Cognitive behavioral therapy (CBT) is the leading evidence-based treatment for bulimia nervosa. A new “enhanced” version of the treatment appears to be more potent and has the added advantage of being suitable for all eating disorders, including anorexia nervosa and eating disorder not otherwise specified. This article reviews the evidence supporting CBT in the treatment of eating disorders and provides an account of the “transdiagnostic” theory that underpins the enhanced form of the treatment. It ends with an outline of the treatment's main strategies and procedures.
Keywords: Cognitive behavioral therapy, Eating disorders, Anorexia nervosa, Bulimia nervosaThe eating disorders provide one of the strongest indications for cognitive behavioral therapy (CBT). Two considerations support this claim. First, the core psychopathology of eating disorders, the overevaluation of shape and weight, is cognitive in nature. Second, it is widely accepted that CBT is the treatment of choice for bulimia nervosa 1 and there is evidence that it is as effective with cases of “eating disorder not otherwise specified” (eating disorder NOS), 2 the most common eating disorder diagnosis. This article starts with a description of the clinical features of eating disorders and then reviews the evidence supporting cognitive behavioral treatment. Next, the cognitive behavioral account of eating disorders is presented and, last, the new “transdiagnostic” form of CBT is described.
Eating disorders are characterized by a severe and persistent disturbance in eating behavior that causes psychosocial and, sometimes, physical impairment. The DSM-IV classification scheme for eating disorders recognizes 2 specific diagnoses, anorexia nervosa (AN) and bulimia nervosa (BN), and a residual category termed eating disorder NOS. 3
The diagnosis of anorexia nervosa is made in the presence of the following features:
The overevaluation of shape and weight; that is, judging self-worth largely, or even exclusively, in terms of shape and weight. This has been described in various ways and is often expressed as strong desire to be thin combined with an intense fear of weight gain and fatness.
The active maintenance of an unduly low body weight. This is commonly defined as maintaining a body weight less than 85% of that expected or a body mass index (BMI; weight kg/height m 2 or weight lb/[height in] 2 × 703) of 17.5 or less.
Amenorrhea, in postpubertal females not taking an oral contraceptive.The unduly low weight is pursued in a variety of ways with strict dieting and excessive exercise being particularly prominent. A subgroup also engages in episodes of binge eating and/or “purging” through self-induced vomiting or laxative misuse.
For a diagnosis of bulimia nervosa 3 features need to be present:
Overevaluation of shape and weight, as in anorexia nervosa.Recurrent binge eating. A “binge” is an episode of eating during which an objectively large amount of food is eaten for the circumstances and there is an accompanying sense of loss of control.
Extreme weight-control behavior, such as recurrent self-induced vomiting, regular laxative misuse, or marked dietary restriction.
In addition, the diagnostic criteria for anorexia nervosa should not be met. This “trumping rule” ensures that patients do not receive both diagnoses at one time.
There are no positive criteria for the diagnosis of eating disorder NOS. Instead, this diagnosis is reserved for eating disorders of clinical severity that do not meet the diagnostic criteria of AN or BN. Eating disorder NOS is the most common eating disorder encountered in clinical settings constituting about half of adult outpatient eating-disordered samples, with patients with bulimia nervosa constituting about a third, and the rest being cases of anorexia nervosa. 4 In inpatient settings the great majority of cases are either underweight forms of eating disorder NOS or anorexia nervosa. 5
In addition, DSM-IV recognizes “binge eating disorder” (BED) as a provisional diagnosis in need of further study. The criteria for BED are recurrent episodes of binge eating in the absence of extreme weight-control behavior. It is proposed that BED be recognized as a specific eating disorder in DSM-V. 6
Anorexia nervosa, bulimia nervosa, and most cases of eating disorder NOS share a core psychopathology: the overevaluation of the importance of shape and weight and their control. Whereas most people judge themselves on the basis of their perceived performance in a variety of domains of life (such as the quality of their relationships, their work performance, their sporting prowess), for people with eating disorders self-worth is dependent largely, or even exclusively, on their shape and weight and their ability to control them. This psychopathology is peculiar to the eating disorders (and to body dysmorphic disorder).
