E108 - What is Psychosis? (with Prof. Richard Bentall)

Richard Benthall is a Professor of Clinical Psychology at the University of Sheffield in the UK. His research has focusued on the psychological mechanisms of severe mental illness (such as hallucinations and delusions) and social factors that affect these conditions. His work along with the work of other researchers is responsible for psychological interventions (such as cognitive behaviour therapy) being included in national guidelines for the treatment of psychotic type conditions.

Bentall along with some collaborators also published a meta analysis in 2012 suggesting strongly that trauma is an important risk factor for developing psychotic conditions:
https://pmc.ncbi.nlm.nih.gov/articles/PMC3406538/

Interviewed by Dr. Alex Curmi. Dr. Alex is a consultant psychiatrist and a UKCP registered psychotherapist in-training.

If you would like to invite Alex to speak at your organisation please email thinkingmindpodcast@gmail.com with "Speaking Enquiry" in the subject line.

If you would like to enquire about an online psychotherapy appointment with Dr. Alex, you can email - alexcurmitherapy@gmail.com.

Give feedback here - thinkingmindpodcast@gmail.com - 
Follow us here: Twitter @thinkingmindpod Instagram @thinkingmindpodcast

Transcript

e108-what-is-psychosis-with-prof-richard-bentall

[00:00:00] Welcome back. Trying to understand the origin and nature of severe mental health conditions like schizophrenia and bipolar is a problem psychiatry and neuroscience is still very much struggling with and a problem I definitely faced in my clinical experience. My guest today is a prolific researcher and the author of several books on this topic such as Doctoring the Mind, Why Psychiatric Treatments Fail and Madness Explained.

Although some people have labeled him as aligned with the so called anti psychiatry movement, My experience was that I was speaking to someone who was and is heavily involved and invested in the research in this area, not dismissive of the potential biological processes at play, and someone who was not anti diagnosis as such, but who wanted our diagnoses to be better, more sophisticated, and more reflective of an underlying biopsychosocial reality.

Today I'm in conversation with Professor Richard Bentall, a professor of clinical psychology at the University of [00:01:00] Sheffield. His research has focused on the psychological mechanisms of severe mental illness like hallucinations and delusions and social factors that affect the development of these conditions.

His work, along with the work of other researchers, is responsible for psychological interventions such as cognitive behavior therapy being included in the national guidance for the treatment of psychotic type conditions. Bentall, along with some collaborators, also published a meta analysis in 2012.

Suggesting strongly that psychological trauma confers a highly consistent risk towards developing psychosis as an adult. Today we discuss what are the symptoms in detail which characterize psychotic episodes and manic episodes and the way they overlap. Some of the problems with our diagnoses like schizophrenia, bipolar disorder, or schizoaffective disorder.

The role of psychiatric genetics. and statistics and beginning to [00:02:00] help us understand more deeply the origin and nature of these conditions, the role of the environment and trauma and how these conditions develop, psychological mechanisms underlying things like hallucinations and delusions, the five kinds of delusions that present themselves most commonly all around the world, how people with severe mental illness can be approached psychologically, and how the design of mental health systems could be influenced by different key examples from around the world.

This is the Thinking Minds podcast, a podcast all about psychiatry, psychology, psychotherapy and self development. As always, thank you very much for listening. If you'd like to support the podcast, check out the links in the description. And now here's today's conversation with Professor Richard Bentall.

Professor Richard Bentall, thank you so much for joining me. [00:03:00] Oh, you're welcome. Um, it's nice to be here. We're going to discuss a few interesting topics today. Actually, earlier today I was, I had a conversation with Dr. Mark Horowitz criticizing the biomedical model of depression and perhaps some of the limitations and dangers of antidepressants.

Today we're going to be talking more about conditions like psychosis. and bipolar, and I'm curious to see what differences may arise between talking about psychosis and bipolar and depression, uh, again, limitations of the biomedical model, and perhaps some questions about how we should be rethinking our public mental health system.

It'd be great to know a little bit about your story, how you got involved with studying the things that you did and where that journey took you. I guess the first thing to say is that I'm at the, probably the opposite end of my career than you're at. So I've been around for a long time. That said, I don't plan to stop anytime soon because I really [00:04:00] enjoy what I do.

I qualified as a clinical psychologist in 1983, having previously done a PhD in experimental psychology at Bangor University. And my PhD didn't have anything to do with clinical psychology directly. It was actually in, as I say, experimental psychology applied to child development, actually. Although I'd originally hoped that it would have a focus on mental illness.

So I'd actually been spent about six months visiting daily, almost, um, a big psychiatric hospital of the old kind, Denby Hospital in North Wales. And it was just winding down. I think it had about a thousand patients when I was. There, but it was rapidly emptying, but I don't know why, but I was fascinated by schizophrenia.

I'd read Lang actually as a student, which helped. And I did a little research project as part of my clinical training, which was on, on hallucinations actually. [00:05:00] After I qualified, I worked for two years in NHS forensic services where most of the patients were psychotic. And I had no idea what I was doing.

I think it was psychiatrists who were very much in control at the time. They thought I was nice, but useless. There's this nice young man called Richard Benton. He wants to work with people with psychosis. We don't think he can do any harm. Let's let him do it. But at the time, it was just at the time where people were beginning to think about maybe whether psychological therapies could be applied to psychosis.

And for some reason, it was something about the culture, but other people and Um, of threat in the country were also interested, particularly down in, in your neck of the woods in, in slam Kings, what's King's college now. But so people like Philip Garrity, for example, Liz Kuypers and David Fowler, and we formed a kind of loose network of, and we just found our way, I think.

