Basing the evidence
These essays have offered up a number of propositions about mental health issues:
· It’s always the same underneath – look carefully enough, and far enough down, and the problems always revolve around the same themes.
· Fundamentally mental health is about self-esteem. It’s the collapse of self-esteem, or its precarity, that causes people so much distress they become ill.
· Mental health is about the cross-over between what’s inside people’s heads and what’s outside them. Problems arise when the contents of people’s heads become too vivid, too real. Health is about being able to engage with external reality without too much disruption to you – being able to meet what’s outside your head half-way.
· And it’s fundamentally about other people because they’re the most significant feature of what’s going on outside your head. Other people are where you get your self-esteem, or it goes to die.
So, it’s time to provide some evidence. But first you have to look at the nature of evidence in Psychology. And, here, psychology includes psychology, psychiatry, psychotherapy and every form of treatment for mental health issues you’ve ever heard of.
Science is supposed to be objective, measurable and repeatable. In Psychology it almost never is.
Psychology started recognising its replication crisis in 2010[1] (obviously in reality it was always there, it just wasn’t taken seriously). It became clear that research, including experiments going back decades which had been regarded as seminal, just didn’t turn out the same way when repeated. One recent review looked at 100 pieces of research published in one year in three leading academic journals and found only around a third of the results could be reproduced.[2] That’s an immediate, unqualified fail for repeatability.
Psychology isn’t like physics: you don’t have unimpeachably objective measures like weight or speed to work with. The gold standard of objectivity in experimental research in medicine and the social sciences is the double-blind Randomised Control Trial. That’s a study in which you set up randomly allocated experimental and control populations and neither participants or observers know who’s in which group. The objective is to eliminate the different forms of bias that could influence the reported results. Generally, that degree of ‘blindness’ isn’t even possible to attempt in experimental settings: if you’re exploring the effects of social exclusion in a cybergame, for example, it’s usually plain, to subjects and observers, who’s who and (at least to the observers) what’s being done. And when Psychology tries to adopt the RCT, the objectivity often gets compromised for ethical or practical reasons: if you’re doing a comparative study between different forms of therapy in a clinical setting, for example, its arguable participants are entitled to know which one they’re getting and, in practice, they’ll probably guess (it’s not like a choice between two identical looking white pills). And when it comes to scoring the results, they’re judged by questionnaires which have been validated by subjective assessment - assessment by numbers of clinicians, but still subjective (and, actually, very unimpressive levels of corroboration). So, no to objectivity too, certainly when judged by the standards of physical science.
Even with RCTs, there’s plenty that can go wrong with experimental design - sample size and selection, scoring and calculation and controlling for extraneous factors. Psychology isn’t like chemistry: you can’t put people in a test tube and isolate them from everything apart from the experimental variable you’re trying to introduce. The idea is that the statistical power of the experiment will get you beyond the effect of different confounding variables on individuals to a true, smoothed-out picture of what’s going on. But psychological research tends to be time consuming and expensive, and it is often statistically underpowered: sample sizes are not big enough. And when researchers try to compensate by using meta studies (reviewing and pooling the results of multiple previous studies) they introduce the risk of inconsistencies in experimental design and methodology. And, in practice, when you put things together the results break down (replication again). When you look at the results of comparative studies between different forms of therapy, it seems they’re all about the same:[3] it appears that what predicts effective treatment isn’t the form of therapy, it’s something else that isn’t being captured in the research.
And that’s where the biggest problems lie for Psychology: knowing what you’re measuring. People’s states of mind are inherently and irreducibly subjective; attempts to turn them into quantifiable data risk absurdity. How bothered have you been in the last two weeks by worrying too much about different things or having little interest or pleasure in doing things? Those are questions from two of the most widely used diagnostic tools for anxiety and depression.[4] Different answers might be the result of quite significant differences between people (at least at that time, on that day). But there really isn’t much reason to believe that the given answers (at that time, on that day) can tell you very much about whether any of those individuals are suffering from depression or an anxiety disorder.
A good way to illustrate the problem is by exception: taking one of the rare situations where Psychology can be reasonably confident of what it’s measuring. Several studies have shown that if you tell people they are about to receive acupuncture and they may experience certain side effects (including muscle ache and pain around the acupuncture site) a statistically significant number will report those symptoms even though they were actually ‘treated’ with retractable needles which didn’t break the skin at all. In these narrow terms, that is pretty good evidence because it’s clear what it’s telling you: it’s telling you what people thought (or said) - in this case that they were experiencing physical pain that couldn’t possibly exist (called the nocebo effect).[5]
In 2018, a study out of UCL reviewed 20 years’ worth of national cross-sectional surveys and, to nobody’s surprise, found very significant increases in reported long-standing mental health conditions amongst children and young people in the UK (a six times increase in England).[6] That chimes with massive increases in the figures for people, especially the young, presenting to medical services for help with mental health conditions.[7] They’re both taken as evidence of an epidemic of mental health issues, especially amongst the young, and that’s all been very widely reported. But, actually, all you can be sure that either of these data points is measuring is an increase in the numbers of people who believe that they (or their children) are suffering from a mental health condition.[8]
The same 2018 UCL study found that when you use recognised clinical assessments (rather than self-reporting or going to the doctor) no significant increase appeared in the numbers of children and young people experiencing mental health conditions over the same 20-year period. And that chimes with the results of the Mental Health of Children and Young Persons (MHCYP) survey (which is the largest study of the mental health of the young in England involving clinical assessment) over the same period and the Adult Psychiatric Morbidity Survey (APMS), which does the same job (with clinical assessment) for adults, which also showed broadly static levels of mental disorder over the same period.
