meaning right and I think we know that humans do that so it's a problem and I've really no idea how we solve that I'm very worried how many people you know some extraordinary proportion of us children have a disorder you know are deems to have a disorder of some kind or other I mean an order of magnitude greater than was true two generations ago and I think that's because there are a lot of adverse experiences so some of those that recurrent some others do with particular social and other arrangements now and people are finding ways to you know make sense of those and the medical very what's the one we've got so it's a matter for real concern something I've got no good answer but something I will say is something I like about the evolutionary approach and what I always sort of tell my students is evolution is about variation and they don't get this I think eeveelution is about this is how you should be and I say you know the whole if you'd read the first hundred pages of Darwin's you know on the Origin of Species it's about how pigeons are really different from each other and none of them is not a proper pigeon they're all like normal pigeons but they're just really varied and and I kind of keep saying what evolution is going to produce the sort of spectrum of variation and so for me that that can be a valuable thing to say look you know these things are kind of so vary a bit from individual to individual because it's complex and and also because that's what evolution does it's about population thinking understanding this continual variation that fails to answer your question well I'm gonna call there cuz we've got a coffee break now for 15 minutes and thank you very much Dahlia thank you everybody [Applause]

Channel: EPSIG UK

Views: 4,292

Keywords:

Id: krTh0mJkyeU

Length: 61min 1sec (3661 seconds)

Published: Wed Apr 10 2019