And eventually that led to clinical trials of CBT for psychosis. [00:06:00] And eventually that led to. CBT being recognized by NICE as a treatment for people with a diagnosis of schizophrenia. I was always very fascinated with the phenomenon themselves. Why is it that people hear voices? What's going on when people hear voices?

I had a particular idea about voices, which I still think It's probably mostly true, which is that they are a form of inner speech. When I was doing my PhD studying children, I was interested in inner speech, this whole process by which we learn to talk to ourselves in our heads, which is a normal part of child development.

So when I was doing my training, an actual kind of thought was surely hallucinations must be something going wrong with inner speech. So that's how I started on that. But over the years, I've been involved in a lot of different aspects of psychosis. I've been involved in, mainly I focused on studies to try and understand the psychological mechanisms, but they broadened out to include, for example, neuroimaging a little bit, although I don't really know how to turn on [00:07:00] the MRI scanner, so somebody else has to do that bit.

I've been involved in psychiatric epidemiology. I'm a bit of a kind of, I like all types of methodologies actually. I've done qualitative research as well and clinical trials, as I mentioned before. And one of the things which has emerged out of that is the discovery that trauma in the history of patients with, it's like, it appears to be quite important.

And I would say a causal factor, not the only causal factor, but a causal factor in psychosis. And that was an idea which even about 15 years ago was heavily resisted by a lot of mental health professionals, but I think it's now pretty much widely accepted. Yeah, I think I started training very much in the era where it was just about being accepted into the mainstream, more or less.

But taking a step back just for a moment, what do you think is a useful working definition of psychosis? Oh gosh, that's a really good question. So by psychosis, when [00:08:00] I, clinical psychologists these days and quite a lot of psychiatrists prefer the term psychosis. Precisely because it doesn't give, I think, a sort of air of a kind of diagnostic precision, which isn't really merited in this area.

So psychotic disorders, I think of as the disorders in which there is in some sense the individual appears to lose touch with reality. Typically, the diagnoses which people receive. in this area are Schizophrenia and Bipolar Disorder or Schizoaffective Disorder sometimes, Delusional Disorder sometimes.

As I think you probably know, I've always been skeptical about the scientific and actually clinical value, utility of those kind of diagnoses. They don't really cleave nature of its joints. And I've been involved in quite a lot of research which actually shows that. What do you mean by that? Ideally, we'd want to classify, so actually just taking a step back, I should first of all say I'm not against classification.

Classification is an essential [00:09:00] kind of part of science. Only science makes progress only when it's got a kind of working classification system. So the name classification of the species led to ultimately to Darwin's theory of evolution. Mendeleev's classification of the periodic table of chemicals in the periodic table led to an understanding of the atomic structure of elements.

So we do have to have a classification system, and ideally we want classification to, I can't remember what his phrase it was that cleaved nature of these joints, but basically we want to actually have it corresponding to reality. And ideally, That, uh, if it's a categorical classification, it should each diagnostic label or species description or whatever should actually identify a naturally occurring kind of entity or individual.

And for a long time, psychiatry and to some extent clinical psychology labored under the idea that. The kind of [00:10:00] diagnosis we find in the DSM system do cleave nature of this joint. So let's get schizophrenia and bipolar disorder in completely separate conditions, for example. There's a mountain of evidence which suggests that's not true, of all sorts of different sorts.

We mentioned just a minute ago the trauma being a causal factor. heavily implicated in both bipolar disorder and actually major depression and schizophrenia. It's actually a sort of basically linked to just about every poor mental health outcome. If we look at genes, nobody's found specific genes for schizophrenia or bipolar disorder.

There are genetic effects, but they don't, they're not diagnostically specific ones. But it's worth maybe pointing out that, uh, this, there seems to be a lot of heritability in those conditions in that. For example, taking psychosis as an example, if one person, one, uh, with twins separated at birth One twin develops psychosis at some point in their [00:11:00] lives.

The other twin is, uh, my understanding has a 50 percent likelihood of developing psychosis. That's correct. Okay, I'm going to challenge that. Okay. Okay, I challenge you on two points. The first one is just a numbers one. There are two ways of calculating concordance rates in those kind of studies. which is the extent to which you get a similarity between the two twits, you know.

One's called the program wise method and one's called the pair wise method. Without going into the maths, which would be a tedious discussion. Geneticists tend to prefer the program wise method, which gives higher numbers. I prefer the pair wise, and it's not because I'm against genetics, just because it makes more sense to me.

So the pair wise concordance rate is 25%. So roughly speaking, somewhere in that region, 20 to 25%. So, if you have schizophrenia and you have an identical twin, then that identical twin has a 25 percent chance of [00:12:00] Which is still significant given that the baseline rate in the population is 1%. It's much higher than in the general population.

So there is a genetic effect, I don't think there's any doubt about it. Okay. But also I think there's been a huge misunderstanding about the nature of that effect. So I've become very fascinated with psychiatric genetics in recent years, and I've been trying to do, get funding to do some research of my own, but it's been quite difficult.

But anyway, people think of heritability as, as basically a causation index. And you often see that actually written in textbooks, for example. So it said there was a meta analysis which said that by Sullivan, which is actually quite old, but the most recent, which claims that the heritability of schizophrenia is 83%.

If you think that means that 83 percent of the cause. is genetic, then you think, wow, this is something which is almost entirely genetic, but it doesn't mean that. [00:13:00] And the reason for that is because technically a heritability coefficient is actually a correlation coefficient. And it's just a correlation.

