Cognitive Behaviour Therapy
Principles, Procedures and Evidence Base
The Birth and Early Development of Behavior Therapy
Some early examples of what later became called behaviour therapy were seen in the USA in the 1920s and 1930s. John Watson, the father of behaviourism, published a famous article in 1920 describing how fear was installed and later generalized in a healthy young male baby, little Albert. However, less well known is a paper by one of Watson’s graduate students, Mary Cover Jones (1924), describing the treatment of a three-year-old boy with a phobia of white rats, rabbits, furs and feathers. She used direct conditioning and modelling principles with a successful outcome. Another early example of applying principles of learning psychology is the paper on enuresis from 1938 where Hobart Mowrer and his wife presented the ‘bell-and-pad’ method based on classical learning principles, in what very well may be the first randomized controlled trial within psychiatry as a whole.
In the 1940s a South African psychiatrist, Joseph Wolpe, started to develop a treatment based on his theory of reciprocal inhibition. His reasoning was that a response innervated by the parasympathetic branch of the autonomic nervous system (e.g. relaxation), could be made to inhibit a response innervated by the sympathetic branch (e.g. fear or anxiety). In the 1950s he developed systematic desensitization (Wolpe 1958), which is still considered an evidence-based method for specific and social phobia (see also Chapter 12). Wolpe originally worked with real phobic stimuli, but since it was difficult to arrange confrontation with live stimuli he later chose to present the stimuli verbally to the patient who imagined being in the real situation.
During the same period developments were taking place in England and the USA. In London, Hans Eysenck, the first professor of psychology at the Institute of Psychiatry, heavily criticized psychoanalysis in a paper from 1952, concluding that it was no more effective than no treatment, i.e. the passage of time. In the 1950s and 1960s various people working at the Institute or collaborating with researchers there developed new treatment methods and methodologies for use in clinical research. Perhaps the most important of these was Monte Shapiro, who was head of the clinical teaching section at the psychology department. In that position he was responsible for the clinical training of a large number of early British behaviour therapists and was the first to develop a way to perform behaviour analysis of the individual patient. Shapiro’s focus on the experimental study of the individual patient also led him to develop the first single case design (A–B–A design) as early as the 1950s (Shapiro and Ravenette 1959), at least 10 years ahead of American developments in this methodology.
The English development of behaviour therapy was very much fuelled by a critique of the dominant psychological treatment at that time, psychoanalysis and psychodynamic therapy. The critique was both at a theoretical level, arguing that the psychoanalytic theories were either proven wrong or not subjected to scientific scrutiny, and at a practical level, highlighting the weak effects of psychoanalysis and the lack of evaluation in most instances.
Jack Rachman, a South African psychologist, moved to London in 1959 and worked with Eysenck. Together they started the first scientific journal in behaviour therapy, Behaviour Research and Therapy in 1963, which to a large extent was run by Rachman even if Eysenck was the chief editor initially. In 1974 Rachman succeeded Shapiro as head of the clinical section of the psychology department. He made important theoretical contributions in several areas, but his early research has primarily focused on obsessive-compulsive disorder (OCD) where he performed experimental studies on exposure and response prevention (ERP) at an early stage. The ERP method itself was developed in 1966 by another English psychologist, Victor Meyer, at Middlesex Hospital in London. The same year the first randomized controlled trial (RCT) of psychological treatment for agoraphobia (Gelder and Marks 1966) came out of the Institute of Psychiatry.
In the United States there was one isolated event in 1954 when Ogden Lindsley (one of Skinner’s students) reported on the use of operant learning principles in the treatment of schizophrenic patients at a hospital in Waltham, Massachusetts, using the term behaviour therapy for the first time. This was, however, not a journal article and basically went unnoticed by clinicians within psychiatry. A decade later Theodore Ayllon and Nathan Azrin published a paper, and in 1968 a book, on the token economy, a ward programme for severely hospitalized schizophrenics at the Anna State Hospital in Illinois. They showed that a combination of social (praise) and tangible (tokens) reinforcers could activate these patients to a very large extent. From the mid-1960s a large number of RCTs started to come out of American universities, first with mainly analogue (student) samples and later with patients fulfilling diagnostic criteria. The RCTs initially focused on specific phobias, but then on social phobia, agoraphobia, generalized anxiety disorder, obsessive-compulsive disorder, depression and schizophrenia. In children the early research efforts were focused on externalizing disorders such as conduct disorder, oppositional defiant disorder and attention deficit hyperactivity disorder (ADHD). Research on anxiety disorders in children did not start until the 1990s even if a few specific phobia studies were published earlier and the first RCTs on depression also came in the early 1990s.