In anorexia nervosa, patients become underweight largely as a result of persistent and severe restriction of both the amount and the type of food that they eat. In addition to strict dietary rules, some patients engage in a driven form of exercising, which further contributes to their low body weight. Patients with anorexia nervosa typically value the sense of control that they derive from undereating. Some practice self-induced vomiting, laxative and/or diuretic misuse, especially (but not exclusively) those who experience episodes of loss of control over eating. The amount of food eaten during these “binges” is often not objectively large; hence, they are described as “subjective binges.” Many other psychopathological features tend to be present, some as a result of the semistarvation. These include depressed and labile mood, anxiety features, irritability, impaired concentration, loss of libido, heightened obsessionality and sometimes frank obsessional features, and social withdrawal. There are also a multitude of physical features, most of which are secondary to being underweight. These include poor sleep, sensitivity to the cold, heightened fullness, and decreased energy.
Patients with bulimia nervosa resemble those with anorexia nervosa both in terms of their eating habits and methods of weight control. The main feature distinguishing these 2 groups is that in patients with bulimia nervosa attempts to restrict food intake are regularly disrupted by episodes of (objective) binge eating. These episodes are often followed by compensatory self-induced vomiting or laxative misuse, although there is also a subgroup of patients who do not purge (nonpurging bulimia nervosa). As a result of the combination of undereating and overeating the weight of most patients with bulimia nervosa tends to be unremarkable and is within the healthy range, BMI = 20–25. Features of depression and anxiety are prominent in these patients. Certain of these patients engage in self-harm and/or substance and alcohol misuse and may attract the diagnosis of borderline personality disorder. Most have few physical complaints, although electrolyte disturbance may occur in those who vomit or take laxatives or diuretics frequently.
The clinical features of patients with eating disorder NOS closely resemble those seen in anorexia nervosa and bulimia nervosa and are of comparable duration and severity. 7 Within this diagnostic grouping 3 subgroups may be distinguished, although there are no sharp boundaries among them. The first group consists of cases that closely resemble anorexia nervosa or bulimia nervosa but just fail to meet the threshold set by the diagnostic criteria (eg, binge eating may not be frequent enough to meet criteria for BN or weight may be just above the threshold in AN); the second and largest subgroup comprises cases in which the features of AN and BN occur in different combinations from that seen in the prototypic disorders—these states may be best viewed as “mixed” in character—and the third subgroup comprises those with binge-eating disorder. Most patients with binge-eating disorder are overweight (BMI = 25–30) or meet criteria for obesity (BMI ≥ 30).
Consistent with the current way of classifying eating disorders, the research on their treatment has focused on the particular disorders in isolation. Wilson and colleagues 8 have provided a narrative review of the studies of the treatment of the 2 specific eating disorders as well as eating disorder NOS, and an authoritative meta-analysis has been conducted by the UK National Institute for Health and Clinical Excellence (NICE). 1 This systematic review is particularly rigorous and, as with all NICE reviews, it forms the basis for evidence-based guidelines for clinical management.
The conclusion from the NICE review, and 2 other recent systematic reviews, 9,10 is that cognitive behavioral therapy (CBT-BN) is the clear leading treatment for bulimia nervosa in adults. However, this is not to imply that CBT-BN is a panacea, as the original version of the treatment resulted in only fewer than half of the patients who completed treatment making a full and lasting recovery. 8 The new “enhanced” version of the treatment (CBT-E) appears to be more effective. 2
Interpersonal psychotherapy (IPT) is a potential evidence-based alternative to CBT-BN in patients with bulimia nervosa and it involves a similar amount of therapeutic contact, but there have been fewer studies of it. 11,12 IPT takes 8 to 12 months longer than CBT-BN to achieve a comparable effect. Antidepressant medication (eg, fluoxetine at a dose of 60 mg daily) has also been found to have a beneficial effect on binge eating in bulimia nervosa but not as great as that obtained with CBT-BN and the long-term effects remain largely untested. 13 Combining CBT-BN with antidepressant medication does not appear to offer any clear advantage over CBT-BN alone. 13 The treatment of adolescents with bulimia nervosa has received relatively little research attention to date.
There has been much less research on the treatment of anorexia nervosa. Most of the studies suffer from small sample sizes and some from high rates of attrition. As a result, there is little evidence to support any psychological treatment, at least in adults. In adolescents the research has focused mainly on family therapy, with the result that the status of CBT in younger patients is unclear.
Preliminary findings have been reported from a 3-site study of the use of the enhanced form of CBT (CBT-E) to treat outpatients with anorexia nervosa. 14 This is the largest study of the treatment of anorexia nervosa to date. In brief, it appears that the treatment can be used to treat about 60% of outpatients with the disorder (BMI 15.0 to 17.5) and that in these patients about 60% have a good outcome. Interestingly and importantly the relapse rate appears low.