It's a partial correlation coefficient to be precise. So what it's talking about is the correlation between the whole genome, basically, and the person's risk of disease. Anybody who's done a stats course knows what they're taught on day one of, about correlations. They don't necessarily apply causality.

So the The pathway, the causal mechanisms, which go from genes to psychosis, are potentially very complicated. And the weird thing is geneticists just don't seem to understand this. Or most don't. Believe me, I've had lots of discussions with them about it. Clearly once, giving a talk about OS to a genetics conference, which was quite memorable.

I started my talk when I talked to them by saying, at the conference, where I said, I think I've probably been invited here to annoy you. And at the end, somebody stood up and [00:14:00] said, you have succeeded. So what, what, what did the geneticists think? The essential point, let me just give you a concrete example.

Okay. When you get down to the DNA level, there are basically two types of genetic variations, which appear to be linked to severe mental illness. One type is called common variant, and these are very, these are hundreds of tiny genetic variations which produce each a very tiny increase of risk. So putting those two aside for one minute, and I could talk about those if you like, but what a lot of geneticists got more excited about is what are called deletions.

And this is when. that due to various accidental processes, that whole chunks of DNA get eliminated or duplicated. Actually, it's, they're called copy number variations because sometimes you get more of one copy of a whole chunk of DNA. And there is something called DiGiorgio syndrome, which is linked to, I think it's a [00:15:00] big deletion on the 11th chromosome, if memory serves me well, and this has a very high risk of psychosis.

So somebody with. DiGeorgiou syndrome. Initially, initial studies seem to indicate the increased risk might be as much as 8 or 12 times. It's not as high as that in recent studies, but it's still quite substantial. Yeah. Bigger than any other genetic variation which has ever been found. So why is this? The temptations to go with something about chromosome 11, which actually causes psychosis.

But if you look at people with DiGiorgio syndrome, what you find is that first of all, that many, almost all of them have learning disabilities. They've got facial disfigurements. They can't speak properly usually because of misalignment of a jaw. They also are extremely vulnerable to, to other physical ailments, particularly heart disease and so on.

So imagine what it's like growing up like this. You're somebody with a learning disability, you look odd, you can't speak [00:16:00] properly, what kind of world do you grow up in? We know about people with learning disabilities is that something which very often happens to them get is that they get very badly bullied.

And we also know from other studies that bullying appears to be one. factor which influences psychosis. What's going on here? It could be that the genes are playing a causal role, but it could be that the genes lead somebody to experience a particular type of environment being bullied, which in itself then leads on to the illness.

Now we don't actually know Which of those two possibilities is true? And I should say there are even more complicated possibilities than either of those two. And the reason for that is because genetic researchers just haven't bothered to measure the environment. Now, you could actually criticize people like me, psychological researchers, you say, Oh, you guys, you haven't measured genes either.

And all I would say about that is, We should, I want to, I've been [00:17:00] trying to get money to do it for a very long time and nobody will give it me. The essential point is you can only understand how these things work if you measure both. You can't just look at heritability estimates and just say environment is what's left.

It's not like that. Because the way that genes and environment interact are very, very complicated. Yeah. And I'll give you another, another example of thought experiment this time about the heritability estimate, which will explain that maybe might indicate why the correlational issue is very important.

You can actually have heritability of a hundred percent. of something which is largely determined by the environment. So, if you imagine a world where everybody smokes exactly 20 cigarettes a day. Nobody smokes 19, 18, 17, or 22, 23. They all smoke exactly 20. What would be the higher probability of lung cancer in that environment?

Would be pretty much close to 100 percent, because the only difference between those who smoke, who get lung cancer, and those [00:18:00] who don't, would be genes. Because everybody gets the same environment. But wouldn't there be other, in that example, other environmental differences that you'd have to account for?

I said almost a hundred percent of thought experiments, but yeah, the point would be that. Or there's a, there's a, I think there's an opposite example, which might be useful and also make your point, which is like. Uh, you could get a hundred people who all have a genetic vulnerability to alcoholism and you could put them on an island with no alcohol and no one would become an alcoholic.

The phenotype would never come out. So I think it's very important going forward to do research which combines genetics with environmental measures, but it's just very difficult to convince. It seems my experience, geneticists and the MRC and people like that, we should do that, but that's what I think we should do.

Yeah, when I've spoken to Catherine Lewis, who is a professor of statistics actually, but in the genetics [00:19:00] lab at the IOP, the Institute of Psychiatry at Kings, she was advocating for this more multidiscipline multidisciplinary approach where psychologists, psychiatrists, geneticists can collaborate more to try and.

Yeah. Across these disciplines. I can't, that gets 100 percent of my endorsement, so that's why I think, and if we did that, the interesting thing is we would find out much more about the mechanistic pathways, because the mechanistic pathways are going to be important in terms of helping people. For example, I mentioned that I'd spent a lot of my time working on CBT for psychosis.

I've put an awful lot of effort into that, I've been involved in a lot of clinical trials. And I like to think it's probably done some good in the world because it has been taken on by NICE. But the reality is that although in the early days we told ourselves we were basing our CBT ideas [00:20:00] on some kind of understanding of the psychology of symptoms, mostly that wasn't really true.

It was developed very pragmatically. And I think CBT, the evidence is It's not panacea, it helps people, some people it doesn't help at all. One of the things is we need to know, to be able to predict who it helps. But if we could understand these mechanisms, we would be much, in a much stronger place because we might be able to find psychological interventions which actually work for particular types of mechanistic pathways.

