Experimental psychopathology and psychological treatment
DOI:10.1093/acprof:oso/9780199228768.003.0023
Abstract and Keywords
This chapter describes changes and developments experimental psychopathology and psychological treatment during the past fifty years. Since the first emergence of psychology as an experimental science in the late 19th century and early 20th century, its application to psychological disorders had been a prominent goal. Two key developments during the 1950s were Hans Eysenck's review of the efficacy of psychological treatment and Joseph Wolpe's book on the use of systematic desensitization in psychotherapy.
Keywords: experimental psychopathology, psychological treatment, psychological disorders, Hans Eysenck, Joseph Wolpe, desensitization, psychotherapy
The behaviour therapy revolution
Since the first emergence of psychology as an experimental science in the late nineteenth and early twentieth century, its application to psychological disorders had been a prominent goal (McDougall 1944). The story is complex, and only some of its themes can be explored here, but it is convenient to take as a starting point two major developments that took place during the 1950s: the publication of Eysenck's (1952) review of the efficacy of psychological treatment and Wolpe's (1958) book on the use of systematic desensitization in psychotherapy.
Seen in context, these were revolutionary developments. Earlier in the century, psychological understanding and treatment of psychological problems had been dominated by psychoanalysis and its theory that neuroses arose from unconscious conflicts. This was despite the fact that a great deal of laboratory work had been done on fear conditioning and extinction, building on Pavlov's seminal work on classical conditioning and Skinner's work on operant conditioning. In particular, there had been a number of experimental demonstrations of Pavlovian fear conditioning in humans, for example the famous case of Little Albert, whose relationship with harmless white rats was severely disrupted by the pairing of exposure to such a rat with a loud and startling noise. The implications seemed clear enough: neurotic (fearful) behaviour is caused by conditioned reflexes. Neutral stimuli that just happen to be present at the time (the bell in Pavlov's original experiment, the white rat in Watson's experiment) become ‘conditioned stimuli’ through contiguity. Anxiety was seen as a learned response, occurring in reaction to signals that predict (i.e. have in the past been followed by) injury or pain (unconditioned stimuli). The resultant fear acts effectively as a motivator of behaviour: any behaviour that brings about a state of relief, safety, or security (reduction of fear) will be strengthened (Mowrer 1939). Yet, despite these advances, (p.284) most ‘application’ to clinical problems consisted of studies of verbal conditioning and the re-statement of dynamic principles and paradigms in learning theory terminology (Dollard and Miller 1950). Such mapping between an old and a new field had little impact on clinical practice.
So why the change in the 1950s? The first reason was the blistering attack on the prevailing psychodynamic orthodoxy, both theory and practice, by Eysenck (1952). He pointed out that there was no evidence that psychodynamic treatments produced outcomes that were any better than spontaneous remission. Eysenck pointed, by contrast, to early outcome studies of behaviour therapy that showed a large percentage of patients had sustained improvement—a larger proportion than those who would have improved spontaneously. He argued that, in as much as the dynamic formulations were testable, they were wrong—but mostly their formulations were untestable. Eysenck's attack on current clinical practice had far-reaching effects. It put evidence of efficacy as the central tenet of what the field of psychotherapy should be concerned with, an approach that now seems wholly justified in the light of the subsequent success of behavioural and cognitive therapies (Roth and Fonagy 2004).
The 1950s had thus already seen an increased interest in the use of behaviour therapy when, in 1958, Stanford University Press published Psychotherapy by Reciprocal Inhibition by Joseph Wolpe. Experiments on cats had shown that conditioned fear of certain locations could be gradually overcome by feeding them closer and closer to the original location (eating being incompatible with fear—hence ‘reciprocal inhibition’). Wolpe suggested that fear conditioning in humans could be similarly counter-conditioned through reciprocal inhibition and, furthermore, that such reciprocal inhibition was the main basis of psychotherapy's effects. In humans with fears and phobias, imagination of gradually more fearful phobic objects or situations under conditions of an incompatible state (deep muscle relaxation) would lead to large reduction in the symptoms of fear. Wolpe's results showed that the majority of patients responded very well to this approach, and the popularity of experimentally based behavioural work gained ground with dramatic speed.
By the end of the 1960s many empirical studies showed compelling evidence for the efficacy of behaviour therapy (Rachman 1971). An example is the successful treatment of the severe and intractable problem of obsessional-compulsive neurosis using exposure (Hodgson et al. 1972), an approach derived specifically from response prevention theories in animals. The dominance of behaviour therapy in the field of clinical psychiatry and psychology seemed complete.
(p.285) Meanwhile, however, it was becoming clear that not all the components that were thought essential for behaviour therapy to work were either necessary or sufficient. Rachman (1977) cited six factors that showed the conditioning theory of fear acquisition to be incomplete. (1) Many people fail to acquire fears in what are undoubtedly fear-evoking situations (e.g. air raids). (2) It is often difficult to produce conditioned fear reactions in humans, even under controlled laboratory conditions (Hallam and Rachman 1976). (3) Not everything is as easy to condition to an unpleasant unconditioned stimulus as everything else, so the equipotentiality premise is clearly untenable (Seligman and Hager 1972). Rachman was able to cite some old and some new experiments. First, Valentine (1946) had succeeded in producing a fear of a caterpillar in an 8-year-old child, but failed to make her fear a pair of opera glasses. More recently Ohman and colleagues (1975) had carried out fear conditioning experiments and had found it easy to establish pictures of snakes as conditioned stimuli, but not pictures of houses. Consistent with these laboratory findings (4) the distribution of fears in normal and neurotic populations is difficult to reconcile with the conditioning theory; epidemiological studies find more fear of spiders and snakes that fear of dentists in Western populations, for whom exposure to pain in the dental clinic was far more prevalent than pain associated with meeting spiders and snakes. (5) The clinical reports of phobic patients do not report large traumatic events at the origin of their phobia. (6) Fears can be reduced by vicarious and cognitive processes, so it seemed likely that they might be acquired by similar processes of modelling and by simple verbal instruction or warning.
Cognitive processes in anxiety disorders
This reference to cognitive influence in the origin and reduction of fear picked up a growing theme in the anxiety literature (Mathews 1990). Meichenbaum, Gilmore, and Fedoravicius (1971) had shown that speech anxiety could be reduced by getting participants to make explicit their ‘self-talk’, without exposing participants to the feared situation. Yet, despite the growing doubts about the sufficiency of older fear conditioning models, the clear demonstrations that a cognitive approach to anxiety might improve outcomes beyond those achieved by behaviour therapy did not happen until the 1980s when, influenced by Beck's cognitive approach to depression (see below), a number of young experimental psychopathology researchers began to look at hitherto intractable psychological fear responses and to use the cognitive approach in their analysis of what maintained the disorder and what, therefore, might alleviate it.
(p.286) Prominent among these was David M. Clark. He had noticed that one of his patients with panic disorder was hyperventilating, and hypothesized that such hyperventilation would, through its physiological effects, including a decrease in carbon dioxide in the blood, be responsible for some of the cardinal panic symptoms such as dizziness and faintness. Clark tried hyperventilating himself and noticed many of the symptoms that his panic patients reported. However, he did not have a panic attack. Why not? He reasoned that it was not the dizziness itself that was important but the catastrophic interpretations that patients made of these symptoms which produced the belief that something physically catastrophic was about to happen, such as a stroke or heart attack. This cognitive theory of panic disorders (Clark 1986) became a hugely influential model. It led to therapies in which patients were actively encouraged to reproduce the symptoms in the clinic as a way of testing the extent to which these indicated serious physical illness or merely the misinterpretation of innocuous body sensations.
However, Clark and his colleague, Paul Salkovskis, noticed that in some cases even patients' exposure to and then control of their own symptoms did not produce the desired change. For example, they had taught patients to control their breathing as an alternative to hyperventilation. Salkovskis (1991) suggested that such controlled breathing may, for some patients, represent a safety signal. That is, that patients believed (a) that something catastrophic was about to happen and (b) that the catastrophic event had been prevented by them controlling their breathing. The patients' dysfunctional beliefs about the meaning of the bodily symptoms was thus maintained: ‘If I had not sat down quietly and breathed slowly I would have had a heart attack’. The challenge for the cognitive therapist was to find ways in which people could identify the safety behaviours and then do precisely the opposite. So if ‘safety’ is sitting down quietly, they are encouraged to do vigorous exercise. This shift in therapy, to identify safety behaviours and challenge them, assisted the development of a number of effective treatments for a range of conditions: hypochondriasis (Warwick et al. 1996), social phobia (Clark and Wells 1995; see Butler and Hackmann 2004), post-traumatic stress disorder (PTSD) (Ehlers and Clark 2000; see also Foa et al. 1995), and generalized anxiety disorder (Borkovec and Newman 1999). Recent laboratory work has confirmed the causal status of cognitive biases (Hirsch and Mathews 2000; Mathews and MacLeod 2002) and the way in which cognitive processes have different impacts on psychopathology (Holmes and Hackmann 2004; Wegner et al. 1987) that are often transdiagnostic (Harvey et al. 2004). Cognitive treatments for anxiety have continued to make use of these laboratory findings (see Bennett-Levy et al. 2004; Craske and Barlowe 2001).
(p.287) Psychological models and treatment of depression
In the 1950s and 1960s behavioural psychologists had concerned themselves largely with anxiety-based disorders. They had seen depression as being outside their range of relevance because the cardinal features of the disorder were not behavioural, but seemed biological (eating disturbance, sleep disturbance, inability to concentrate, loss of energy, agitation, and retardation). The behavioural aspects (loss of interest in activities) seemed secondary to the biological syndrome and the other symptoms (guilt, suicidal ideation) were more ‘subjective’ and therefore thought less open to behavioural treatments. However, this did not prevent some psychologists, in the late 1960s and early 1970s, from formulating depression in behavioural terms: as a low rate of response-contingent positive reinforcement (Ferster 1973; Lewinsohn 1974; for a review see Williams 1992).
A. T. Beck had been writing about a cognitive formulation of depression for some time, but little attention was paid to his theory. It seemed obvious to many psychologists that negative thinking was a predominant feature of depression, but this was seen as a symptom of depression rather than part of the causal pathway. Intriguingly, it was probably not Beck's writings that aroused the interest in cognitive models of depression, but Seligman's ‘learned helplessness’ theory (Seligman 1975), derived from his work with animals. Seligman argued that the emotional, cognitive, and behavioural components of depression were due to the perception of response-reinforcement independence that would lead to an expectation of future helplessness—a cognitive theory.
Seligman's model provided a coherent, plausible, and testable account of the onset of depression, but the treatment implications of the model were very sketchy: giving patients experiences that would ‘reverse helplessness’. The result was that, in the mid-1970s, a cohort of experimental and clinical psychologists had a coherent cognitive theory, but no equivalent coherent therapy to accompany it. It is perhaps no surprise that the publication of Beck's successful cognitive therapy trial for clinical depression (Rush et al. 1977), followed soon after by the manual (Beck et al. 1979), should arouse immediate interest.
Beck proposed that the cognitive aspects of depression—negative thoughts about the self (‘I'm a failure’), the world (‘Everyone is against me’), and the future (‘Nothing will ever improve’)—were a critical aspect of both what caused depression in the first place, and then maintained the depression. Beck's approach was to encourage patients to ‘catch’ their negative thoughts and ideas, to write them down and treat them as hypotheses to be tested against (p.288) the evidence. He worked collaboratively with patients to set up behavioural experiments to test out the validity of the negative thoughts.
In parallel with these clinical developments, laboratory research was demonstrating the validity of the underlying theory, using new information-processing paradigms to demonstrate links between severity of depressed mood and memory bias (Teasdale et al. 1980), and showing that interrupting negative thoughts could indeed alleviate depressed mood (Teasdale and Fennell 1982). The 1970s saw a burgeoning of interest in laboratory studies on depression, and ended with the overwhelming conclusion that depressed mood could have an enormous effect on negative thinking, interpretations, and memory; that the negative thinking of depressed people was characterized by stable and global attributions for events (Abramson et al. 1978); and that self-referent negative thinking was the hallmark of depression and would need to be challenged if depression was to be treated psychologically (Beck et al. 1979). A number of randomized controlled trials (RCTs) substantiated the claim that challenging negative thoughts could alleviate depression and prevent recurrence over at least two years (for a review of early studies see Williams 1992), and subsequent trials have borne out this optimism (DeRubeis et al. 2005; Hollon et al. 2005).