So, there’s a rather significant question mark: are people getting iller or just more likely to think they are? Is there a kind of nocebo effect at large in the population? And nobody knows the answer.
Well, this is mental health, after all, so does it matter whether people who think they’re ill fit the recognised categories of diagnoseable disorder or not? If they (or their parents) think they’re ill and it’s causing enough distress and unhappiness, doesn’t it need attention?
The answer is probably yes, but it may not be the same attention as someone who does hit clinical thresholds in relation to recognised conditions, which is the attention they’re getting at the moment. And the point here is the ambiguity of the data. All you can be sure of is that dramatically more people are presenting for help, but this is unhesitatingly taken across the board as evidence of a massive increase in the underlying levels of disease. And a moment’s thought will reveal that Psychology has a very obvious financial interest in the data being treated that way.
Certainly, Psychology doesn’t help the situation by the way it sometimes carries out and presents the research. The MHCYP (carried out in 1999, 2004 and 2017) is described by the NHS as the best source of data on trends in the mental health of young people in England.[9] In 2020, and twice since, in response to the Pandemic the NHS commissioned short-form, mini ‘wave’ follow-ups to the 2017 survey. These short-form follow-ups did not include clinical assessment, which had been part of all previous MHCYP surveys.
The results of the original MHCYP series and the ‘wave’ follow-ups are routinely presented together to show a very dramatic increase in problems with mental health in the young since the Pandemic. When you take account of the disparity between results from self-reporting and clinical assessment in the 2018 UCL study, and the dramatic increases in demand for mental health services against trends seen with clinical assessment in the APMS and previous versions of the MHCYP, you realise how problematic that is.
The MHCYP was the best evidence of trends in mental health for children and young people – well, it isn’t now. The follow up ‘wave’ studies include footnotes saying that they shouldn’t be used for comparative purposes with the original series but it was entirely predictable that it would be and that it would show dramatically higher levels of (self-reported) issues. And it was, and it did, and that’s the narrative everyone has taken on board, and they regard it as scientifically proven (and we don’t know whether it’s true or not).[10] The use of expedited short-form follow-ups (with smaller samples and no clinical assessment) can be explained as a practical expediency during the Pandemic but it is extremely bad science.
It is issues around subjectivity and the potential for suggestibility, confirmation bias and the predictive quality of people’s beliefs and assumptions to distort the results of psychological research that appear to be at the heart of the replication crisis. When modern researchers investigated, they were often able to uncover undetected confounding factors in earlier studies that were to do with participants’ subjective expectations. So, for example, an early casualty was the famous Stanford Prison Experiment in the early 70s once it became clear that the results had been significantly influenced by ‘demand characteristics’ (participants readiness to take up the roles they felt had been prescribed for them). Not least, it came to light, because a number had been explicitly ‘coached’ by experimenters. So, quite a different perspective on people’s apparent willingness to descend into Lord of the Flies cruelty when placed in positions of power without accountability.[11]
Psychology is learning that things ‘work’ for people, to an extent much greater than previously realised, if they think they’re going to work. Thousands of pieces of research were pre-determined by the mind-set of the people coming into the experiments and nobody knew. And that doesn’t just involve ‘catastrophic implications’[12] for decades of psychological research, it implies that it is likely to be impossible to eliminate, or control for, the determinative effects of participants’ subjective expectations of outcomes in future research. These issues have led some researchers to conclude that ‘psychology is fundamentally incompatible with hypothesis-driven theoretical science’.[13]
But, in reality, the practical limitations of quantitative data in mental healthcare have always been blindingly obvious. Take the controversy around the effectiveness of antidepressants. There can hardly be a more urgent topic, we’ve had 70 years of constant research and the debate still rages. A professor called Irving Kirsch has dedicated his career to proving, with empirical data, that antidepressants are no more effective than a placebo.[14] But, at the same time, antidepressant prescriptions have gone off the scale: more than tripling in England between 1998 and 2018[15] (with reports of further significant increased usage during the Pandemic[16]). If Kirsch is right, there are people who should probably be in prison, if he’s wrong why is he lecturing at Harvard with a chair at three other universities?
The problem is falsifiability, which is meant to be another hallmark of science. The science is always too weak, there’s too many holes in it: no-one can ever deliver the knock-out punch. You can see it in the related chemical imbalance theory of depression (related because antidepressants were supposed to work by correcting the imbalance). Well-reported research in 2007 showed that when you reduce serotonin levels (and other neurochemical candidates) in the brains of people who are not suffering from depression, they don’t become depressed.[17] By the standards of psychological research that’s a clear outcome, yet it is estimated that 80% of the general public believe that it has been established that depression is caused by a chemical imbalance in the brain, it’s regularly referred to on popular websites, many GPs believe it and its even endorsed by researchers in the area[18] – you just can’t kill the beast.
Something very irrational is going on. Thousands of people (researchers, university staff, journalists, commentators, clinicians) have been involved for decades in conducting research, teaching it and promoting it and nothing of any real practical value has come out the other end. There is not a single meaningful contribution to the treatment of mental health which has been the product of experimental research. Nothing is being done because it has been discovered (intentionally) or proven by experimental research. The ‘discovery’ of antidepressants, for example, was an accident resulting from the development of drugs to treat tuberculosis and all the research since has been an attempt to retrofit a purpose that was never intended.[19]
On those terms, the history of experimental research in mental healthcare looks like a monumental waste of time and money. And, in practice, it is a complete side show: the drugs aren’t there because research has demonstrated a clear, compelling case for them (because it hasn’t). The defining features of mental healthcare don’t come from research at all, they are the product of a tacit consensus amongst healthcare professionals and other interested parties as to what they want mental healthcare to look like. In fact, that consensus survives precisely because it can’t be defeated by evidence and it is independent of theory, so it can be what people want it to be (and what they want it to be is ideas that allow them to feel better about themselves).
These aren’t new or controversial ideas. This kind of critique of research in Psychology has been there from the beginning: it’s embedded in the literature and no academic conference is complete without it. The trouble is when people hear it, they just nod sagely (because it really is unanswerable) and carry on doing exactly what they were doing. In part, that’s because that’s what they are qualified to do, that’s what their degree and training is and that’s how they feed their children and keep a roof over their heads (also, to be fair, no-one has actually come up with a demonstrably better way of approaching the subject). But there is more to it than people just working their ticket.
Acquiring knowledge, finding out, isn’t solely an intellectual exercise. ‘Knowing’, feeling like we know enough about what’s going on, is an essential human need (especially if it’s an area you’re supposed to be an expert about). The ‘thirst for knowledge’ is like a thirst: the ‘epistemophilic instinct’[20] is similar to a physiological drive in its urgency and, ultimately, its significance for well-being. People feeling that things around them make enough sense to fit into their schema of things – feeling that they can demonstrate that to themselves and other people - is vital for mental health. So, even something as supremely rational seeming as looking for evidence has its own instinctive, unconscious life and can get recruited quite easily for quite irrational purposes.
There is something bonkers about all those millions of pages of research that aren’t going to prove anything. And the lure is the trappings of science: the hypotheses, data, regression analysis, confidence intervals help people to capture the feeling that they ‘know’. Debates in mental health are like debates about whether England should have picked Jonny Bairstow for this year’s Ashes. If you don’t care, you don’t have a view and most people don’t care about mental health (unless and until they find they have to). If you care enough to want to have a view, you already have it – you already ‘know’ what you think because the discomfort of not ‘knowing’ on something that matters to you is too strong. And no-one is ever going to prove their case.
People use the statistics that suit to make their point about cricket in the same way as they do with Psychology (and the more complex and arcane the statistics, the better, because that marks you out as someone who really ‘knows’). At least with Bairstow, the test series happened so there was some opportunity to evaluate whether you were right or not. With experimental Psychology, it just goes on for ever.
Psychology is a broad church; it encompasses tea-and-biscuit family counsellors in Cotswolds villages and the medicine trolley in a closed psychiatric ward. And everyone can believe honestly and honourably what they want to because they can always point to data to support it: depression medication increases the risk of suicide, particularly in the young[21] but reducing access to antidepressants results in increased suicide rates amongst 10-19 years olds.[22] The result is an endless stalemate.[23] Psychology is a tolerant, open-minded field, animated by debate and that’s lovely but, in the meantime, we have no effective response to mental illness which is one of the worst things that can happen to anyone and appears to be happening dramatically more often around the world.
The history of experimental research in Psychology hasn’t just been a side show which has failed to deliver any practical results, it has been a distraction. ‘The failure of psychology is… that it makes inordinate use of methods that are a mismatch for the aspirations of researchers in the field, at the expense of valuable empirical research’ and that has ‘led to experiments that are uninformative and frivolous’.[24] The irrational, ‘inordinate’ use of methods that can’t be made to work here has got in the way of finding out.
But the same researchers (and these are full-on career scientists) who supplied that verdict on the history of experimental Psychology talk about the opportunities for a more observational approach: looking at what happens, noticing, looking for ‘obvious regularities’, seeing what’s ‘widespread, robust or ephemeral’, what ‘strongly affects our actions or life outcomes’.[25]
Evidence is always all around us and we constantly use it perfectly validly through observation, inference and intuition. We don’t need statistical analysis to conclude that the sun will rise in the east tomorrow (though it appears to be 100% so far) or that we should get out of the way of a moving car or that our partner is upset with us. Psychology has plenty of valuable evidence to offer about mental health and mental illness. It has just allowed itself to become so distracted by its need to look and feel like it ‘knows’ that it’s obscured what it is there to be read.
Sometimes that takes the form of research involving quantifiable data, other times it’s paying attention to what people can be seen to do and say, particularly what they say and do obliquely (this is Psychology after all). And there’s a lot to be learned too about mental health and illness from the way Psychology itself behaves, from the way people giving and receiving mental healthcare approach it.
What emerges when you draw some of these strands together is a compelling and consistent picture that ought to have a great deal more influence in how we think about mental health and treat it. It isn’t always evidence that’s objective, measurable, repeatable and falsifiable in the way that science is meant to be but, it turns out, neither is the science and the picture that’s emerged from that is inconsistent and incoherent and irrelevant to how we actually treat the subject anyway.
With that out of the way, we can look at some of the evidence that’s already there and begging to be taken notice of.
∞
It’s all the same thing: Something else that can help people, especially experts, feel that they ‘know’ is really detailed taxonomy: an elaborate system of classification with everything neatly labelled and nice clear lines in between. Mental healthcare has two! The Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD). They’ve both been around a long time, gone through several iterations and swollen enormously in the process. Each now has well over 300 different diagnoses, some of which go back to the origins of psychiatry in the 1890s.
But the dividing lines between diagnoses are anything but clear. Medical diagnoses are evaluated according to reliability (can people apply them consistently in the same circumstances?) and validity (do they represent something that actually exists, in the way a daffodil exists and is different from a dandelion?). The reliability of psychiatric diagnoses is not at all good: people cannot agree about how to apply them.[26] When the latest version of DSM was being developed it became necessary to change the previously used rating system so that the same patient being given the same diagnoses half the time was regarded as ‘good’ reliability and the same diagnosis being picked only a quarter of the time or even 1 in 5 times was ‘acceptable’.[27] Try that with heart disease.
And the validity is not regarded as good either: a 2019 study concluded that psychiatric diagnoses were ‘scientifically meaningless’ in that they weren’t identifying conditions that were really separate things at all.[28] Which is consistent with the results of genetic research. Genome analysis does not show different genetic configurations for different mental conditions, it indicates a common genetic base across different disorders - ‘a shared genetic liability among major psychiatric traits’.[29] And this isn’t an ‘early-days-so-far-as-we-can-tell’ kind of thing: the genome is mapped: you can see it and both at a base level, and higher pathway levels (groupings of genes working together), it looks like the same biological mechanisms across the different disorders.
The 2018 flagship study by the world-wide Psychiatric Genomics Consortium, comparing the genetics of depression with other mental disorders, found ‘significant positive genetic correlations between major depression and every psychiatric disorder assessed’ and concluded that ‘major depression is not a discrete entity at any level of analysis’. With the academic restraint required of a group of around 80 of the world’s leading geneticists, they confined themselves to ‘the current classification scheme for major psychiatric disorders does not align well with the underlying genetic basis of these disorders’.[30]
It doesn’t align well with the epidemiology either. The 2007 APMS examined co-morbidity (the incidence of people being diagnosed with more than one mental disorder at the same time). The results were remarkable. There were no negative correlations between any conditions: the presence of any condition increased the likelihood of another condition being present. And with the more common anxiety and depression based disorders there was more than a 50% chance of being diagnosed with at least one of a range of other disorders, including eating disorders, OCD and personality disorders.[31]
Similar results were obtained in the US with more than half the participants who received a diagnoses receiving more than one.[32] The largest ever study into co-morbidity, in Denmark which concluded in 2019, found that the risk of co-morbidity was pervasive across all mental disorders and that this risk persisted over time: you were more likely to develop a different diagnoseable disorder up to 15 years after receiving a diagnosis.[33] And that’s consistent with the Dunedin study (the longest running longitudinal study dedicated to mental health – that is, you follow a group of people through their lifetimes and submit them to regular psychological testing) which has found that 80% of people who receive a diagnoses go on to receive a different one at some stage and a pattern of people who receive a diagnosis experiencing regularly changing diagnoses over the course of their lives.[34]
That is quite compelling evidence: the people saying these are all different conditions can’t use their own diagnoses consistently and think it’s OK if they match 20% of the time, the evidence from genetics is that it seems to be the same biological mechanisms and processes involved all the time, if you’ve got one, it’s very likely you’ve got others and if you’ve had one it’s much more likely you’ll get another (even when it looks like you’ve got better in between). People may want to think they can identify 300 different flavours (because their desire to feel they ‘know’ has taken some irrational directions, and even though they can’t in fact identify them reliably at all) but it’s looking like it’s all ice cream underneath.
And that view has gained traction, leading to calls for a more dimensional or transdiagnostic approach to mental health.[35] The trouble is, so far, that hasn’t really led anywhere. It may be a more intellectually honest position to take but it hasn’t resulted in any practical developments in how to respond to mental illness. It sits comfortably enough within the free-floating, loose consensual, value-driven framework to mental healthcare which has come to pervade all the different fields of Psychology, but probably the best defined position to emerge from it to date is that mental illness isn’t really a ‘thing’ (isn’t ice cream at all): what we have thought of as mental health issues are the results of normal human distress in response to inequality, injustice, abuse and neglect.
Whatever the philosophical merits of that argument, it’s a little defeatist when it comes to responding to mental illness – we just have to accept it’s a product of the way things are and there are severe limitations on what can ever be achieved until we evolve a fairer, kinder, more equal society. And even if these are normal human responses, they arise in certain circumstances, they’re not inevitable and they invariably involve terrible human distress. There is ice cream, and it would be useful to understand more about what the ingredients were (it might be useful if you want to try to move towards a kinder, safer society too). So, what underpins all these different diagnoses? What’s always there when people need help?
It's about self-esteem: the best evidence about what’s going for people who are suffering from mental health issues is what they say. And if you talk to someone who is in trouble, or listen to or read anything by people talking about their mental health problems, what they always say is that they don’t like themselves. It’s become easier to hear all the time as people have grown more prepared to talk about their mental health experiences. Have a look next time you’re reading an account of struggles with mental health: you’ll always find it.