There is a growing body of research on the treatment of binge-eating disorder. This research has been the subject of a recent narrative review 15 and several systematic reviews. 1,16,17 The strongest support is for a form of CBT similar to that used to treat BN (CBT-BED). This treatment has been found to have a sustained and marked effect on binge eating, but it has little effect on body weight, which is typically raised in these patients. Arguably the leading first-line treatment is a form of guided cognitive behavioral self-help as it is relatively simple to administer and reasonably effective. 18
Until recently, there had been almost no research on the treatment of forms of eating disorder NOS other than binge-eating disorder despite their severity and prevalence. 7 However, recently the first randomized controlled trial of the enhanced form of CBT found that CBT-E was as effective for patients with eating disorder NOS (who were not significantly underweight; BMI >17.5) as it was for patients with bulimia nervosa with two-thirds of those who completed treatment having a good outcome. 2
In summary, CBT is the treatment of choice for bulimia nervosa and for binge-eating disorder with the best results being obtained with the new “enhanced” form of the treatment. Recent research provides support for the use of this treatment with patients with eating disorder NOS and those with anorexia nervosa.
The remainder of this article provides a description of this transdiagnostic form of CBT.
Although the DSM-IV classification of eating disorders encourages the view that they are distinct conditions, each requiring their own form of treatment, there are reasons to question this view. Indeed, it has recently been pointed out that what is most striking about the eating disorders is not what distinguishes them but how much they have in common. 19 As noted earlier, they share many clinical features, including the characteristic core psychopathology of eating disorders: the overevaluation of the importance of shape and weight. In addition, longitudinal studies indicate that most patients migrate among diagnoses over time. 20 This temporal movement among diagnostic categories, together with the shared psychopathology, has led to the proposal that there may be limited utility in distinguishing among the disorders 19 and furthermore that common “transdiagnostic” mechanisms may be involved in their maintenance.
The transdiagnostic cognitive behavioral account of the eating disorders 19 extends the original theory of bulimia nervosa 21 to all eating disorders. According to this theory, the overevaluation of shape and weight and their control is central to the maintenance of all eating disorders. Most of the other clinical features can be understood as resulting directly from this psychopathology. It results in dietary restraint and restriction; preoccupation with thoughts about food and eating, weight and shape; the repeated checking of body shape and weight or its avoidance; and the engaging in extreme methods of weight control. The one feature that is not a direct expression of the core psychopathology is binge eating. This occurs in all cases of bulimia nervosa, many cases of eating disorder NOS, and some cases of anorexia nervosa. The cognitive behavioral account proposes that such episodes are largely the result of attempts to adhere to multiple extreme, and highly specific, dietary rules. The repeated breaking of these rules is almost inevitable and patients tend to react negatively to such dietary slips, generally viewing them as evidence of their poor self-control. They typically respond by temporarily abandoning their efforts to restrict their eating with binge eating being the result. This in turn maintains the core psychopathology by intensifying patients' concerns about their ability to control their eating, shape, and weight. It also encourages more dietary restraint, thereby increasing the risk of further binge eating.
Three further processes may also maintain binge eating. First, difficulties in the patient's life and associated mood changes make it difficult to maintain dietary restraint. Second, as binge eating temporarily alleviates negative mood states and distracts patients from their difficulties, it can become a way of coping with such problems. Third, in patients who engage in compensatory purging, the mistaken belief in the effectiveness of vomiting and laxative misuse as a means of weight control results in a major deterrent against binge eating being removed.
In patients who are underweight, the physiological and psychological consequences may also contribute to the maintenance of the eating disorder. For example, delayed gastric emptying leads to feelings of fullness even after patients have eaten only modest amounts of food. In addition, the social withdrawal and loss of previous interests prevent patients from being exposed to experiences that might diminish the importance they place on shape and weight.
The composite “transdiagnostic” formulation is shown in Fig. 1 . This illustrates the core processes that are hypothesized to maintain the full range of eating disorders. When applied to individual patients, its precise form will depend on the psychopathology present. In some patients, most of the processes are in operation (for example, in cases of anorexia nervosa binge-purge subtype) but in others only a few are active (for example, in binge-eating disorder). Thus, for each patient the formulation is driven by their individual psychopathology rather than their DSM diagnosis. As such, the formulation provides a guide to those processes that need to be addressed in treatment.