If required to bet that these ideas about the mechanisms linking trauma to hallucinations will turn out to be correct, I would not offer more than a very small portion of my salary as a stake. However, they at least offer plenty of opportunities for further research.
Dialogue with a Sceptical Voice
Experience tells me that, for many of my colleagues, the account I have just given of environmental contributions to psychosis is likely to be seen as very controversial. (When I recently gave a talk about my research on paranoid patients’ perceptions of their parents, a much respected psychologist – ironically, a researcher who studies expressed emotion – became very heated and said that she thought my ideas were ‘dangerous’.) Indeed, the suggestion that environmental influences could be important has been so effectively censored over the last few decades that these kinds of effects are scarcely mentioned in most textbooks of psychiatry or clinical psychology.
It seems a good idea, therefore, to anticipate some of the objections that might be made to the ideas I have introduced in this chapter. For example, an advocate of an exclusively neo Kraepelinian approach to psychiatry might argue that the biological findings we have considered in the previous chapter are somehow more fundamental or tangible than the findings on environmental influences.
Perhaps the idea behind this objection is that biological variables, such as sequences of DNA, brain scans or chemical assays, are more easily defined, more objective phenomena than traumas or emotional maltreatment. However, when making this claim, our sceptical friend surely cannot doubt that traumas and frosty relationships really happen. Moreover, he seems to forget that biological observations usually require many inferential steps. Geneticists routinely manipulate their data, inflating observed concordance rates to estimate the number of participants in their studies who have not yet become psychotic but who will do so in the future. Homovanillic acid is measured in the cerebrospinal fluid instead of making more difficult measurements of dopamine in the brain. The areas of the brain that appear to be activated during a functional neuroimaging experiment depend on the choice of the ‘off-task’ used as a comparison condition. Neither biologist nor psychologist, it seems, has unfiltered access to the world.
Perhaps what my sceptical friend really means is that biological abnormalities create the necessary conditions for psychosis, whereas specific environmental factors merely increase the risk of symptoms. This assumption lies behind stress-vulnerability theories of schizophrenia such as Paul Meehl’s schizotypy model or the model of relapse proposed by the UCLA research group. My reply is that this is an assumption, rather than a fact. However strong the overall influence of heredity on psychosis, no one has yet proven that everyone who becomes psychotic must carry particular genes. When we look at other biological correlates of psychosis, in all cases we find people who go mad but who do not show the relevant characteristics. Some schizophrenia patients do not have enlarged cerebral ventricles, some do not perform poorly on neurocognitive tests, and so on.
A deeper problem with my sceptical friend’s attempt to prioritize biological causes over environmental factors is that it assumes that these two types of influences are independent of each other and can be easily distinguished. However, in Chapter 7, I pointed to evidence that environmental influences literally shape the brain. For example, I cited evidence that some brain structures change in volume following the experience of trauma or sexual abuse. We have also seen that disruption to attachment relationships can have long-term effects on the chemistry of the brain. As neuroscience progresses, it seems likely that many more of these kinds of relationships between behavioural and biological variables will be uncovered by researchers. Perhaps eventually the distinction between what is psychological and what is biological will cease to be important. Reviewing the evidence for these kinds of relationships that has been gathered so far, University of Auckland psychologist John Read has argued that many of the biological abnormalities apparent in psychotic patients are similar to those demonstrated in the victims of trauma.69 These abnormalities therefore might be seen as evidence supporting the hypothesis that adverse experiences contribute to psychosis, rather than as evidence that is inconsistent with it.
A related argument that my sceptical friend might make is that environmental factors are simply not strong or specific enough to be considered major determinants of psychosis. Dealing with the strength issue first, it is now possible to calculate a single value, known as an effect size, which can be used to compare the magnitudes of different kinds of influences. The larger the effect size, the greater the magnitude of the effect being measured, with effect sizes over 0.50 usually being considered quite large. Walter Heinrichs of York University in Toronto, Canada, has recently calculated effect sizes for comparisons between schizophrenia patients and ordinary people on a wide range of neurocognitive and biological measures.70 For example, for the Continuous Performance Test (see Chapter 8), the average effect size was 0.67, whereas, for the backward masking test (also described in Chapter 8), it was 1.27. For levels of homovanillic acid (a metabolite of dopamine) in the cerebrospinal fluid of medication-free patients, the effect size was not only small but in the wrong direction (-0.11). For the density of dopamine D2 receptors, the average effect size was 0.93 but the effect sizes obtained in individual experiments varied between large in the expected direction (2.44) and moderately large in the wrong direction (-0.57). For structural neuroimaging studies comparing the size of the prefrontal brain in schizophrenia patients and ordinary people, the average effect size was 0.33. From the review of the literature on sexual abuse in psychotic women carried out by Goodman and her colleagues, I have calculated effect sizes for the influence of childhood sexual abuse varying between 0.70 and 2.04, depending on various assumptions about the rate of abuse in the general population. These values compare well with those found for more widely recognized factors influencing psychosis.
The question of whether the environmental effects I have identified are specific enough to be counted as major determinants of psychosis cannot be answered quite so simply. It is true, for example, that a very sizeable minority of the population is insecurely attached whereas, my sceptical friend would surely point out, only about 1 per cent of the population receives treatment for a psychotic illness. However, it is worth recalling that more people have psychotic experiences than receive psychiatric treatment. For this reason, we might really be trying to explain the experiences of 10 per cent of the population, rather than 1 per cent. Moreover, given the difficulty in defining effects such as family relationships and trauma, it is possible that any apparent lack of specificity reflects the limitations of our current measures. (Perhaps attachment relationships have to be disrupted in a specific, as yet undefined way. Although this counter-argument may sound weak, it is worth remembering that it has taken forty years to progress from George Brown’s original studies to our current knowledge about expressed emotion.) We must also consider the importance of interactions between different risk factors for psychosis. These interactions may be between different environmental factors. (For example, someone who is securely attached might cope reasonably well with a violent or sexual assault, whereas someone who has a dismissing-avoidant attachment style, and who has grown to maturity in a stressful and threatening environment, might respond by developing paranoid ideas.) There may also be interactions between environmental and biological variables, for example between attachment style and genes, or between trauma and early brain damage. (In the last chapter, I described how gene–environment correlations can inflate the impact of environmental influences on psychosis.)
The Origins of Psychosis
This brings me nearly to the end of my outline of an alternative to Kraepelin’s paradigm. The findings we have considered in the last three parts of this book are summarized in Table 18.2, which draws together what is known about the developmental pathways and psychological processes that lead to specific kinds of madness. Of course, many pieces of these par
allel and overlapping jigsaws remain missing. Nonetheless, in the case of each type of complaint we can see a tentative outline of the steps that lead from the cradle to the psychiatric clinic.
For the last thirty years, neoKraepelinian psychiatrists, like their predecessors in the first era of biological psychiatry, have promoted the idea that madness is a brain disease. In an earlier chapter, I described this paradigm as a cultural system, because it reflects the starting assumptions made by researchers, rather than conclusions drawn from scientific evidence. Recent neurodevelopmental models of schizophrenia and bipolar disorder are the latest manifestation of this cultural system. In making this claim, I am not attempting to deny that there is value in the neurodevelopmental approach. However, the grip of the Kraepelinian paradigm on the minds of researchers has been so great that many have ignored the evidence on psychological or social influences, denied that they are important, or treated them as peripheral to the problem of explaining madness, by diagnosing traumatized psychotic patients as suffering from an atypical form of PTSD.71
Despite the enormous efforts made by biological researchers, the effects that have been detected (for example, on the influence of foetal exposure to the influenza virus or on obstetric complications) have usually been small in magnitude, and the findings have often been inconsistent. Moreover, the link between these factors and the positive symptoms appears to be especially tenuous. This observation parallels the results obtained from the psychological studies I reviewed in Chapter 7, which have found only modest relationships between the severity of cognitive deficits and negative symptoms, and almost no relationship at all between deficits and hallucinations and delusions.
Table 18.2 Probable pathways to different kinds of madness.
In contrast, although much less effort has been made to unravel the psychological and social origins of psychosis, the evidence that is available is consistent and points to effects that are at least as strong as those measured by biological investigators. Moreover, this evidence suggests that a socially stressful environment, attachment and other family difficulties, and trauma, all have an impact on positive symptoms, especially persecutory delusions and hallucinations.
It seems that we are presented with something of a paradox. Although there can be no doubt that early brain damage confers vulnerability to madness, it seems to have very little impact on those symptoms which, since Schneider, have been considered the hallmark of psychosis. Perhaps Bleuler72 and Michael Foster Green73 are right in concluding that muted cognitive and emotional signs lie at the core of madness, and that the more flamboyant symptoms that have perplexed modern researchers are merely the individual’s reaction to these subtle deficits. Or perhaps cognitive deficits impede the individual’s ability to cope in times of crisis, creating challenges that the individual’s social reasoning skills are unable to meet.
Why adolescence is a high-risk period
With this possibility in mind, let us now return briefly to the question of why psychotic symptoms so often appear in adolescence and early adulthood. NeoKraepelinians would have us believe that this age trend can be explained entirely in terms of biological processes unwinding within the person – the biological time-bomb theory. But, of course, this high-risk period is a part of the life cycle that places special demands on the developing person’s social and cognitive skills. The American psychologist G. Stanley Hall, author of the first considered psychological theory of adolescence, famously borrowed the German phrase Sturm und Drang (storm and stress) to characterize it.74 This view of the transition to adulthood as a period of emotional turmoil has since been assimilated by popular culture, so that most people, at least in Western countries, accept it as self-evident. If someone calls you an adolescent you are unlikely to feel flattered.
Unfortunately, adolescence is perhaps the least understood developmental stage. There are not even agreed conventions for defining where it begins and ends. (Hall’s view was that it extended from puberty to when full adult status was attained, which he took to be around the age of 25.) Nor is it clear that it really is a time of storm and stress, at least for the majority of young people. It has been estimated that only 15 per cent of adolescents report severe emotional disturbance.75
Nonetheless, the time between puberty and settled adulthood clearly involves a number of developmental tasks, each of which may be seen as a hurdle at which the weak or ill-prepared may fall. In the ten years after puberty, the emerging adult must remodel her relationship with her parents, begin to explore sexual and emotional relationships with potential partners, and decide on a career path. Overarching these tasks, there is the need to establish an identity, a process that the developmental psychologist and psychoanalyst Eric Erikson portrayed as the central problem of adolescence in his famous book Childhood and Society.76
British psychologists Chris Harrop and Peter Trower note that these challenges seem to provoke in adolescents traits that are reminiscent of characteristics often seen in psychotic patients: marked shifts of mood coupled with equally dramatic shifts in self-esteem; self-consciousness, egocentricity and grandiosity; magical thinking and a preoccupation with powerful role models and the fable of one’s own life.77 In a study of normal adolescents carried out by Patrick McGorry and his colleagues in Melbourne, Australia, it was found that these kinds of characteristics are extremely difficult to distinguish from the typical prodromal symptoms of psychosis.78
It seems unlikely that the high risk of psychosis experienced by adolescents and young adults is unrelated to these experiences. As American psychologist Barbara Cornblatt has pointed out, children who suffer from the kinds of subtle social and cognitive difficulties that have sometimes been found in pre-psychotic children may be poorly equipped to face the developmental tasks of adolescence.79Although she based this hypothesis on American psychological research that has emphasized the role of cognitive deficits in psychosis (described in Chapter 7 of this book), her proposal seems even more plausible when what we know about social-cognitive processes is added into the equation. Children who reason abnormally about the causes of important events in their lives, or who are unable to guess accurately the thoughts of others, will be especially handicapped when faced with the demands of complex social situations. It is therefore quite unnecessary to assume that, in every case, disasters must trip up the child on his pathway to adulthood (although, as we have seen, this sometimes happens). The mere inability to adapt to new demands, to cross hurdles which brothers, sisters, friends and neighbours stride over with relative ease, may be sufficient to instil a sense of failure and personal inadequacy, magnify pre-existing social-cognitive peculiarities, and provoke thoughts that are increasingly psychotic.
Of course, the mere appearance of psychotic experiences is not enough to turn a troubled adolescent into a patient. For this to happen the experiences must be negatively valued and perceived as at odds with the individual’s culture. This suspicion may arise suddenly, or may develop over a long period of time. Often people wrestle with a variety of explanations of what is happening, sometimes entertaining several contradictory hypotheses at the same time (‘Maybe I’m special? Maybe I’m going insane? Maybe this is just a phase I’m going through?’). My colleague Tony Morrison has argued that the outcome of this process often depends partly on the person’s metacognitive belief system (for example, a person who has unrealistic views about the efficiency of the human mind may be especially troubled by the discovery that his own mind is playing tricks on him).80 Often, the decision that ‘something is seriously wrong’ is reached with the help or persuasion of other people: typically, a combination of parents, friends and diagnostic experts.
As I look back, I can remember many patients who, at the end of this process, have seemed stranded in the no man’s land between childhood and the adult world. Often very sensitive individuals, most have seemed aware, at least at some level, of the nature of their predicament. Sometimes, the pain of being stuck and forced to watch as peers have moved on to
get jobs, marry and buy cars and houses has been almost palpable. The task of helping young psychotic people to liberate themselves from this state of frozen development is one of the greatest challenges that a psychologist can face.
A Final Caveat
Over forty years ago, the philosopher Karl Popper pointed out that complex chains of interacting processes can rarely be reduced to simple causal laws:
Madness Explained Page 56