It’s in the diagnostics too. One of the surprising features of DSM and ICD as classification systems for mental health is that they have very little to say about ideas in the minds of sufferers: they are very focussed on behavioural aspects and the mind is largely left as a black box. The only real exception, in both, is ideas of worthlessness and guilt which are identified as core symptoms of depression.
And it’s significant that this theme of disliking yourself has been identified with mental illness from the earliest days of people trying to understand the subject. The ancient Greeks had the idea of the Furies who pursued the guilty. Freud came up with the idea of a dynamic mechanism in our own minds which actively seeks to find fault and punish us. He called it the super ego, it’s now a commonplace in mental health and well-being, more often referred to as the inner bully or internal critic or imposter syndrome.
This capacity for the mind to turn upon itself is at the heart of theories of depression in psychoanalysis[36] and CBT[37]. And it’s recognition of the importance of negative self-thoughts in mental health that’s behind the drive to destigmatise mental illness (we don’t have to do that with pneumonia). That’s implicit too in themes of self-care, self-compassion, self-forgiveness and learning to love yourself.
And the idea of self-esteem as the engine of mental health is consistent with the social factors and experiences that Psychology has associated with mental illness. Being the wrong side of social inequality, injustice, discrimination, prejudice, neglect, abuse or violence is unlikely to make people feel good about themselves. Neither will relationship breakdown nor losing the people you depend on most for your emotional security through bereavement.
When things are going badly, people blame themselves, that self-persecuting impulse (inner bully, super ego, inferiority complex) kicks in, they turn on themselves, they can come to despise themselves, to loathe themselves. Naturally enough, things tend to get worse in a recession.[38] And there is quantitative data for this too: the kind of ‘scientific’ evidence people want to see, lots of it, backing up associations between self-esteem and depression[39] and other mental health conditions.[40]
But Psychology keeps dancing around the issue. However prevalent the messaging around destigmatisation, kindness and self-care (and what that really has to tell you about the significance of self-esteem in the context of mental health), what Psychology has to say expressly on the subject is invariably mealy-mouthed, wishy-washy and empty: ‘Low self-esteem isn't a mental health problem in itself. But mental health and self-esteem can be closely linked. Some of the signs of low self-esteem can be signs of a mental health problem. This is especially if they last for a long time or affect your daily life’.[41] What?!
Yet, actually, everyone else does get it: self-esteem has become what people intuitively mean by mental health. ‘Mental health’ has come to be shorthand for ideas about the self. When people talk about taking care of their mental health, they don’t generally think of themselves as talking about illness, and it’s not just feeling sad or unhappy; it’s a particular kind of unhappiness. Invariably, they mean things that make them feel bad about themselves, and the confusion and loneliness and sense of being lost that causes. Psychology somehow just can’t bring itself to spell that out and say this is key, this is an ingredient, across the board, this is where the hurt is, this is where the damage is done. If it did, it might be in a position to do a much better job of responding to what is the most important issue in mental healthcare.
It's about the contents of your own head becoming too intense: Psychology has had an obsession for years with the idea of identifying and counting personality traits. The theorized total has varied between 27 and four, and the topic is a little out of fashion now, but the legacy has been the “Big Five” personality traits: extroversion, agreeableness, openness, conscientiousness, and neuroticism.
Neuroticism is associated with emotional intensity and instability: its hallmarks are anxiety, self-consciousness, negative feelings about the self, complaint, irritability and anger. You don’t have to accept the idea of a trait of neuroticism at all to recognise the value of the evidence this seam of research found, again and again, that people who inhabit these states of mind more are significantly more likely to develop mental illness.[42] That’s right across the piece - mood disorders, substance abuse, somatic disorders, eating disorders. And they’re more likely to suffer from a range of serious physical conditions too. I know – who would have guessed? Unhappy people get ill. But it’s an important part of the picture and it needs stating because, amazingly, it’s not a very prominent feature of our approach to mental health.
You can see the process in play whenever people do get unhappy: the contents of their own heads get louder. There’s a sliding scale, which features at different points across the spectrum of mental health conditions: rumination to intrusive thoughts, rigid thinking, catastrophising, obsessive ideas, compulsive rituals. People’s ability to make realistic assessments of situations becomes compromised by the pressure of their own over-insistent ideas: distortions and delusions (everyone has them) magnify and proliferate. In the end, if it goes far enough - in full-blown psychosis - the contents of people’s heads become so urgent, so insistent, they can’t tell what’s real and what’s them anymore. The boundary between what’s inside them and what’s outside breaks down altogether: their own thoughts get externalised. The walls start talking and, because the underlying mechanism is self-loathing, it’s not nice things they’re saying.[43]
The same process underlies anhedonia: loss of the ability to enjoy things you used to (something primarily associated with depression but experienced across mental illness). People become too distracted, too pre-occupied with their own thoughts to be able to participate in what’s happening around them. They can’t get beyond the intensity of what’s happening in their head, or the need to numb it – in that sense, they have become too self-centred (and not in a fun way).
There’s good practical evidence of what’s going on for people who are suffering from mental health issues to be found in the tools they use to get away from it. Maybe the most popular is mindfulness: millions of app downloads, work-based courses and a place on NICE recommended clinical treatments. What the many different forms of mindfulness have in common is a deliberate tuning of attention in relation to your own thought processes.
Sessions generally start by re-directing awareness to the physical senses in an effort to rebalance between what’s going on inside people and what’s around them. The main objective of mindfulness is usually detached observation of your own thought processes, trying to teach people to step outside their own internal experiences, so that they are able to notice them and live with them without being overwhelmed. That’s about trying to dial down the intensity of what’s happening inside people’s heads. There’s no explicit attempt to ‘cure’ anything or to change the ideas people have (though often the idea that these techniques will lead people towards a more stable emotional life is around). It is solely about managing a symptom and its popularity shows just how much harm and distress it causes people when things get too loud inside your head.
If mindfulness is the most popular tool for managing mental health, the most popular clinical treatment, by a long way, is medication. More than three times as many people receive medication as therapy[44] and one in six adults use antidepressants in a year[45] (and significant numbers of people take psychiatric medication other than antidepressants). Clearly a lot of people find something helpful about antidepressants, but nobody understands why: nobody knows how antidepressants have the effect they do.
One explanation research has come up with is that (in common with other types of psychiatric medication) antidepressants disrupt the natural processes of the brain to produce a kind of sedative or dampening down effect.[46] Certainly ‘blunting’ (a reduction in positive as well as negative emotion) is a well-known side effect.[47] It’s estimated that around half the people who take pills experience blunting[48] but demand for medication continues to escalate.[49]
That suggests that, even at the price of numbness, what people who are ill want most is respite from the excruciating intensity of their own internal experience. Like mindfulness, medication doesn’t set out to cure anyone, it doesn’t change anything to do with the underlying issues, but just turning down the motor racing inside their heads can allow millions of people to feel better (and maybe to become better).
It's to do with other people: If the contents of your own head get too loud, they drown out what’s going on around you and the most significant element of what’s outside people’s heads is other people.
The Grant Study at Harvard is the most comprehensive longitudinal study ever attempted. It has tracked more than 250 Harvard graduates over the course of their lifetime with regular clinical assessment, psychometric tests, physical examination and interviews with close family members. Over the course of 80 years, the single most consistent finding has been that the most reliable predictor of mental (and physical) health and well-being has been the quality of subjects’ relationships.[50] That’s pretty straightforward evidence: if you can’t get on with the people around you, you’re more likely to get ill.
There’s evidential value to be derived too from the theories formulated by people who have spent their careers working in mental health. Many of the most influential theories about what goes wrong with mental health revolve around relationships with other people. The idea of narcissism is one of the best known (everybody ‘gets’ that one). Attachment theory focuses on the formation of bonds with other people as the blueprint for ‘normal’ development and mental health.[51] Social rank theory holds that depression derives from feelings of defeat and entrapment as a result of experiencing yourself as ‘lower rank’ than others.[52]
Difficulties with what psychologists call ‘Theory of Mind’ (the ability to form realistic pictures of what’s going on for other people – their desires, beliefs, intentions and emotions – as well as your own) have been associated with many different mental health issues[53] and the idea has found a clinical application in Mentalization Based Therapy.[54] Autism has always been a controversial classification and it’s now associated with such a wide description of behaviours and states of mind that it’s become very difficult to define. But according to the National Autistic Society (which is determined to resist generalised or stereotypical descriptions of the condition), the most consistent characteristic seems to’ be difficulty 'reading' other people - recognising or understanding others' feelings and intentions’.[55]
None of this should be surprising because, really, the whole of Psychology is about you and other people. And, when you examine them, the different approaches and treatments employed in mental healthcare fall neatly one side or the other of a line that can be drawn between the self and others. So, one of the benefits of group work, and support forums, most regularly cited by users is the insight they offer into other people’s predicaments (and the realisation that they are very often the same as yours). Users say it’s seeing other people more clearly, and a sense of shared experience, that helps them.
And developing a better sense of other people is at the heart of the approach of, ostensibly, one of the most influential figures in clinical Psychology. When you take the psychoanalytic Gormenghast overtones out of the picture, what Melanie Klein was saying is that problems arise from people’s difficulties in managing their own aggression, anxiety and self-dislike stemming from their inability to deal with other people’s separateness and independence. The way out, she said, is to be able to come to terms with separateness and independence, to recognise our own capacity to cause damage to other people, to be able to sustain remorse and find consolation in the idea of making good, doing better. Those are ideas which underpin several of the essays in this series.
But most elements of mental healthcare put the focus very much on the self side of the equation. CBT, for example, is about examining your patterns, your beliefs, your thought processes largely in isolation from anything that’s going on for other people. And the spiritual opposite of Melanie Klein is Carl Rogers’ person centred therapy: a highly individualistic approach, with the pathway to health built around discovering a robust enough sense of self through the experience of your own subjective experience being recognised, understood and validated.
Klein and Rogers epitomise a struggle in clinical Psychology (a battle for its soul really), between adaptation to an external reality (which largely consists of other people) and validation of an internal subjectivity. Rogers’ approach has won. Not because of the overwhelming evidence in its favour. Quite the opposite: when you can’t produce convincing evidence that what you have works, you’ve got to give people what they want. Freud was sexy (literally), Rogers is soothing and feels self-affirming and empowering – without convincing evidence, telling people their problems derive from being too self-centred just isn’t going to wash in the same way.
Mental healthcare, at a level that transcends different disciplines, qualifications, and theoretical frameworks, has adopted the highly subjective, individualistic themes of person-centred therapy. Whatever the specific treatment people actually receive, those are the values that flavour how people encounter mental healthcare and have increasingly come to determine how people generally think about mental health.
That hasn’t happened because there’s a consistent, compelling evidence base to support it. You will find evidence to support pretty much any aspect of this approach but then again you can find evidence to support pretty much any aspect of mental healthcare you want. It has happened because in the absence of a coherent evidence base, it’s a free-for-all and what has emerged is a tacit consensus, between clinicians, users and commentators, that this is what they want mental healthcare to look like.
But the most basic piece of evidence of all about mental healthcare is that the way we are dealing with mental health isn’t working well at all. The statistics mental healthcare professionals keep quoting all the time seem to be telling us that the problem just keeps getting worse all the time (and there is surely only so much blame you can heap on social media and the smart phone?). And, at the moment, we seem to be unable to find any convincing evidence that any of the different forms of treatment we use might be more effective than the others, in a way that could lead us forward on a clearer path. That can feel pretty hopeless. It ought to – imagine if that was the picture with cancer. If it doesn’t, that tells you something peculiar is going on with mental healthcare.
But there might be hope of things improving, and quite soon: evidence might save us yet and from an unlikely source.
∞
One of the other consequences of Psychology’s problems in proving its effectiveness in mental healthcare is that it’s very difficult for it to make a compelling case for investment. That’s been a factor in the levels of medication - at best only around one in four who encounter therapy on the NHS recover and, traditionally, therapy has been a great deal more expensive than pills. So, more than three times as many people receive medication on the NHS as therapy[56] and different categories of psychoactive medication are being prescribed to around 15 million people in England (some of whom might be receiving more than one category).[57] But what we are seeing now is the advent of online and automated therapy and that can make talking cheaper than pills.
The NHS app now allows users to self-refer for NHS Talking Therapy Services. That’s a big change in terms of access to mental healthcare but for most, at least in the first instance, they’re not going to find themselves sitting in a room with a psychotherapist. They’re likely to be offered guided self-help through an interactive website or an online programme. That’s just the beginning, AI is coming: the NHS has been using chatbots in referrals for mental healthcare for a while and is already trialling delivery of therapy via chatbots.[58] And there have been millions of downloads of the dozens of different meditation, mindfulness and well-being apps available in the private sector.
There is an implicit admission of failure in Psychology’s emphasis on individuality, difference, divergence and in the shift towards a social injustice model of mental health. The idea of ‘cure’ is being abandoned: this is from a Lancet editorial late last year titled ‘Recognising the right to be different’: ‘The aim of reducing the effect of mental illness generally is certainly laudable, but society must also accept that not all individuals with mental disorders can achieve mental health and that wellbeing should encompass living with mental illness. Recognising and accepting difference is the core of ending stigma and discrimination’.[59]
Follow this logic and you only need a handful of clinicians managing automated resources to manage the ‘symptoms’ of social distress and divergence as cost effectively as possible (after all people aren't going to get ‘better’ and you could use this money in cancer wards or heart clinics or education where it could make a difference). In theory, what could be left to Psychology would be contributing ideas to what society might look like in order to optimize mental health: what it is that people need in order to be well. But you’re definitely going to need some solid evidence on that.
And this is likely to be the shape of things, at least in the short term: a continuing heavily medicated backstop to mental healthcare, with increasing reliance on the promotion of lifestyle and digital resources, including AI, for symptom management. Ideas of ‘normal’ ‘better’, ‘well’ will continue to recede and with them ideas of active therapy as an agent for change or development. It is (and will become more) about providing each individual with readily accessible tools to manage their personal mental health profile (which will, in fact, be the same anxiety and depression related symptoms for everybody) and medicating them when they can’t.
But over time, there is the potential for more ambitious goals to emerge naturally, automatically. As technology moves more into the frame it will enable real-time data, feedback matching different interventions against outcomes, on a massive scale. Completely unprecedented levels of data will become available, together with completely objective, automated processes for analysing it. You will be able to see what works, where, when. There is the prospect of a complete, bottom-up, re-boot of the subject. Psychology would become genuinely evidence based and it’s the best kind of evidence: just looking at what happens and counting. Gradually, like monkeys typing Shakespeare, a truly effective model of mental healthcare could emerge.
There is valid evidence about mental health in Psychology and all around us but we haven’t been able to makes sense of it: there isn’t a subject we’ve done a worse job of studying. It’s just too close to home, we’re too invested: these are versions of us at our worst we’re talking about. Here, even more than the rest of life, we want things to be the way we want them to be. With this subject, we want to be right too badly, to be right.
Taking the human experts – and their desperation to ‘know’ - out of the equation may be the only way we can find our way to a functional system for mental healthcare. There’s a great deal of concern about the uses AI will be put to. Probably all justified but it also has the potential to stop us hiding from ourselves so effectively.