And actually I'm In this late stage of my career, working on that, we've just done, uh, with colleagues who just done a feasibility trial of EMDR for psychosis, which is a, a therapy, which is not some people count it as a form of CBT, but it's not really, it's quite different, but it's a therapy which aims, targets the traumatic memories, which people have when they've had trauma in their lives and our [00:21:00] idea is to see whether that would actually also help to change psychotic symptoms.

Yeah, that could be really interesting. So, I mean, I'd like to offer my, how I think about psychosis and schizophrenia and you can tell me what you think about that. So I guess, and this really correlates with what I've seen clinically, so I've seen people a lot who come into A& E, for example, with psychosis, some kind of emergency has happened.

And what I see is that the following seem to go together a lot, and that would be erratic, chaotic, disorganized behavior, disorganized speech, thinking patterns, so like sentences which aren't coherent, which don't seem to make sense, hallucinations. So the. sensation that you're either hearing a voice that doesn't seem to be coming from any person or seeing something that other people don't see.

And then delusional beliefs, which we could spend a three hour podcast trying to define, but we could I'm writing a book [00:22:00] on it at the moment. Well, there you go. Yeah. But, but roughly delusion, delusional beliefs. Beliefs which are seemingly quite bizarre, quite rigidly held, that it's hard really to talk someone out of, or reason with, or entertain any kind of doubt around, and importantly, which seem to be out of keeping with that person's cultural context.

And I see those I have seen those go together quite a lot, and then what I see typically is either the person makes a kind of complete recovery that can be with medication, but it might be that if the person was, for example, using a drug like cocaine, amphetamines, cannabis. And that seems to have induced those symptoms, which we know it can.

Sometimes it resolves spontaneously within a few days, without needing to add any medication, just providing basically a safe environment. So that's what I think of as psychosis. And then what I [00:23:00] think of as schizophrenia is, if psychotic episodes keep happening recurrently, so those symptoms that I described.

If they keep happening throughout life, sometimes they might happen several times and then resolve, uh, sometimes with treatment, sometimes without. Uh, sometimes what I've seen quite a lot, uh, in hospital is people who have those symptoms I mentioned, kind of at a low level, but chronically. So they might experience hallucinations chronically, unusual beliefs chronically, but also importantly, the speech and the behavior and a lack of, perhaps a lack of self care as well, I should add, what are called negative symptoms.

And my basic beliefs around the causation are like, perhaps a, no, a likely genetic vulnerability, but then trauma, life stress. Other stress, other environmental influences like drugs, and perhaps if those stresses, traumas, the drugs weren't present, perhaps those symptoms [00:24:00] might never show. So that, that's my thinking about the etiology and the presentation and how, how does that align or not align with your thinking?

First of all, straight observation of people having psychiatric emergencies, but I think you're describing them pretty much as they often are. It's interesting, I'm thinking about your description and you, you put a lot of emphasis on disorganization and particularly disorganized speech. Now, I've actually done a few studies of disorganized speech, but hardly anybody else has.

Weirdly, it was much studied by psychologists and linguists actually in the 1980s And then I don't know why it just went out of fashion, but it is that feature of psychosis is one which is not particularly well understood. But one thing which we do know about it is it's highly linked to emotion. For example, I've done [00:25:00] two or three studies where, and actually with both bipolar patients and schizophrenia patients.

Where we simply ask the patients who have a history of that type of disorganization, we ask them to talk about things. And in one condition, we ask them to talk about something banal, what they've seen on television. And the other one, we ask them to talk about something emotionally stimulating, such as what were the events which surrounded them going into hospital.

And it, what happens is you just get the disorganized speech when they're talking about the emotional stuff. And other people found that as well. So we know that is quite linked to emotion and that there's some evidence that When you're emotionally charged up, like that is very particular, very difficult to calibrate your speech to the needs of the listener.

So in a typical conversation, you're very, you're unconscious, like we're having now, you're unconsciously monitoring whether you think the other person has understood. [00:26:00] And that process is somehow disrupted when the person becomes emotionally aroused. Interesting thing is that it seems that if you find somebody who's disorganized like that, and incredibly difficult to listen to them, I've tried, what happens is you can pay attention for a very short period of time, but then after a while, your brain just drops out because it's so difficult to see what they're saying.

But if you come back to them later when they've, when their symptoms have ameliorated a bit, and if you ask them what they were talking about, they can tell you. And it's usually emotionally distressing stuff. So what you're describing as that kind of crisis is actually to some extent an emotional crisis.

And it usually is something like that, which leads to a acute psychotic episode. Now, in terms of saying, if they keep having that, and if it persists, we call that schizophrenia. Fair enough. But that's just a convention, really. And [00:27:00] what's not clear is that schizophrenia, as I mentioned before, is distinct from bipolar disorder.

In fact, one of my PhD students just published a really neat study of that in Schizophrenia Bulletin, where we took data, we got, managed to get data from others, from previously conducted studies. We had over a thousand patients data, I think, and we show that in many ways, when you look at the structure of the symptoms for schizophrenia in bipolar patients, the underlying structure is actually the same.

So they look different, but it's probably different presentations of the same thing, basically. And Yeah, so have I answered your question? I'm not sure I understood the last point, and I'd just like to, just before we go into that, I'll just inform the listener, so we, I explained a little bit about how psychosis is typically thought of, so When we're talking about bipolar, we're talking about conventionally we would describe individuals who have what's called manic episodes.

That's the hallmark and that would be [00:28:00] episodes of weeks to more rarely months where the person has very high levels of energy, fast pressured speech, their thoughts might be racing. They may then have psychotic symptoms on top of the mania, which I suppose is where it becomes diagnostically uncertain.

But more typically in mania, the, the delusional thoughts, they would have, they would be congruent with that person's mood. So if they feel really elated, they might feel they're like a prophet. They might hear voices that correspond with that, for example, hearing the voice of God. But it's also important to point out that there are many cases of Mania, and I've seen them, what we would call mania, that don't have psychosis, so the person is just really high in energy, they speak very quickly, they stop sleeping, they might sleep two to four hours a night when normally they would sleep eight, their appetite is vastly reduced, they may also engage in, in more [00:29:00] risky behaviours than normal, that might be like sexual behaviours, gambling, spending money, things like that.

So that's, and then obviously, but obviously, uh, aside from manic episodes, they might have episodes of depression, which people tend to understand what that is. So could you just explain to me now that we know that that's that last point, what, how, how are they more similar than we realize? It's quite technical.

So just as a bit of background, I mentioned already, but there are lots of reasons for thinking that these. Disorder is not entirely separate. And one clear reason for thinking that is because of schizoaffective patients. So schizoaffective patients are patients who have a mixture of both sets of symptoms.

And I would say I've seen a lot of that. Yeah, I would actually say it's probably more common than pure schizophrenia or pure bipolar. So, there's a lot of debate about how to make sense of that. And I guess a good thing I would say is that when I first started writing this [00:30:00] area, and I wrote stuff, which is skeptical about psychiatric diagnoses.

My first paper on that was in the late 1980s. People thought I was nuts, I'm pretty sure. But I don't think they think I'm nuts now because I think nearly everybody thinks the classification system is not really working. So the question then is what do you do to fix it? And the answer to that is nobody knows, but there are plenty of ideas in town which have been explored.

A lot of these ideas involve using big statistics, using quite powerful statistical tools to make sense of the actual, the idea is you find the actual pattern of symptoms. Try and find actual syndromes, that's groups of symptoms which go together. So really what we would have wanted from the DSM in the.

first place, a true statistical manual. Yeah. And one idea which has emerged out of that is that maybe all patients with psychosis, maybe all of them can be described in terms of five dimensions of symptoms. It's just that different patients have [00:31:00] different profiles. So five dimensions are positive symptoms, negative symptoms, disorganization symptoms, depression, and mania.

And the idea is that all this is just a hypothesis, although there's quite a lot of evidence to support it, but I wouldn't say it's a kind of proven. So the idea is that all people with psychosis, whatever their diagnosis, they've just got different profiles within that kind of five dimensional space.

And so they all belong to some similarity, essentially, even though they look different. This is where it gets a bit technical. Okay. So. Those ideas are based on something called a latent variable model. And what that means is this, that we think the symptoms are expressions of something hidden, which we can't see, a disease process.

Statistical terms, that's represented by a method which we call factor analysis, which is a technique which estimates what that thing might look like [00:32:00] from the pattern of the symptoms. And those who produce, who support the five dimensional model, they basically say there are five latent factors. So you could say that psychosis has got five separate disease processes, which appeared to different degrees in different people.

There is an alternative power, which is to say that. Maybe there is no hidden process as such. Maybe what happens is you get patterns of symptoms because one symptom causes another, which causes another. Possibly in the kind of typical way, and that's called a network model. And there are very powerful statistical techniques, which have only been around for about 10 years, which allow you to see whether the data you get from patients fit that model.

Now, what you get is a fancy picture of a network, however symptoms appear to be influencing each other. Now, I should say there are lots of caveats to this. Is it causal? We don't know, but basically you can get these compounds. And that's what we [00:33:00] used looking at schizophrenia and bipolar patients. So what you're looking at is the interconnections between the symptoms.

The extent to which each and what we're assuming is that each of the symptoms are influencing each other and you can measure the influence. But it's an assumption that's what's happening. If you do that, it turns out that the pattern of influence is the same in bipolar disorders and in schizophrenia patients.

So all of the symptoms look very different. on the surface, the interconnections between them are the same and they're strikingly the same. That would, would also fit the idea that we've got one general psychosis, which is expressed in different ways under different circumstances. I don't know if I believe that by the way, I just think it's a, a tenable hypothesis.

Actually, I'll go a little bit further. I'll stick my neck out and say what I think is likely to turn out to be true, but we don't know how to test it.

So you've got these two different [00:34:00] kind of ideas. One is that symptoms go together because they're caused by some underlying process, the disease. The other is that symptoms go together because there's a characteristic way in which symptoms influence each other. We don't have a statistical test to tell the difference between the two of those.

At the moment, we can do a network analysis of our patient's data, and we get a nice picture, and we can do a factor analysis, and we get a nice picture, and they both look convincing, but there's no way of telling which one's really convincing. Reality, almost certainly, it's some hybrid of both, but there are some underlying processes which influence symptoms, but also the symptoms influence each other.

And we, it's beyond our statistical ability at the moment to construct and test such models, but I wouldn't be surprised if it becomes possible in the future because one of the big unsung stories of my [00:35:00] career, not my career, my personal contribution, although I did make one small contribution in this area, is just the phenomenal advance in analytic methods, which are available.

So the statistical techniques that we can use now are. Incredible compared to what I was taught when I was a student. Yes, and with the onset of AI, I imagine the ability to analyze large data sets is just going to improve exponentially from here. Yeah, I know lots of people who think that, and I know lots of people who are dabbling in machine learning as a kind of technique.

I think, let's see. I know some skeptics about it as well. The big argument about that, but one of the big problems with that is that you can set your AI to start analyzing stuff and it can give you an answer, but you don't know why it's got the answer. It'll come up with predictive models. At the moment, I've just been, just on a small toy project, tried with a friend who's, colleague who's into that stuff.

We tried to see [00:36:00] where we could predict outcomes in longitudinal data set. This gets three patients that would, that's ongoing at the moment. But we'll probably end up with some combination of variables, which look like if you. Combine them in a particular way to predict a bad outcome, but then you go, why, um, Yeah, it's always worrying when the AI, the AI, if it does something and you have no idea why, I, I heard some anecdotes once that Uh, some AI algorithms were influencing, were, were designed to run, help run the stock market in China.

And then one day they just decided to shut the market down and they had no idea why that was happening. I've not heard that, but I guess it's not implausible. I use AI in my work a little bit. but I find it's really helpful for some things. I think you've got to treat it cautiously at the moment anyway.

Okay. So we've already dived deeper in terms of the possible causation of conditions like psychosis and bipolar disorder, but now moving on to, you know, what to do. And, and I [00:37:00] think one of the real revolutions of the past 10, 15, 20 years. in, in treating patients with these conditions is actually saying there is value to understanding how you got to where you are and there is psychological meaning to be made.

out of the symptoms that you have. So if, for example, you have a paranoid delusion, it may not make a lot of sense at face value. At face value, it may seem like you've lost contact with reality. But I think there's a lot and there's a lot of what your work proposes is there's a lot of we can understand about a person through that.

So I wonder if you could talk to me a little bit about what might be some of the underlying psychological mechanisms underlying things like delusions or, or, or hallucinations. Let's take delusions. Delusions I find endlessly fascinating, partly because of something you alluded to earlier, which is the difficulty of distinguishing them from extreme ideologies, for [00:38:00] example.

And that's why trying to write a book to try and solve that problem because writing a book really makes you think about stuff. But basically, if you take, we've actually published a, uh, a review of the world literature on this, which is, I say, we are one of my students and I, which it turns out that across the world are five major delusions, which come up all the time.

Paranoid delusions, reference delusions. Grandiose delusions, delusions of control, and religious delusions. Now, it has to be said that there are a lot of very obscure kind of delusional syndromes which seem to really excite neuropsychologists, like misidentification syndromes and things like that, but Capgras syndrome, but But putting those two on side, because they're very rare, it's these five main types.

But if you think about them, they've all got some kind of meaning attached to them, because they're addressing what are the real existential dilemmas of life. One of the things we have to decide in our [00:39:00] everyday lives, It's just constantly it's who to trust. And we know, for example, from studies that if you meet somebody who you've never met before, you meet somebody for the first time, you form an immediate impression about whether they're trustworthy within about 200 milliseconds.

So it's faster than you can think. So something is going on pretty automatically. You're just picking up autumn cues, automatically making that judgment. Of course, you can then adjust the judgment later, dividing the world into those who are trusted. Those who are not is very important for survival. And paranoid patients seem to be doing it slightly differently.

And we know that typically paranoid patients have had genuine victimization experiences. There's a nugget of truth in paranoia all the time. And actually, it's quite rational to be paranoid under some circumstances. That's paranoia. Reference, we all have to scan our environment over, over time and decide which stimuli are [00:40:00] relevant to us.

That's essential to survival. And looks like in reference delusions, that is somehow not going quite, quite right. What's, what's happening when someone has a reference delusion? What do they typically report or describe? I should say these are probably very little research by psychologists, but a very common thing, for example, is people saying that.

The radio announcer or the TV presenter is actually addressing a message directly to them. So there's some significance in, they might think that the actual presenter is actually talking directly to them or there's some secret message encoded in what they're saying. That will be a reference delusion.

Next one is delusions of grandiosity. Again, what's in everyday life. All of us, we human beings are social species and we naturally compare ourselves to other human beings and we compare ourselves in a kind of social ranking, [00:41:00] whether that's a good thing or not. We do it pretty automatically.

Interestingly enough, it turns out that being low social rank or thinking yourself as low social rank predicts, is a predictor of every type of psychiatric disorder there is. That's an important issue, and obviously it's going wrong in some way with grandiose patients. Delusions of control. Most of us have concerns about autonomy.

How much are we in control of ourselves and our lives? And so you can see how that connects. And finally, religious delusions. What is the meaning of life? And what happens after I die? So although delusions express come out in a kind of weird way, all of them have got, all the main ones have got some link to core existential themes.

And so far as we know, the only one which has been studied intensively is paranoid delusions. They seem to be related to life events as well. In a way, I suppose the important point is [00:42:00] that, that psychosis is part of humanity in some sense. It's part of that sort of. It's a kind of expression, a dysfunctional expression of all the things which people worry about in their lives.

Yeah. I suppose what you're saying is all of those different delusions reflect really fundamental needs and uncertainties. We all have. Yes. The uncertainty about, the uncertainty about who to trust, the uncertainty about how relevant different things are, the uncertainty about how much we're worth, the uncertainty about what life is all about and its meaning.

And I guess when people's needs go really, really unmet, traumatic experiences, this delusion can come up perhaps in consequence. We know quite a bit about some of the psychological mechanisms involved, particularly in paranoia. Low self esteem seems to be very important, but actually. part of paranoia. If you fit, if you feel low in social rank and you feel that you're very low self esteem, then you also feel very vulnerable and open to [00:43:00] persecution by others.

And we also know, interesting thing I think is that, uh, persecutory delusions are particularly related to attachment style as well. That's part of the same story really. Probably most of your listeners know that the bonds we form with our caregivers when we're young provide a kind of template for. how we handle relationships as we grow up.

In the case of paranoid patients, patients tend to be what we call insecurely attached or anxiously attached, which is they tend to assume that they are not worthy and that nobody would really love them basically is what it boils down to. But this seems to be a kind of template which they've learned through life.

The thing that strikes me on discussing these topics and reflecting about my time, particularly in inpatient services, hospitals, less so in the community, but definitely in hospitals, is there was not much [00:44:00] of an emphasis on trying to understand someone psychologically in a hospital, certainly not trying to understand someone's unmet needs, and not A lot of emphasis on trying to create an environment which is non threatening.

It's complicated because a lot of times when people are in acute psychotic states or manic states they can be dangerous, it can be an unpredictable environment, so there's this immense, there's this need for safety and risk aversion and that's important, but at the same time it felt like a lot could have been done to make an environment more positively nurturing.

more comfortable, less threatening where, where possible. And I feel perhaps it's something we're really missing in our public mental health system. I 100%, 1000 percent agree with you because actually, I'll tell you just a kind of [00:45:00] interesting thing, which came out of one research study I was involved in. I mentioned before that I'd made a small contribution to statistics.

It wasn't really my contribution, but I was involved. It was, I had a lot of data on the quality of the relationship between the therapist and the patient, what we call therapeutic alliance in one of our trials. And I got talking to a wonderful statistician called Graham Dunn about how to analyze the data.

And I thought I knew how to analyze the data using simple regression techniques. Graham was almost, I wouldn't say he was appalled. I expected him to say, yeah, go ahead and do that. But he said, there's a fundamental problem in statistics here, which has never been solved, which is not one for something you actually want to hear from a statistician.

And amazingly, he then took 10 years to solve it. Basically, what he wanted to show, what he wanted to show was that the therapeutic alliance was causal in whether the therapy [00:46:00] worked. And to do that, it was necessary to eliminate what we call confounding is third variables, which could confuse the picture.

So for example, suppose there was something which made you like the therapist, but also made you get better. It would then look as if the amount you liked the therapist made you get better, but that wouldn't be true. It would just be a spurious correlation. I'd thought of that, of course. So I built my models.

So they had all sorts of control variables in everything I could measure, basically. But as Graham said, there's always something you haven't measured. And maybe it's the color of the wallpaper in the room where the therapy was carried out would influence whether the patient felt good with the therapist and also what, what Also, whether they'd get better or not.

Anyway, Graham developed or adapted, because they were first used in economics, a methodology called instrumental variable analysis, which actually allows under some circumstances, and this is one of them, you could, to show causation. Absolutely. And it [00:47:00] turned out the therapeutic alliance was causal. We'd proven that the Therapeutic Alliance did make a big difference.

It was the main feature of whether CBT worked. And that's even in cases of psychosis or just CBT generally, or psychosis? So it was just with psychosis, it was just with psychosis. But a big trial, 300 of patients. And interestingly enough, and this was a very important finding, just as important really, we found that people had a bad Therapeutic Alliance, CBT made them worse.

So, having a therapy from somebody you didn't have a good relationship with was damaging. But that got me thinking, which is, because we put a lot of effort into the idea that CBT is a kind of technical, I often think it's a bit like a prescription pad for some psychologists. This is our technique. It's manualized.

You just follow the manual. Yeah. To be fair, most psychologists don't quite think of it like that. The manual guides you, but essentially it doesn't work unless you've got this high quality relationship. [00:48:00] Yeah. What about the relationship between the doctor, the prescriber, and the patient? What about the relationship between the nurses on the wards?

I don't know if this is true, but I would actually bet a very considerable portion of my meager wealth on it. Which is that if you could improve all of those, you would actually massively improve patients outcomes. If you could make sure that every relationship in, say, inpatient wards was of the right sort, from the cleaning person up to the consultant, then you'd get much better outcomes.

Have you ever heard of the Trieste model? I have, yeah. I don't know that much about it. Weirdly enough, I was in Italy, actually giving some lectures in Padova recently, which isn't that far from there. And I did wonder about going over, but no. The reason I bring it up, because it seems to be the best example to my mind of trying, of attempting this, of really having a model where [00:49:00] Uh, the mental health system firstly was directly integrated with the social services.

So this sort of much more fluid, um, but also it was person centered really from the ground up, uh, with patients given opportunities to participate in their local community a lot more readily and more easily. Uh, and really, uh, de emphasization on coercion, mandatory use of medication, use of the Italian equivalent of the Mental Health Act.

And it would, by all appearances they achieve very, very good results. And I had someone who really ran their system for a number of years, Dr. Roberto Medzina on the podcast a few months ago. Oh, I wish I had seen that. Yeah. Yeah. I mean, I can send it to you, but that seems to be the best instantiation of what you described is what, what if we had a mental health system where [00:50:00] relationship really was put as a priority.

Do you know about Soteria? No. Okay. So it's another thing, which is a bit similar. So it was created by Lauren Mosher, who was an American psychiatrist who I knew just before he died. He was an extraordinary person, actually. So the story goes that Mosher was appointed to be head of schizophrenia research at NIMH in the United States.

And just before He was about to take up the post. He was on holiday in London and he'd heard about R. D. Lange and, and Philadelphia House and Lange's experiments with the kind of therapeutic community for psychosis. So he decided to visit. Now, Lang couldn't organize a piss up in a brewery. He was hopeless.

And although he did actually spend quite a lot portion of his life actually pissed, apparently. So Moshug visits Lang's facility and he was appalled because there was, the place was in chaos and nobody had decided who was going to cook the next meal, that sort of thing. [00:51:00] But he thought, what if he did it properly?

And he got a grant from NIHR to set up this thing called Satiria House in Oakland, I think it was. And it was just a place where people with first episode psychosis could go. And they were cared for by people who were not mental health professionals, but were selected purely for their interpersonal skills.

And they were more or less told, just be kind. And they kept medication to an absolute minimum. And it had very good results again. But it didn't plea, please. Mosher's master's at NIMH and he eventually had to leave his schizophrenia, his position as chief of schizophrenia research. Was big arguments over it.

Robert Whittaker, the American journalist has written quite a lot about it. And as I say, I talked to Mosher a bit about it himself before he died, but there are different versions of exactly what, why he ended up passing ways with [00:52:00] NIMH, as you can imagine, depending on which side you're on. But, you know, it sounds like a very similar thing that it was a.

Just providing a very kind, supportive, but structured, with some rules, environment where people could, felt safe. And it strikes me that, you know, we can, we can, there's two worlds, and we could get lost in both worlds, and one world is what's causing these problems, whether it's depression, psychosis, mania.

And as, as our listeners will be aware, if they've listened to this whole conversation, it's really complicated and very mysterious still. And then there's this other world of like, given that, given that people have these problems, what is the best way to interact with them? What is the best way to treat them?

What is the best way to contain and help them make progress in different ways? And often we can get so lost in the first world and maybe get bought into certain ideas or small silos of it. that [00:53:00] we lose some of the more common sense paths forward. I'd agree with that. But the other thing I'd like to say about that is that if you have a kind of psych, I've in recent years began to think of myself more as a social scientist than a biomedical scientist.

But if you think you're applying the tools of, of social science to understand these. strange behaviors. And if you apply them properly, the picture which you get is actually quite a humane one, right? I think it was Wittgenstein says something like the best picture of human nature as a human being, which is the sort of you.

It turns out that if you do the research trauma, life experiences, bad stuff, which happens to people, this is all very important. And in terms of a causal factor, and it also turns out that if you're working clinically people actually want to talk about those things. And they really like to have them, their stories taken [00:54:00] seriously.

And that's usually therapeutic for them. You don't have to be a kind of genius psychotherapist to actually just. treat the patient as a human being and just say, tell me what happened. What's the story which got you to that position? Just providing the opportunity for them to tell that story is actually very helpful for people, but also it provides a lot of information for a clinician.

And as I say, one of the weird things is that what you find when you do that is often you can relate it to psychological models we have for how these senses develop. Absolutely. And it just reminds me of what I've learned in psychotherapy training, which is that. like 80 percent of the effectiveness is just the basic skills of listening, empathy, mirroring, clarifying, just being a good listener gets you most of the way there.

And then there's some really clever ideas that you can put on top of that sometimes, and that can work or not work. But there's a foundation that's Late in my day, I became a fan of Carl [00:55:00] Rogers. Yes. The American psychologist who wrote a lot about that, Roger was a huge optimist. He thought that you could actually, the task of the therapist, people could cure themselves.

And if you provided the right circumstances, and he thought that the task of the therapist was just to Provide the right circumstance. I don't agree. I think that's going too far. But he pointed out there are some essential listening skills, some essential therapeutic skills. You have to empathize with the patient.

You have to be genuine with the patient. You have to show them positive regard. and look for the, which means looking for the best in them. And if you can do those things, then you'll go, you you're well on your way to being a good therapist. But what's underappreciated is how difficult that is. I was once asked by a journalist, if I had a superpower, what would it be?

And I said, being able to show empathy, genuineness, and unconditional positive regard on a Friday afternoon is a superpower, which is my [00:56:00] glib answer. Because it is in a way. You're tired. You've been doing clinical work all week, or in my case, you've been up to your eyes and trying to help PhD students and lecturing and grant writing, stuff like that.

And then you're in this room with this person. It's Friday afternoon. You're exhausted. They're probably, it's probably not the best time for them. It's so easy for your brain to suddenly start thinking about the shopping list. Actually, directing your attention and giving your attention to people in a way which is helpful is not easy.

Yes. And I think Rogerian philosophy actually runs pretty deep and a lot of why it takes so much discipline to do that kind of therapy is not just what you mentioned, but also the emphasis on being non directive of listening and not, not reflexively telling a person what to do or giving them advice or telling them, giving them a prescription takes an enormous amount of discipline.

And if you don't think it does, [00:57:00] I suggest try doing that with your partner for. for a month. Try not telling your partner what to do for a month and email me and let me know how you get on with that. So, yeah, no, I think it is hard to do that. And it's, and as I say, it's a real, a real skill to be able to do it.

And some people are not suited to it. Yeah. And perhaps if there's one, one takeaway, it's that, you know, how can we devise a way where more people working in mental health settings, particularly the acute settings have these skills and. It can be more deeply embedded in the ethos of the organization somehow.

Yeah. Big question, maybe perhaps for another podcast. Yeah. That would be a big task, but if you could do it, it would make a huge difference. Well, Richard, thank you very much for coming on today. We're out of time, but it's been wonderful to speak to you. I look forward very much to your book about delusions.

Let me know when it's out and perhaps we can, you can come back on and we can talk [00:58:00] about it. And thank you very much for spending some time with me today. No, it's been a pleasure and thank you for your interest and yeah, okay. It's been very nice to meet you.