Understanding and treating the process: the third wave of behavioural cognitive psychotherapy
Most cognitive theorists assumed that the important mediator of change in therapy was changing the patients' degree of belief in their negative thinking. For example, in panic disorder somebody might believe, ‘I'm about to have a heart attack’; in depression somebody might believe, ‘I'm not good enough’. But evidence began to emerge from the treatment of depression that would show that the theory on which it is based was, in some respects, insufficient.
We have seen already the remarkable success of cognitive therapy for depression. Many studies have found that it is as effective as antidepressant medication in reducing depression. More importantly, the effects of cognitive therapy last. By contrast, when people are taken off their antidepressant medication the depression tend to recur within one to two years. But this is a paradox because studies have found that the degree of belief in dysfunctional attitudes and negative thoughts changes equally during both pharmacological and psychological treatment. The successful mediator of change in cognitive therapy cannot therefore be the change in degree of belief. DeRubeis had argued that during cognitive therapy the patient learns a raft of compensatory strategies to use when mood starts to be disturbed. But what is it that these compensatory strategies do?
(p.289) Safran and Segal (1990) had suggested that the critical change process was ‘decentring’: the ability to observe one's thoughts and feelings as temporary, objective events in the mind, as opposed to reflections of the self and necessarily true. From this decentred perspective ‘the reality of the moment is not absolute, immutable, or unalterable’ (p. 117). Similarly, John Teasdale (1988) had argued that in depression an important process that maintained the disorder was ‘depression about depression’. The person interprets their mood as meaning that they are inadequate in some way. Notice how this represents a shift in perspective from the idea that it is degree of belief in thoughts that maintains a disorder to the way a person reacts to their own thoughts and feelings. If this was so, then either cognitive or behavioural treatments might be equally effective in bringing about this changed perspective (Dimidjian et al., 2006; Jacobsen and Martell 2001). The scene was set for the development of what has become known as the ‘third wave’ of behavioural/cognitive treatments that share the perspective (cf. Acceptance and Commitment Therapy— ACT) that change in the form or frequency of negative thoughts was not the mediator of change (see Hayes et al. 2005). For example, Teasdale, Segal and Williams (1995) had suggested that individual differences in vulnerability to depression would depend on differential reactivity to sad mood, differences in the tendency to react by adopting a ruminative and avoidant mode of mind (cf. Wegner et al. 1987). This mode of mind is toxic because it assumes thoughts to be a true reflection of reality, striving to close perceived discrepancies between actual and ideal representation of the self through the use of memory and by checking against future images of what might happen (Segal et al. 2002). So training depressed patients while in remission to recognize the activity of this mode of mind, and instead to disengage from it and switch to a more adaptive mindful mode, would reduce subsequent vulnerability to relapse. On the basis of both this theory and the clinical evidence from Kabat-Zinn's work using mindfulness training for patients with chronic pain, and Linehan's use of mindfulness as part of dialectical behaviour therapy for patients with a diagnosis of borderline personality disorder, they tested the impact of mindfulness-based cognitive therapy (MBCT) in two RCTs, showing that this approach halved the risk of relapse for patients with a history of three or more previous episodes (Ma and Teasdale 2004; Teasdale et al. 2000).
Concluding remarks
From the early days of behaviour therapy, theoretical and clinical developments have gone hand in hand. The maxim that theories not only have to be true, they also have to be useful, is no more borne out than in this field. We have identified three stages in the development of theory and practice: behavioural, cognitive, (p.290) and the recent third-wave approaches such as MBCT and ACT. Importantly, these latter approaches do not seek to replace behavioural and cognitive approaches, but to exist alongside them, testing out which strategies are most effective and what the essential moderators and mediators of change are. This combination of investigating underlying theory and improving practice in the light of both laboratory and clinical research is the hallmark of psychological treatment development over the last fifty years, and the hundreds of thousands of patients who have benefited over the years are a testimony to its success.
No comments:
Post a Comment