[1] https://www.psychologytoday.com/gb/basics/replication-crisis
[2] https://www.science.org/doi/10.1126/science.aac4716
[3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3665892/
[4] https://patient.info/doctor/patient-health-questionnaire-phq-9 and https://patient.info/doctor/generalised-anxiety-disorder-assessment-gad-7
[5] https://pubmed.ncbi.nlm.nih.gov/24780405/
[6] https://www.researchgate.net/publication/327581172_Mental_health_and_well-being_trends_among_children_and_young_people_in_the_UK_1995-2014_Analysis_of_repeated_cross-sectional_national_health_surveys
[7] https://www.bmj.com/content/380/bmj.p324
[8] The average consultation period for a GP is nine minutes - https://bmjopen.bmj.com/content/7/10/e017902. If you have someone presenting, persistently, with what looks like it might be a mental health condition, they’ll get passed through to specialist mental health services (or the waiting list) so the numbers of people being passed through to mental health services via GPs isn’t much help in bridging the gap between people believing they are ill and being ill.
[9] https://digital.nhs.uk/services/data-access-request-service-dars/dars-products-and-services/data-set-catalogue/mental-health-of-children-and-young-people-mhcyp
[10] In fact, on closer analysis it appears that the overall ‘at risk’ categories of young people hadn’t increased between the original series and the ‘wave’ follow-ups https://jonathancoppin.substack.com/p/are-mental-health-professionals-crooks
[11] https://pubmed.ncbi.nlm.nih.gov/31380664/
[12] https://www.theguardian.com/science/2023/jun/24/do-life-hacks-work-the-truth-is-well-never-know?
[13] https://pubmed.ncbi.nlm.nih.gov/34878945/
[14] https://www.frontiersin.org/articles/10.3389/fpsyt.2019.00407/full
[15] https://bjgpopen.org/content/5/4/BJGPO.2021.0020
[16] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9273730/
[17] https://pubmed.ncbi.nlm.nih.gov/17389902/
[18] https://www.nature.com/articles/s41380-022-01661-0
[19] https://link.springer.com/article/10.1007/bf02041242
[20] https://pubmed.ncbi.nlm.nih.gov/6194129/
[21] https://digitaleditions.telegraph.co.uk/data/1311/reader/reader.html?#!preferred/0/package/1311/pub/1311/page/25/article/NaN
[22] https://childmind.org/article/antidepressants-and-teen-suicides/
[23] ‘When, last month, I put it to Strack that these findings undermined his conclusions, he simply pointed to a “considerable number of studies in which the effect was demonstrated”, though it wasn’t clear which particular studies he had in mind, or why their findings should override those of the two recent mega-studies.’ https://www.theguardian.com/science/2023/jun/24/do-life-hacks-work-the-truth-is-well-never-know?
[24] https://journals.sagepub.com/doi/abs/10.1177/17456916211026947
[25] Ibid
[26] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2922387/
[27] https://www.psychiatrictimes.com/view/two-fallacies-invalidate-dsm-5-field-trials
[28] https://www.sciencedaily.com/releases/2019/07/190708131152.htm
[29] https://pubmed.ncbi.nlm.nih.gov/31813014/
[30] https://www.nature.com/articles/s41588-018-0090-3
[31] https://digital.nhs.uk/data-and-information/publications/statistical/adult-psychiatric-morbidity-survey/adult-psychiatric-morbidity-in-england-2007-results-of-a-household-survey
[32] https://pubmed.ncbi.nlm.nih.gov/15939839/
[33] https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2720421
[34] https://pubmed.ncbi.nlm.nih.gov/32315069/
[35] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7027356/
[36] https://www.free-ebooks.net/psychology-culture/Mourning-and-Melancholia/pdf?dl&preview
[37] https://psycnet.apa.org/record/1986-28931-001
[38] https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-016-2720-y
[39] https://www.frontiersin.org/articles/10.3389/fpsyt.2022.913067/full
[40] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5747942/
[41] https://www.mind.org.uk/information-support/types-of-mental-health-problems/self-esteem/about-self-esteem/
[42] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5428182/
[43] https://www.hra.nhs.uk/planning-and-improving-research/application-summaries/research-summaries/hearing-voices-in-psychosis-vip-how-do-they-change-over-time/
[44] https://www.mentalhealth.org.uk/explore-mental-health/statistics/people-seeking-help-diagnosed-mental-health-problems-statistics
[45] https://www.theguardian.com/society/2018/aug/10/four-million-people-in-england-are-long-term-users-of-antidepressants
[46] https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0030240
[47] https://www.verywellmind.com/can-antidepressants-make-you-feel-emotionally-numb-1067348
[48] https://www.cam.ac.uk/research/news/scientists-explain-emotional-blunting-caused-by-common-antidepressants
[49] https://www.nhsbsa.nhs.uk/statistical-collections/medicines-used-mental-health-england/medicines-used-mental-health-england-quarterly-summary-statistics-october-december-2021
[50] https://www.theatlantic.com/ideas/archive/2023/01/harvard-happiness-study-relationships/672753/
[51] http://labs.psychology.illinois.edu/~rcfraley/attachment.htm
[52] https://pubmed.ncbi.nlm.nih.gov/30594043/
[53] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8858823/
[54] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2816926/
[55] https://www.autism.org.uk/advice-and-guidance/what-is-autism
[56] https://www.mentalhealth.org.uk/explore-mental-health/statistics/people-seeking-help-diagnosed-mental-health-problems-statistics
[57] https://nhsbsa-opendata.s3-eu-west-2.amazonaws.com/mh-annual-narrative-final.html
[58] https://www.digitalhealth.net/2022/03/london-patients-ai-chatbot-mental-health-support/
[59] https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(22)00344-3/fulltext