The second time behaviour therapy was used in scientific writing was in 1958 when Arnold Lazarus, a South African psychologist, published an article in the South African Medical Journal describing ‘New methods in psychotherapy’. The following year Eysenck published the important article ‘Learning theory and behavior therapy’ in the Journal of Mental Science. In the early 1960s the concept behaviour therapy came to be used more generally by the proponents of this new form of treatment that was primarily based on learning psychology principles (both classical and operant conditioning) and later social learning as described by Albert Bandura (1969). This was also seen in the two scientific journals that were launched in the 1960s: Behaviour Research and Therapy in 1963 and Behavior Therapy in 1968, as well as a number of books on this new form of therapy.
The Birth and Early Development of Cognitive Therapy
Aaron T. Beck, an American psychiatrist who was initially trained in psychoanalysis, worked mainly with depression in the 1960s. He realized that psychoanalytic treatment was not very successful with depressed patients and started to collect clinical data on the way depressed patients characteristically think about themselves and the future. Based on these data he developed a cognitive theory of depression and a treatment that focused on changing negative thoughts and associated behaviours. His first book on depression was published in 1967 and his first on cognitive therapy in 1976. The first RCT on cognitive therapy in depression (Rush et al. 1977) showed that cognitive therapy was better than the antidepressant imipramine, a finding that aroused both interest and critique. The treatment manual of cognitive therapy for depression was published in 1979 and it has been used in numerous RCTs since then. Thus, the first diagnosis for which cognitive therapy was shown to work was depression.
Beck and others then turned their interest to anxiety disorders and a book on this theme was published by Beck, Emery and Greenberg in 1985. One of the European researchers, who after a behavioural training, became interested in cognitive therapy is the second author of this chapter, David Clark at Oxford University. After spending a semester of training at Beck’s Institute of Cognitive Therapy in Philadelphia, Clark started to use cognitive therapy in his work with panic disorder patients and in 1986 he published a well-known article on the theoretical model of cognitive therapy in panic disorder. His first RCT on panic disorder was published in 1994, showing that cognitive therapy was more effective than imipramine and applied relaxation. Clark then turned his interest to social phobia, health anxiety and post-traumatic stress disorder (PTSD).
The third diagnosis for which cognitive therapy was applied is eating disorders, bulimia nervosa in particular. This work was initially carried out by the English psychiatrist Christopher Fairburn at Oxford University in the 1980s. In the early 1990s the English psychologists Nick Tarrier and Gillian Haddock at Manchester University and the psychiatrists David Kingdom and Douglas Turkington started to explore how cognitive therapy could be used with psychotic patients, something that American researchers in the area initially discouraged. However, the last 20 years of research in this field shows that it is possible to apply cognitive interventions for psychotic patients, sometimes with excellent results even if the mean effect sizes reported in meta-analyses are not that large.
CBT: The Combination of Two Therapies
There were probably a number of factors that together helped the combination of cognitive and behaviour therapy (CBT) gain acceptance. From a behaviour therapy perspective, the cognitive component in the emotional reaction had, since Peter Lang’s articles and chapters in the 1960–70s, been considered as important as the behavioural and the physiological components. Development of treatment methods mainly focused on the behavioural (e.g. exposure in vivo and social skills training) and physiological components (e.g. progressive relaxation and biofeedback). However, as early as the 1970s there were attempts to develop more cognitive techniques as illustrated in the books by Mahoney (1974) and Meichenbaum (1977). Thus, there were a number of behaviour therapists who greeted cognitive therapy with open arms, realizing that it brought them a number of new interventions that could be used in the treatment of their patients.
From the cognitive perspective, Beck realized in his work with depressed patients that the severely depressed could not be treated with purely cognitive interventions to start with. They needed behavioural interventions to get activated so that they responded to the cognitive therapy. Another factor often mentioned by Beck was that when he wanted to present his treatment at scientific congresses the only place he was welcomed was at the Association of the Advancement of Behavior Therapy, from 1973 onwards. Thus, even if not all cognitive therapists wanted the combination, they did not seem to have a problem with the term CBT.
Another factor that may have contributed to this marriage of two therapies is the history of world congresses in this field. The first World Congress in Behaviour Therapy was held in Jerusalem in 1980, followed by Washington, DC in 1983, Edinburgh in 1988 and the Gold Coast (near Brisbane) in 1992. Regarding cognitive therapy the first world congress was in Umeå, Sweden in 1986, then in Oxford, 1989 and Toronto, 1992. The two congresses in 1992 were held only two weeks in between and only a few people went to both. This situation was unsatisfactory and a group of people started to work on the idea of a joint congress, among them the Danish psychologist Irene Oestrich and the authors of this chapter. We managed to persuade influential people to endorse the idea of a joint congress with two scientific committees: the behavioural led by Lars-Göran Öst and the cognitive by David Clark. In 1995 the World Congress of Behavioral and Cognitive Therapies was held in Copenhagen as the first joined congress. This was a success and since then all the CBT world congresses have been joined (with only one scientific committee though) (Acapulco, 1998; Vancouver, 2001; Kobe, 2004, Barcelona, 2007; and Boston, 2010).
Characteristic Features of Cognitive Behaviour Therapy
CBT is a broad form of psychotherapy with the following factors as the most characteristic features: