Bulimia

Bulimia is a psychiatric illness that is characterized by bouts of overeating, usually followed by self-induced vomiting or excessive use of laxatives. Most people suffering from bulimia are girls or women between the ages of 15 and 30. In some cases, the symptoms coexist with those of anorexia nervosa. Repeated vomiting can lead to dehydration and loss of potassium, causing weakness and cramps, and also causes tooth damage due to the gastric acid in vomit. Treatment of bulimia includes supervision and regulation of the person’s eating habits, psychotherapy, and, in some cases, antidepressant drugs.

Read more:

Bulimia nervosa - non-technical

Bulimia nervosa in detail - technical

Topics covered:

  • Introduction
    • Origins of the concept
    • Events since 1980
  • Diagnosis and classification
  • Clinical features
    • Specific psychopathology
    • Dieting and binge eating
    • Purging and other forms of weight control
    • Attitudes to shape and weight
    • General psychopathology
    • Physical features
  • Relationship to other disorders
    • Anorexia nervosa
    • Obesity
    • Depression
    • Anxiety disorders
    • Alcohol abuse
    • Personality disorders
    • Diabetes mellitus
  • Epidemiology
  • Aetiology
    • Development of the disorder
    • Predisposing factors and processes
    • The contribution of genetic factors
    • Maintaining factors and processes
  • Assessment
  • Treatment
    • Studies of pharmacological treatment
    • Antidepressant medication
    • Studies of psychological treatment
    • Cognitive-behavioural therapy Interpersonal psychotherapy
    • Exposure with response prevention
    • Management of bulimia nervosa
    • Step 1
    • Step 2
    • Step 3
    • Step 4
  • Course and outcome 
  • References

Introduction

Origins of the concept

The history of the diagnosis bulimia nervosa begins as recently as 1979. It was in this year that Russell published his now seminal paper ‘Bulimia nervosa: an ominous variant of anorexia nervosa' (1) in which he described 30 patients (28 women and two men), seen between 1972 and 1978, who had three major features in common. First, they had recurrent episodes of uncontrolled overeating; second, they regularly used self-induced vomiting or laxatives as means of weight control; and third, they had a morbid fear of becoming fat. Russell described many other features shared by these patients, including a history of anorexia nervosa (present in 80 per cent), the presence of severe depressive symptoms, and the fact that in most cases their body weight was in the healthy range. He noted that the disorder tended to run a chronic course and that it was ‘extremely difficult to treat'. Finally, he suggested that this clinical picture should be viewed as a syndrome, distinct from anorexia nervosa, and he proposed the term ‘bulimia nervosa'.

It is difficult to exaggerate the importance of Russell's paper. Its greatest contribution was perhaps its prescience in that it separated a subgroup of patients from among the range of eating disorders seen in clinical practice that was just starting to become more common, it identified its central features, and it gave it an appropriate name. In the late 1970s there was beginning to be interest in subdividing patients with anorexia nervosa into those who simply restrict their eating and those who also intermittently overeat, vomit, or take laxatives (see: anorexia nervosa). There had also been scattered reports of an anorexia-nervosa-like state in which overeating was the dominant feature, and terms such as ‘bulimarexia' and the ‘dietary chaos syndrome' had been suggested. None of these reports, however, provided the detailed clinical description or the clarity of conceptualization that were the hallmark of Russell's paper.

Events since 1980

Events gathered pace in the 1980s. In 1980 a syndrome termed ‘bulimia' was included in DSM-III. (2) This was intended to denote the type of patient that Russell had described, although its diagnostic criteria proved to be overly inclusive. In 1987, the criteria were refined and brought more in line with Russell's original concept. The syndrome was also renamed bulimia nervosa.(3) Also in the early 1980s evidence mounted that bulimia nervosa might be common and this led to a spate of studies of its prevalence. At the same time reports were published describing the successful treatment of these patients, the two most promising approaches being a specific form of cognitive-behavioural therapy and the use of antidepressant drugs. By the mid-1980s, both treatments had been tested in the first of what has become a large series of controlled trials.

Now, three decades later, the diagnosis bulimia nervosa is included in both DSM-IV (4) and ICD-10,(5) its prevalence is established, aspects of its aetiology are beginning to be understood, and much has been learned about its treatment.

Diagnosis and classification

Bulimia nervosa is one of the two main eating disorders recognized in DSM-IV and ICD-10. The other is anorexia nervosa. Three features must be present to make a diagnosis of bulimia nervosa.

  1. Recurrent episodes of ‘binge eating' (defined below).
  2. The regular use of extreme methods of weight control (e.g. highly restrictive dieting, self-induced vomiting, the misuse of laxatives and diuretics, or overexercising).
  3. A characteristic set of attitudes to shape and weight, at the heart of which is the judging of self-worth in terms of shape and weight. These attitudes are expressed as an intense dissatisfaction with shape and weight, a fear of weight gain and fatness, and, in many cases, a pursuit of weight loss and thinness. In addition to the three required diagnostic features, there is an exclusionary criterion. This is that the person must not meet diagnostic criteria for anorexia nervosa. In practice, this means that the patient must not fulfil the low-weight criterion for anorexia nervosa (a body mass index below 17.5).

Both DSM-IV and ICD-10 acknowledge the existence of eating disorders other than anorexia nervosa and bulimia nervosa. These may be described as ‘atypical eating disorders'. (6) They have been poorly characterized and little has been written about them. Many appear to be incomplete expressions of anorexia nervosa or bulimia nervosa (either in severity or form), but others are qualitatively different. In DSM-IV, all eating disorders other than anorexia nervosa and bulimia nervosa are placed in the single category ‘eating disorder not otherwise specified'. In ICD-10, various additional eating disorder categories are recognized (e.g. atypical anorexia nervosa, atypical bulimia nervosa, overeating associated with other psychological disturbances), although these concepts have never been adequately defined or differentiated. 

In DSM-IV, a provisional new eating disorder diagnosis was introduced which is termed binge eating disorder. It is placed in an appendix designed for possible future diagnostic concepts. Patients with binge eating disorder have recurrent episodes of binge eating, like those seen in bulimia nervosa, but they do not fulfil the diagnostic criteria for either anorexia nervosa or bulimia nervosa. This is a controversial category about which much remains to be learned. (7,8)

Clinical features

The great majority of patients with bulimia nervosa are female and most are in their twenties (although the age range is between 10 and 60 years). In considering the psychopathology of the disorder, a distinction may be drawn between its ‘specific' and ‘general' features. The former comprises features that are largely peculiar to eating disorders (e.g. self-induced vomiting), whereas the latter consists of features seen in other psychiatric conditions (e.g. depressive symptoms). The clinical features of bulimia nervosa are similar in men and women (9) and in those with and without a history of anorexia nervosa.

Specific psychopathology

Dieting and binge eating

The eating habits of patients with bulimia nervosa are characterized by strict dieting punctuated by repeated episodes of binge eating. The dieting is extreme, in that little tends to be eaten, and it is governed by multiple self-imposed dietary rules. These rules tend to be applied to all aspects of eating, including when to eat, what to eat, and how much to eat. As a result, the food eaten (when not binge eating) is restricted in quantity and range.

Recurrent episodes of ‘binge eating' interrupt this dieting. (The term binge eating denotes discrete episodes of eating that have two characteristics—first an unusually large amount of food is eaten, given the circumstances, and second there is a sense of loss of control at the time. (10)) The frequency and regularity of the binges varies. Some patients have episodes almost every day, whereas in others the episodes are intermittent. In DSM-IV, it is specified that the binges should occur on average at least twice a week, but this is an arbitrary figure that has little discriminatory value. Among those patients in whom the binge eating is frequent, the binges have few, if any, obvious triggers, although there may be circumstances under which binge eating is more likely (e.g. when alone at home). Among patients in whom the binge eating is less frequent, the binges often have clear precipitants. These tend to be of three overlapping types: first there is breaking a personal dietary rule (e.g. exceeding a daily calorie limit or eating a banned food), second there are situations which intensify concerns about shape and weight (e.g. receiving an adverse comment about appearance), and third there is the occurrence of negative moods (often as a result of interpersonal events). All three encourage the temporary abandonment of dietary control.

The amount of food eaten during binges varies, both from patient to patient and from episode to episode. Typical episodes involve the consumption of 4200 to 12 600kJ (1000–3000kcal). (11) The food eaten generally comprises items that are otherwise being avoided. Thus binges tend to be composed of energy-dense high-fat items such as chocolate, ice-cream, and pastries. Binges come to an end as a result of the combined influence of exhaustion, extreme fullness, a diminution of the drive to eat, and the running out of food supplies. In about three-quarters of patients they are immediately followed by measures designed to counteract the effects of the overeating, the most common being self-induced vomiting and the taking of laxatives or diuretics.

The binges are a source of considerable distress. They magnify patients' fears of weight gain and fatness, and they are a source of shame and self-disgust. For this reason most binges occur in private and are kept secret from others. It is the binge eating that eventually drives these people to seek help.

Purging and other forms of weight control

In DSM-IV bulimia nervosa is subdivided into two types, a purging and non-purging type. In the purging type there is self-induced vomiting or the misuse of laxatives or diuretics, or both, whereas in the non-purging type purging is either not present or it is infrequent. The majority of patients seen in clinical practice have the purging form of the disorder and it has been the focus of most research.

Self-induced vomiting is the most common form of purging. In most patients it only takes place after binge eating. It is generally achieved by stimulating the gag reflex, using the fingers or some other long object, although in more established cases it can be accomplished with no mechanical aid. The vomiting is repeated until patients think that they have retrieved all the food that they can. Patients become extremely distressed if they are unable to vomit after binge eating: indeed, if they foresee that they may not have the opportunity to vomit, they tend not to binge. A minority of patients also induce vomiting at other times, for example following smaller episodes of overeating (subjective episodes) or ordinary meals or snacks.

The misuse of laxatives or diuretics is somewhat less common than self-induced vomiting. It takes two forms: one is to compensate for specific episodes of binge eating, like self-induced vomiting, and the other is as a general method of weight control (like dieting), in which case it is not tied to particular episodes of overeating. The number of laxatives or diuretics taken varies considerably, sometimes far exceeding the recommended dose.

None of these methods of purging is an effective method of weight control. Self-induced vomiting results in the retrieval of only about half to two-thirds of what has been eaten, (12) the taking of laxatives has a minimal effect on food absorption, (13) and diuretic taking has none. As a result, a significant proportion of each binge is absorbed. The patient's weight is largely a product of her undereating and overeating, with the two tending to cancel each other out.

Other forms of weight control behaviour are practised by some patients, including overexercising, the spitting out of food, and the taking of repeated enemas or saunas. Overexercising is the most common of these, but it is not nearly as prominent or as obviously pathological as in anorexia nervosa. A minority of patients ruminate, that is repeatedly regurgitate and re-chew food that has been eaten. They may then either re-swallow the food or spit it out. This behaviour is not well understood.

Attitudes to shape and weight

A characteristic set of attitudes to shape and weight is the other distinctive element of the specific psychopathology of bulimia nervosa. Equivalent attitudes are found in anorexia nervosa, and their presence is required to make either diagnosis. These attitudes are often described as the ‘core psychopathology' of eating disorders. They are characterized by an overconcern with shape and weight in which there is a fear of weight gain and fatness that is generally accompanied by a pursuit of weight loss and thinness. Underlying this psychopathology is the tendency to judge self-worth largely, or even exclusively, in terms of shape and weight. Whereas it is usual to evaluate self-worth on the basis of perceived performance in a variety of domains (such as interpersonal relationships, work, sport, artistic ability, etc.), people with anorexia nervosa or bulimia nervosa evaluate themselves primarily in terms of their shape and weight. These attitudes and values constitute a good example of an overvalued idea.

Most features of bulimia nervosa can be understood as being secondary to these attitudes to shape and weight. The dieting, purging, and overexercising are obvious secondary features. In addition, there are direct behavioural expressions of these concerns. For example, many patients repeatedly weigh themselves and scrutinize their appearance in mirrors. Others avoid any knowledge of their weight while being acutely sensitive about their appearance. Some avoid others seeing their body and some even avoid seeing it themselves. This can have a major impact on social and sexual relationships. In the most extreme cases, there is a loathing of body shape (body image disparagement). Unlike in anorexia nervosa, misperception of body shape is rarely seen.

The concerns about shape and weight, and therefore eating, affect others in the patient's immediate environment. Meals are often times of tension and social events which involve eating may be avoided. The feeding of children may be affected, (14,15 and 16) and this can result in growth impairment (17).

General psychopathology

General psychiatric symptoms are prominent in bulimia nervosa, more so than in anorexia nervosa. (18,19) The nature of the comorbid symptoms also differs. Depressive features are particularly striking: indeed, the level of depressive symptoms in bulimia nervosa is equivalent to that seen in major depressive disorder. Depressed mood, feelings of hopelessness and worthlessness, poor concentration, guilt, and suicidal ideation are seen. Anxiety symptoms are also encountered, many of which are directly related to the eating disorder; for example, there is often pronounced anxiety about eating in public. Obsessive– compulsive features are sometimes present, although they are less common than in anorexia nervosa. Similarly, social functioning is less impaired.

A subgroup of patients with bulimia nervosa have ‘impulse-control' problems, such as the overconsumption of alcohol or drugs, or repeated self-harm. Some of these patients also meet diagnostic criteria for borderline personality disorder. The prevalence of such features varies according to treatment setting: they are unusual in community samples, (20) whereas they are more frequent among patients seen in specialist treatment centres. (21)

Much more common than frank personality disorders are two traits which are also seen in anorexia nervosa. (22) The first is low self-esteem. This generally antedates the eating disorder, although it is often exaggerated by it. Many patients with bulimia nervosa describe long-standing doubts about their worth and ability, irrespective of their accomplishments. The second is perfectionism, that is, imposing on oneself inordinately high personal standards in a range of domains (e.g. work, sport, personal conduct). Since many of these standards are unachievable, it is common for these patients to give long histories of having viewed themselves as perpetually failing.

Physical features

There are few physical abnormalities in bulimia nervosa. (23,24) Body weight is unremarkable in the majority of patients and thus features of starvation are rarely seen. Nevertheless, menstrual abnormalities or amenorrhoea are present in about a quarter of patients. These are likely to be secondary to the disturbed eating since they generally respond to the successful correction of the eating disorder. On laboratory testing endocrine abnormalities are sometimes encountered and these tend to be mild versions of those found in anorexia nervosa.

Frequent purging, and especially the combination of vomiting and laxative misuse, results in fluid and electrolyte abnormalities in some patients. These abnormalities are varied in nature but most often consist of some combination of hypokalaemia, hyponatraemia, and hypochloraemia. The patients appear to accommodate to these abnormalities since medically serious complications (e.g. cardiac arrhythmias) are much less common than might otherwise be expected given the laboratory findings. Some patients experience intermittent oedema, particularly if there is a sudden decrease in the extent of their purging.

Localized physical abnormalities include erosion of the dental enamel (especially from the lingual surfaces of the front teeth) among those who have vomited for many years, traumatic calluses on the knuckles of the hand of those who use their fingers to induce the gag reflex (Russell's sign), and enlargement of the salivary glands, especially the parotids, probably as a result of chronic inflammation. A small proportion of patients have raised serum amylase levels usually due to an increase in the salivary isoenzyme.

Relationship to other disorders

Anorexia nervosa

Bulimia nervosa has many features in common with anorexia nervosa particularly the characteristic attitudes to shape and weight, and the forms of behaviour that arise as a result. In most cases, bulimia nervosa is preceded either by frank anorexia nervosa (in about a quarter of cases) or an anorexia-nervosa-like state. Movement from bulimia nervosa to anorexia nervosa is unusual although it occurs, and some patients remain on the cusp between the two disorders.

There is some evidence of co-aggregation between bulimia nervosa and anorexia nervosa with increased rates of both disorders among the relatives of probands with either condition. (25) There is substantial overlap in the risk factors for anorexia nervosa and bulimia nervosa. (26,27)

Obesity

Few patients with bulimia nervosa are overweight or frankly obese. However, there is evidence of raised rates of parental and premorbid obesity. (27) Obesity is an unusual sequel of the disorder although this may be because those at most risk of obesity are less likely to recover (28) and so continue to suppress their weight.

Depression

Depressive symptoms are common in bulimia nervosa and they may antedate the eating disorder. (19,27) Most family studies have found a raised rate of affective disorder among the relatives of these patients but the balance of evidence suggests that the increase in risk is largely confined to the relatives of those probands who are themselves depressed. (25)

Anxiety disorders

Little is known about the relationship between bulimia nervosa and the various anxiety disorders. These disorders are common among patients with bulimia nervosa, and they may antedate the disorder.(19,29) There has been just one family genetic study of the relationship between an anxiety disorder and bulimia nervosa. It found that there was an increased risk of obsessive–compulsive disorder among the relatives of these patients, but this was confined to those probands who themselves had a history of the disorder. (30)

Alcohol abuse

There is a raised rate of alcohol and drug abuse among patients with bulimia nervosa. It rarely antedates the eating disorder but this is to be expected given the age of onset of substance abuse disorders. There is also a raised rate of substance abuse disorders among the relatives of these patients. Once again it seems that this is largely confined to those probands who themselves have a history of substance abuse. (30)

Personality disorders

It is hazardous making personality disorder diagnoses among those with eating disorders. This is because eating disorders have their onset in adolescence and they directly affect many of the characteristics upon which personality is judged. Thus there is a risk of overestimating the presence of personality disturbance. Nevertheless, some patients with bulimia nervosa do seem to have a coexisting personality disorder, the most common form being borderline personality disorder. (18) Little is known about the rate of personality disturbance among the relatives of these patients.

Diabetes mellitus

It was thought that eating disorders were over-represented among those with type I diabetes mellitus. This is probably not the case. Controlled studies in which eating disorder features have been assessed by interview rather than self-report questionnaire (the preferred method and one which minimizes the risk of false-positive ratings) have found no evidence of increased risk. (31,32)

Epidemiology

The fact that it took Russell (1) more than 6 years (1972–1978) to collect 30 cases of bulimia nervosa suggested that the disorder was not common. Therefore it is remarkable that within a few years of the publication of Russell's paper it was evident that bulimia nervosa was an important source of psychiatric morbidity among women.

In the early 1980s large numbers of previously undetected cases were identified using the media. (33,34) These were remarkably similar to Russell's cases, except that almost all were female and a small proportion had a history of anorexia nervosa. Most had kept their eating disorder secret for many years, and because of shame and hopelessness few had sought help. Many thought that they were the only person with this type of eating disorder. Simultaneously, however, and doubtless partly as a result of the media attention, there was also a sharp increase in the number of people requesting treatment for bulimia nervosa such that it was soon the most common type of eating disorder seen in clinical practice. This remains true today. (35,36) 

The marked increase in the number of patients with bulimia nervosa stimulated interest in the prevalence of the disorder. By 1989 over 50 prevalence studies had been conducted, many of which had yielded unrealistically high prevalence figures as a result of using weak assessment and sampling procedures. Gradually methods improved with the result that estimates of the prevalence of bulimia nervosa decreased to more modest and consistent levels with the point prevalence among young adult women (aged 16 to 35 years) being in the region of 1 per cent. (37,38) A similar figure has been obtained for lifetime prevalence. (39) The prevalence of bulimia nervosa among men is not known. Among patient samples, male cases are most unusual. Little is known about the presence of bulimia nervosa in non-Western cultures although it is thought to be uncommon.

There have been few estimates of the incidence of bulimia nervosa, and since these have been based upon clinic rather than community samples, they are likely to underestimate the true figures. Even today, many people with bulimia nervosa do not seek help. The lack of reliable community-based incidence figures also makes it impossible to know whether the disorder has become more common since the 1970s or whether the upsurge in cases in the early 1980s was as a result of undetected cases being more likely to seek treatment. Data from the assessment of women in different birth cohorts (40) suggest that the disorder has become much more common, although other explanations for the apparent increase cannot be ruled out. It is of note that systematic searches of the psychiatric literature prior to the mid-1970s have uncovered very few cases of the disorder. (41) 

Aetiology

Development of the disorder

Anorexia nervosa generally starts in mid-adolescence with a period of voluntary dietary restriction which proceeds to get out of control. As a result body weight falls and a state of starvation develops. Shape and weight concerns may predate the onset of the dieting or develop as weight is lost.

Bulimia nervosa starts in a similar way although the age of onset is typically some years later, and shape and weight concerns usually antedate the dieting. The dietary restriction resembles that seen in anorexia nervosa and it leads to weight loss sufficient to result in anorexia nervosa in about a quarter of cases. (As a result of referral bias, this proportion is higher in cases seen in specialist centres.) In the remaining cases there is also weight loss but it is less extreme. After a variable length of time (generally within 3 years) dietary control breaks down with the patient's dieting becoming punctuated by episodes of overeating. At first, the episodes of overeating may be modest in size and intermittent, but gradually they become larger and more frequent. As a result, the lost weight is regained and body weight returns to near its original level. By this point the disorder tends to be self-perpetuating. At some stage in this sequence of events, self-induced vomiting and laxative misuse may be adopted to compensate for the overeating. In practice, however, both forms of behaviour have the opposite effect since belief in their effectiveness encourages a relaxation of control over eating. In those who vomit this phenomenon is exaggerated by the discovery that the process is easier after eating large amounts of food.

Predisposing factors and processes

There are many risk factors for the development of bulimia nervosa (27,42) and these overlap with those for anorexia nervosa. (26) The risk factors may be usefully divided into a number of categories.

  1. Demographic factors: these are being female, adolescent, and living in a Western society.
  2. Exposure to an immediate social environment that encourages dieting: this includes being brought up in a family in which there is intense interest in shape, weight, and eating as a result of one or more members either having some degree of eating disorder or having a medical condition that affects eating or weight (such as diabetes mellitus). Extreme occupational or recreational pressures to diet also appear to be associated with increased risk (e.g. ballet dancing (43)), although there may also be an element of self-selection. Another important influence is parental and childhood obesity, the rates of which are substantially increased in bulimia nervosa. Both are likely to sensitize individuals to their appearance and weight, and thereby make them prone to diet. There is also some evidence that puberty occurs comparatively early which may also magnify concerns about shape.
  3. Exposure to factors that increase the risk of psychiatric disturbance in general and depression in particular: these include a family history of psychiatric disorder, especially depression, and a range of adverse childhood experiences including parenting deficits and sexual and physical abuse. (27,44,45) It was thought that sexual abuse was especially common among those who develop bulimia nervosa, but the balance of evidence suggests that the rate is no higher than that among those who develop other psychiatric disorders. (27,46) There is some evidence that those who report such a history have more general psychiatric symptoms than those who do not. (46)
  4. Perfectionism and low self-esteem: both traits are common antecedents of anorexia nervosa and bulimia nervosa. Typically these factors interact, resulting in feelings of incompetence and ineffectiveness.
  5. Family history of substance abuse: there is a raised rate of substance abuse in the families of patients with bulimia nervosa. (30) It is not clear how this increases the risk of bulimia nervosa. Clinical observations suggest that some of those who develop bulimia nervosa learn to modulate their mood by consuming large quantities of food, alcohol, or psychoactive drugs.

An important question is how those with anorexia nervosa are able to maintain strict control over their eating whereas this is not true of those with bulimia nervosa. The explanation is unclear but several processes may be of relevance. First, perfectionism is even more pronounced in anorexia nervosa (42) and it may enhance self-control. Second, the vulnerability to obesity found in bulimia nervosa may somehow undermine dietary restraint. Third, the mood lability of bulimia nervosa may also disrupt restraint.

The contribution of genetic factors

The magnitude of inherited influences is unclear. Several studies have found an increased rate of anorexia nervosa and bulimia nervosa among the relatives of probands with bulimia nervosa, (25) whereas other studies have found either no increase or only an increase in atypical eating disorders. (30) The findings of the first twin series of note suggested that the inherited contribution was minimal. (47) More recent studies from the Virginia Twin Registry have yielded inconsistent findings with estimates of heritability varying from negligible (48) to 55 per cent (40) to 83 per cent. (49) A number of factors are likely to have contributed to the variability in the findings, including the small sample sizes (which render twin models unstable) and the difficulty making lifetime diagnoses of bulimia nervosa. It is also of note that, in contrast with other psychiatric disorders, the findings suggest that there may be violations of the equal environments assumption that underpins twin studies. (48,49)

The nature of any inherited vulnerability is a matter of speculation. The liability appears not to be shared with that for other psychiatric disorders, (25) although there is evidence suggestive of shared familial transmission between anorexia nervosa and obsessive–compulsive personality disorder, (30) a construct which overlaps with that of perfectionism. It is possible that there is an inherited abnormality in the regulation of weight and eating habits in view of the raised rates of parental and childhood obesity. This may include susceptibility to adverse consequences of dieting. Dieting is known to affect 5-hydroxytryptamine neurotransmission particularly among women, (50) and there is evidence of abnormalities in brain 5-hydroxytryptamine function in women who have recovered from bulimia nervosa. For example, one study found that acutely lowering brain 5-hydroxytryptamine function resulted in the temporary reappearance of bulimic symptoms (51) and another found abnormal levels of cerebrospinal fluid 5-hydroxyindoleacetic acid. (52) The molecular genetic studies of eating disorders are focusing on the genes involved in 5-hydroxytryptamine neurotransmission but no consistent findings have emerged. (25)

Maintaining factors and processes

Once established, bulimia nervosa tends to run a chronic course with the proviso that it tends to improve over the long term (see below). There are a number of processes which account for its self-perpetuating character which are discussed in this article on anorexia nervosa. They include the ongoing influence of the extreme concerns about shape and weight; the form of these patient's dieting, which encourages binge eating; the moodmodulating effect of binge eating; and the fact that the loss of control over eating perpetuates fears of weight gain.

Assessment

The identification of patients with bulimia nervosa is not difficult so long as the diagnosis is considered. This is important because the shame that characterizes the disorder leads many people to delay seeking help—the average delay between onset and presentation is about 5 years—and when they do present for treatment, some do so indirectly. Thus they may complain of depression, substance abuse, menstrual disturbance, or gastrointestinal symptoms, rather than the eating disorder itself. The best policy is for psychiatrists always to keep in mind the possibility of bulimia nervosa when assessing female patients aged between 16 and 35years. Negative responses to the following two questions should suffice to exclude the disorder. 

  1. Do you have any problems with your eating?
  2. Do you have any problems controlling your eating, i.e. problems with binge eating?

Patients who present directly complain of having lost control over eating and their assessment is generally straightforward.

The best established measure of eating disorder features is the Eating Disorder Examination. (53) This interview is widely regarded as the ‘gold standard' measure of eating disorders, (54) and it may be used either for routine clinical assessment or for research purposes. It defines all key terms and generates operational eating disorder diagnoses. Various self-report questionnaires are available but they provide a more basic level of assessment and they cannot be used to make a clinical diagnosis. The leading self-report measures are the Eating Disorder Inventory (55) and the self-report version of the Eating Disorder Examination. (56)

No assessment is complete without weighing the patient. This needs to be done with considerable sensitivity because of these patients' concerns about their weight. A physical examination is not essential unless the patient is underweight (or there are other medical indications), nor are laboratory tests required except in those cases in which there is reason to suspect that there might be fluid or electrolyte disturbance.

Treatment

Given that bulimia nervosa has only recently been described, there has been a remarkable amount of research on its treatment. Over 50 randomized controlled trials have been completed. (57) Although almost all these studies have been ‘efficacy' rather than ‘effectiveness' studies, there are reasons to think that their findings are of direct relevance to routine patient care not least because the patients studied have been similar to those seen in clinical practice. (58) Nevertheless, there is a definite need for effectiveness research particularly now that main treatment options are clear.

Studies of pharmacological treatment

A variety of drugs have been tested as possible treatments for bulimia nervosa including antidepressants, appetite suppressants, anticonvulsants, and lithium. Only antidepressants have shown promise.

Antidepressant medication

All the major classes of antidepressant drug have been evaluated, including tricyclic antidepressants, monoamine oxidase inhibitors, selective serotonin uptake inhibitors, and atypical antidepressants. The findings have been relatively consistent and may be summarized as follows (adapted from Wilson and Fairburn (57)).

  1. Antidepressant drugs are more effective than placebo at reducing the frequency of binge eating and purging. On average, among treatment completers there is about a 60 per cent reduction in the frequency of binge eating and a cessation rate of about 20 per cent. The therapeutic effect is more rapid than that seen in depression. There is generally little change in the placebo group. The drop-out rate varies but averages about 30 per cent.
  2. The longer-term effects of antidepressant drugs remain largely untested. Almost all of the studies to date have been of their short-term use (16 weeks or less). The findings of the few longer-term studies suggest that outcome is poor whether or not patients remain on medication.
  3. Few studies have evaluated the effects of antidepressant drugs on features other than binge eating and purging. Mood improves as the frequency of binge eating declines but this effect is common to all treatments for bulimia nervosa. Antidepressant drugs do not appear to modify the patient's extreme dieting which may account for the apparently poor maintenance of change.
  4. The different antidepressant drugs seem to be equally effective, although there have been no direct comparisons of different drugs.
  5. With one exception, there have been no systematic dose–response studies. The exception showed that fluoxetine at a dose of 60mg/day, but not 20mg/day, was more effective than placebo.
  6. Patients who fail to respond to one antidepressant drug may respond to another. There have been no drug augmentation studies.
  7. No consistent predictors of response have been identified. Pretreatment levels of depression appear not to be related to outcome.
  8. The mechanism(s) whereby antidepressant drugs exert their ‘antibulimic' effects is not known. The apparent comparability of different classes of drug implicates a common mechanism but this is unlikely to be their antidepressant action since the response is too rapid and the level of depression does not predict outcome.

Studies of psychological treatment

Cognitive-behavioural therapy

The most intensively studied psychological treatment is a specific form of cognitive-behavioural therapy. This was the first promising treatment described (59) and it remains the leading treatment for the disorder. 

Cognitive-behavioural therapy is conducted on an outpatient basis and involves 15 to 20 sessions over about 5 months. It is suitable for all patients bar the small minority (less than 5 per cent) who require hospital admission.

The findings of the studies of cognitive-behavioural therapy (over 20 controlled trials) are summarized below (adapted from Wilson and Fairburn (57)).

  1. The drop-out rate with cognitive-behavioural therapy (about 15 per cent) is about half that seen with antidepressant drugs. The treatment is also more acceptable to these patients than treatment with medication.
  2. Cognitive-behavioural therapy has a substantial effect on the frequency of binge eating and purging. On average, among treatment completers there is about an 80 per cent reduction in the frequency of binge eating, and a cessation rate of about 60 per cent.
  3. The effects of cognitive-behavioural therapy appear to be well maintained. Most of the recent studies have included a 6 to 12 month follow-up period. The relapse rates are low.
  4. Cognitive-behavioural therapy affects most aspects of the psychopathology of bulimia nervosa including the binge eating, purging, dietary restraint, and the overevaluation of shape and weight. In common with other treatments, the level of depression decreases as the frequency of binge eating declines. Social functioning and self-esteem also improve.
  5. Cognitive-behavioural therapy is more effective than delayed treatment (that is, a waiting list control group), other psychological treatments (other than possibly interpersonal psychotherapy; see below), and antidepressant drugs at reducing the frequency of binge eating and purging. Among the other psychological treatments studied have been supportive psychotherapy, focal psychotherapy, supportive–expressive psychotherapy, interpersonal psychotherapy, hypnobehavioural treatment, stress management, nutritional counselling, behavioural versions of cognitive-behavioural therapy, and exposure with response prevention. 
  6. No consistent predictors of response to cognitive-behavioural therapy have been identified. Severity of symptoms at presentation, a history of anorexia nervosa, low self-esteem, and the presence of borderline personality disorder have been associated with worse outcome in some studies but not others. Initial response (over the first 4 weeks of treatment) is emerging as a useful predictor of subsequent outcome.
  7. The mechanism(s) of action of cognitive-behavioural therapy have yet to be established. It seems that the cognitive procedures are required for progress to be maintained since behavioural versions of the treatment are associated with a greater risk of relapse.
  8. There is evidence that the combination of cognitive-behavioural therapy and antidepressant drugs may be more effective than cognitive-behavioural therapy alone in reducing anxiety and depressive symptoms.
  9. Simpler forms of cognitive-behavioural therapy show promise. These include brief versions of the treatment designed for use in primary care and cognitive-behavioural self-help in which patients follow a cognitive-behavioural self-help programme either on their own (pure self-help) or with the guidance of a therapist (guided self-help). (60,61)
Interpersonal psychotherapy

Interpersonal psychotherapy is the leading alternative to cognitive-behavioural therapy. This treatment was originally devised by Klerman et al. (62) as a treatment for depression. It is a focal psychotherapy, the main emphasis of which is to help patients identify and modify current interpersonal problems. The treatment is both non-directive and non-interpretative and, as adapted for bulimia nervosa, (63) it pays little attention to the patient's eating disorder. It is therefore very different to cognitive-behavioural therapy.

In a comparison of cognitive-behavioural therapy, interpersonal psychotherapy and a behavioural version of cognitive-behavioural therapy, interpersonal psychotherapy was found to be the least effective of the three treatments at the end of treatment but the proportion of responders continued to increase such that it was as effective as cognitive-behavioural therapy by 8 and 12 month follow-up (64), and indeed, 5years later. (63) The marked difference between the two treatments in their procedures and temporal influence suggested that, although they were equivalent in their eventual effects, each treatment operated through specific mechanisms. A recent large-scale replication of the cognitive-behavioural therapy versus interpersonal psychotherapy comparison has confirmed these findings.

Exposure with response prevention

This is an adaptation of the behavioural treatment for obsessive– compulsive disorder in which patients with bulimia nervosa are presented with cues that generally precede binge eating or vomiting and are then prevented from engaging in their usual response. It was advocated in the 1980s both as a treatment in its own right and as an element of cognitive-behavioural therapy. It is difficult to administer for a number of reasons: patients find it aversive; it is procedurally complex; and the sessions are time-consuming. The studies of its use suggest that it conveys no benefit over standard cognitive-behavioural therapy. (57,65)

Management of bulimia nervosa

Given the large number of patients with bulimia nervosa and the scarcity of therapists with training in the treatment of eating disorders, a ‘stepped care' approach to management has been advocated. With such an approach a simple treatment is used first and only if this proves insufficient is a more complex and specialized intervention provided. Four steps may be distinguished.

Step 1

Having established the diagnosis, the first decision is whether the patient may be treated on an outpatient basis. The great majority (over 95 per cent of referrals to non-specialist centres) may be managed this way. Exceptions are patients whose level of depression is so severe that they cannot make use of psychological treatment, significant risk of suicide, and physical complications necessitating inpatient or day patient care. Severe substance abuse requires treatment in its own right, although this can sometimes be integrated with the treatment of the eating disorder. For example, it is possible to adapt cognitive-behavioural therapy for bulimia nervosa so that it addresses at the same time the patient's substance abuse.

Step 2

If the patient is suitable for outpatient treatment, guided cognitive-behavioural self-help is the next step. In this context, this involves following a cognitive-behavioural self-help programme under the guidance of a ‘facilitator' (a non-specialist therapist). Three cognitive-behavioural self-help books are available, (66,67 and 68) two of which are direct translations of cognitive-behavioural therapy for bulimia nervosa. There is evidence to support the use of all three books and it is a matter of preference which is used. All three provide information about bulimia nervosa together with a self-help programme. The role of the facilitator is not to provide treatment as such, as in a conventional ‘therapist-led' treatment, but rather to support and encourage the patient to follow the programme. Thus, this is a ‘programme-led' form of treatment, and it is this characteristic that makes it suitable for dissemination. The treatment generally takes about 4 months and involves 8 to 10 meetings with the facilitator, each lasting up to 30min. It is best if the first few appointments are weekly.

Step 2 may take place in a variety of settings including primary care. A third to a half of patients show substantial change, and their progress appears to be well maintained. No consistent predictors of outcome have been identified. Patients who obtain little benefit (and usually this is obvious within 4 to 6 weeks) need to move on to step 3, although it would not be inappropriate to recommend as an interim step a brief trial of an antidepressant drug with those patients who are willing to consider drug treatment (see step 3).

Research on the treatment of binge eating disorder suggests that there may be a subgroup of patients who will respond to ‘pure self-help', that is following a cognitive-behavioural self-help programme with no outside support. (69) In clinical situations in which there will be a delay before therapist/facilitator-led treatment can be provided, pure self-help may have a role. At the very least it provides patients with sound information about the disorder. No adverse effects of pure or guided self-help have been identified, and failure to respond does not seem to affect response to subsequent treatment.

Step 3

Patients who do not benefit from cognitive-behavioural self-help should receive full cognitive-behavioural therapy. In most cases this should be provided on its own, but it may be worthwhile adding an antidepressant drug under any one of the following three circumstances: if significant depressive symptoms are interfering with compliance, if depressive symptoms are persisting despite an improvement in eating habits, and if progress is limited. The drug of choice is fluoxetine (60mg). This can be started without intermediate lower doses and it should be taken in the morning.

Step 4

The fourth step is essentially pragmatic since there are few research findings of relevance. To guide the choice of treatment, reasons for non-response need to be carefully considered. Explanations include failure of cognitive-behavioural therapy, poorly administered cognitive-behavioural therapy, poor patient compliance (which itself needs to be explained), and the influence of outside events.

There are a number of different treatment options under these circumstances, the choice of which depends upon the outcome of the reassessment and the resources available. They include the following.

  • Stop treatment and arrange to re-evaluate the patient after an interval of some months. This can be justified on at least two grounds. First, bulimia nervosa has a general tendency to improve over time (see below). Spontaneous improvement may therefore occur. Second, patients and therapists can show ‘burn-out' after sustained periods of therapeutic work. A break can often be helpful. Deciding to stop treatment should be a joint decision and it is not appropriate with patients who are distressed or with those whose physical or psychological well being is a cause for concern.
  • Change antidepressant drug. As noted above, there is evidence that some patients respond to a change in antidepressant drug. If there is going to be a response to a second-line drug, it is likely to be rapid (within a few weeks). No guidelines are available to govern the choice of drug.
  • Embark upon a new psychological treatment. While one obvious choice is interpersonal psychotherapy, there are no grounds for supposing that patients who fail to respond to cognitive-behavioural therapy will respond to interpersonal psychotherapy. Indeed, there is emerging evidence that this is not the case. An alternative strategy is to change the form of cognitive-behavioural therapy. The re-evaluation of the patient may have resulted in the identification of problems that might be amenable to cognitive-behavioural procedures outside the realm of mainstream cognitive-behavioural therapy for bulimia nervosa. For example, those with extreme concerns about shape might benefit from more emphasis on body image (70) and those with markedly low self-esteem might respond to an approach which focuses on negative self-evaluation. (71)
  • Arrange for day patient or inpatient treatment. In a small minority of cases outpatient treatment proves not to be sufficient, either because the disorder is resistant to treatment or because the patient's life circumstances are interfering with progress. In such cases day patient or inpatient treatment can be useful. Generally this involves a combination of approaches including elements of cognitive-behavioural therapy. Both day patient treatment and inpatient treatment should be followed by treatment on an outpatient basis designed to ensure that progress is maintained.

Course and outcome

Much remains to be learned about the course and outcome of bulimia nervosa. It is clear from epidemiological studies that many people do not present for treatment. The course of their disorder is completely unknown. Those who do present tend to do so after a considerable period of time indicating that among this subgroup the disorder has a tendency to run a protracted course. Conversely, the findings of the treatment studies indicate that the outcome is considerably better than Russell originally suggested, although it must be stressed that even with cognitive-behavioural therapy, the most effective treatment, only about half the patients make a full recovery.

There have been few studies of long-term course or outcome. The studies with longer periods of follow-up have identified proportionately fewer cases at reassessment. (72) At 10-year follow-up, about 10 per cent meet diagnostic criteria for bulimia nervosa and a further 15 per cent have an atypical eating disorder. (73,74) There is no evidence that bulimia nervosa evolves into any other psychiatric disorder, and anorexia nervosa is a very unusual outcome. Body weight barely changes over time and, in contrast with anorexia nervosa, the mortality rate appears not to be raised. The disorder tends to improve during pregnancy but subsequent relapse is common. (75,76) No robust predictors of course or outcome have been identified.

Thus, if one takes a long-term perspective, there seems to be a trend towards recovery. Whether this is an inherent property of bulimia nervosa, whether it is age-related, or whether it reflects the influence of treatment is not known. 

References

1. Russell, G.F.M. (1979). Bulimia nervosa: an ominous variant of anorexia nervosa. Psychological Medicine, 9, 429–48.

2. American Psychiatric Association (1980). DSM-III: diagnostic and statistical manual of mental disorders. American Psychiatric Association, Washington, DC.

3. American Psychiatric Association (1987). DSM-IIIR. American Psychiatric Association, Washington, DC.

4. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th edn). American Psychiatric Association, Washington, DC.

5. World Health Organization (1992). International statistical classification of diseases and related health problems, 10th revision. WHO, Geneva

6. Fairburn, C.G. and Walsh, B.T. (1995). Atypical eating disorders. In Eating disorders and obesity: a comprehensive handbook (ed. K.D. Brownell and C.G. Fairburn), pp. 135–40. Guilford Press, New York.

7. Fairburn, C.G., Welch, S.L., and Hay, P.J. (1993). The classification of recurrent overeating: the ‘binge eating disorder' proposal. International Journal of Eating Disorders, 13, 155–9.

8. Castonguay, L.G., Eldredge, K.L., and Agras, W.S. (1995). Binge eating disorder: current state and future directions. Clinical Psychology Review, 15, 865–90.

9. Carlat, D.J., Camargo, C.A., Herzog, P.H., and Herzog, D.B. (1997). Eating disorders in males: a report on 135 patients. American Journal of Psychiatry, 154, 1127–32.

10. Fairburn, C.G. and Wilson, G.T. (1993). Binge eating: definition and classification. In Binge eating: nature, assessment and treatment (ed. C.G. Fairburn and G.T. Wilson), pp. 3–14. Guilford Press, New York.

11. Walsh, B.T. (1993). Binge eating in bulimia nervosa. In Binge eating: nature, assessment and treatment (ed. C.G. Fairburn and G.T. Wilson), pp. 37–49. Guilford Press, New York.

12. Kaye, W.H., Weltzin, T.E., Hsu, L.K.G., McConaha, C.W., and Bolton, B. (1993). Amount of calories retained after binge eating and vomiting. American Journal of Psychiatry, 150, 969–71.

13. Bo-Linn, G.W., Santa Ana, C.A., Morawski, S.G., and Fordtran, J.S. (1983). Purging and calorie absorption in bulimic patients and normal women. Annals of Internal Medicine, 99, 14–17.

14. Lacey, J.H. and Smith, G. (1987). Bulimia nervosa: the impact of pregnancy on mother and baby. British Journal of Psychiatry, 150, 777–81.

15. Stein, A., Woolley, H., Cooper, S., and Fairburn, C.G. (1994). An observational study of mothers with eating disorders and their infants. Journal of Child Psychology and Psychiatry, 35, 733–48.

16. Stein, A. and Fairburn, C.G. (1989). Children of mothers with bulimia nervosa. British Medical Journal, 299, 777–8.

17. Stein, A., Murray, L., Cooper, P.J., and Fairburn, C.G. (1996). Infant growth in the context of maternal eating disorders and maternal depression: a comparative study. Psychological Medicine, 26, 569–74.

18. Braun, D.L., Sunday, S.R., and Halmi, K.A. (1994). Psychiatric comorbidity in patients with eating disorders. Psychological Medicine, 24, 859–67.

19. Brewerton, T.D., Lydiard, R.B., Herzog, D.B., Brotman, A.W., O'Neil, P.M., and Ballenger, J.C. (1995). Comorbidity of Axis I psychiatic disorders in bulimia nervosa. Journal of Clinical Psychiatry, 56, 77–80.

20. Welch, S.L. and Fairburn, C.G. (1996). Impulsivity or comorbidity in bulimia nervosa. A controlled study of deliberate self-harm and alcohol and drug misuse in a community sample. British Journal of Psychiatry, 169, 451–8.

21. Lacey, J.H. (1993). Self-damaging and addictive behaviour in bulimia nervosa. A catchment area study. British Journal of Psychiatry, 163, 190–4.

22. Vitousek, K. and Manke, F. (1994). Personality variables and disorders in anorexia nervosa and bulimia nervosa. Journal of Abnormal Psychology, 103, 137–47.

23. Mitchell, J.E. (1995). Medical complications of bulimia nervosa. In Eating disorders and obesity: a comprehensive handbook (ed. K.D. Brownell and C.G. Fairburn), pp. 271–5. Guilford Press, New York.

24. Mitchell, J.E., Pomeroy, C., and Adson, D.E. (1997). Managing medical complications. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 383–93. Guilford Press, New York.

25. Strober, M., Lilenfeld, L.R., Kaye, W., and Bulik, C. (2000). Genetic factors in anorexia nervosa and bulimia nervosa. In Feeding problems and eating disorders (ed. P.J. Cooper and A. Stein). Harwood, Switzerland, in press.

26. Fairburn, C.G., Cooper, Z., Doll, H.A., and Welch, S.L. (1999). Risk factors for anorexia nervosa: three integrated case–control comparisons. Archives of General Psychiatry, 56, 468–76.

27. Fairburn, C.G., Welch, S.L., Doll, H.A., Davies, B.A., and O'Connor, M.E. (1997). Risk factors for bulimia nervosa: a community-based case–control study. Archives of General Psychiatry, 54, 509–17.

28. Fairburn, C.G., Norman, P.A., Welch, S.L., O'Connor, M.E., Doll, H.A., and Peveler, R.C. (1995). A prospective study of outcome in bulimia nervosa and the long-term effects of three psychological treatments. Archives of General Psychiatry, 52, 304–12.

29. Bulik, C.M., Sullivan, P.F., Fear, J.L., and Joyce, P.R. (1997). Eating disorders and antecedent anxiety disorders: a controlled study. Acta Psychiatrica Scandinavica, 96, 101–7.

30. Lilenfeld, L.R., Kaye, W.H., Greeno, C.G., et al. (1998). A controlled family study of anorexia nervosa and bulimia nervosa. Archives of General Psychiatry, 55, 603–10.

31. Fairburn, C.G., Peveler, R.C., Davies, B., Mann, J.I., and Mayou, R.A. (1991). Eating disorders in young adults with insulin dependent diabetes: a controlled study. British Medical Journal, 303, 17–20.

32. Peveler, R.C., Fairburn, C.G., Boller, I., and Dunger, D. (1992). Eating disorders in adolescents with IDDM. A controlled study. Diabetes Care, 15, 1356–60.

33. Fairburn, C.G. and Cooper, P.J. (1982). Self-induced vomiting and bulimia nervosa: an undetected problem. British Medical Journal, 284, 1153–5.

34. Fairburn, C.G. and Cooper, P.J. (1984). Binge-eating, self-induced vomiting and laxative abuse: a community study. Psychological Medicine, 14, 401–10.

35. Turnbull, S., Ward, A., Treasure, J., Jick, H., and Derby, L. (1996). The demand for eating disorder care. British Journal of Psychiatry, 169, 705–12.

36. Millar, H.R. (1998). New eating disorder service. Psychiatric Bulletin, 22, 751–4.

37. Fairburn, C.G. and Beglin, S.J. (1990). Studies of the epidemiology of bulimia nervosa. American Journal of Psychiatry, 147, 401–8.

38. van Hoeken, D., Lucas, A.R., and Hoek, H.W. (1998). Epidemiology. In Neurobiology in the treatment of eating disorders (ed. H.W. Hoek, J.L. Treasure, and M.A. Katzman), pp. 97–126. Wiley, Chichester.

39. Garfinkel, P.E., Lin, E., Goering, P., et al. (1995). Bulimia nervosa in a Canadian community sample: prevalence and comparison of subgroups. American Journal of Psychiatry, 152, 1052–8.

40. Kendler, K.S., MacLean, C., Neale, M., Kessler, R., Heath, A., and Eaves, L. (1991). The genetic epidemiology of bulimia nervosa. American Journal of Psychiatry, 148, 1627–37.

41. Russell, G.F.M. (1997). The history of bulimia nervosa. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 11–24. Guilford Press, New York.

42. Palmer, R.L. (1998). Aetiology of bulimia nervosa. In Neurobiology in the treatment of eating disorders (ed. H.W. Hoek, J.L. Treasure, and M.A. Katzman), pp. 143–59. Wiley, Chichester.

43. Garner, D.M., Garfinkel, P.E., Rockert, W., and Olmsted, M.P. (1987). A prospective study of eating disturbances in the ballet. Psychotherapy and Psychosomatics, 48, 170–5.

44. Schmidt, U., Tiller, J., and Treasure, J. (1993). Setting the scene for eating disorders: childhood care, classification and course of illness. Psychological Medicine, 23, 663–72.

45. Welch, S.L., Doll, H.A., and Fairburn, C.G. (1997). Life events and the onset of bulimia nervosa: a controlled study. Psychological Medicine, 27, 515–22.

46. Fallon, P. and Wonderlich, S.A. (1997). Sexual abuse and other forms of trauma. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 394–414. Guilford Press, New York.

47. Treasure, J. and Holland, A. (1991). Genetic vulnerability to eating disorders: evidence from twin and family studies. In Child and youth psychiatry: European perspectives (ed. H. Remschmidt and M.H. Schmidt), pp. 59–67. Hogrefe and Huber, New York.

48. Hettema, J.M., Neale, M.C., and Kendler, K.S. (1995). Physical similarity and the equal-environment assumption in twin studies of psychiatric disorders. Behavior Genetics, 25, 327–35.

49. Bulik, C.M., Sullivan, P.F., and Kendler, K.S. (1998). Heritability of binge-eating and broadly defined bulimia nervosa. Biological Psychiatry, 44, 1210–18.

50. Cowen, P.J., Anderson, I.M., and Fairburn, C.G. (1992). Neurochemical effects of dieting: relevance to changes in eating and affective disorders. In The biology of feast and famine (ed. G.H. Anderson and S.H. Kennedy), pp. 269–84. Academic Press, San Diego, CA.

51. Smith, K.A., Fairburn, C.G., and Cowen, P.J. (1999). Symptomatic relapse in bulimia nervosa following acute tryptophan depletion. Archives of General Psychiatry, 56, 171–6.

52. Kaye, W.H., Greeno, C.G., Moss, H., et al. (1998). Alterations in serotonin activity and psychiatric symptoms after recovery from bulimia nervosa. Archives of General Psychiatry, 55, 927–35.

53. Fairburn, C.G. and Cooper, Z. (1993). The Eating Disorder Examination (12th edition). In Binge eating: nature, assessment and treatment (ed. C.G. Fairburn and G.T. Wilson), pp. 317–60. Guilford Press, New York.

54. Garner, D.M. (1995). Assessment of eating disorder psychopathology. In Eating disorders and obesity (ed. K.D. Brownell and C.G. Fairburn), pp. 117–21. Guilford Press, New York.

55. Garner, D.M. (1991). Eating disorder inventory, Vol. 2. Psychological Assessment Resources, Odessa, FL.

56. Fairburn, C.G. and Beglin, S.J. (1994). Assessment of eating disorders: interview or self-report questionnaire? International Journal of Eating Disorders, 16, 363–70.

57. Wilson, G.T. and Fairburn, C.G. (1998). Treatments for eating disorders. In A guide to treatments that work (ed. P.E. Nathan and J.M. Gorman), pp. 501–30. Oxford University Press, New York.

58. Wilson, G.T. (1998). The clinical utility of randomised controlled trials. International Journal of Eating Disorders, 24, 13–29.

59. Fairburn, C.G. (1981). A cognitive behavioural approach to the management of bulimia. Psychological Medicine, 11, 707–11.

60. Fairburn, C.G. and Carter, J.C. (1997). Self-help and guided self-help for binge-eating problems. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 494–9. Guilford Press, New York.

61. Pasquale, L., Sciuto, G., Cocchi, S., Ronchi, P., and Bellodi, L. (1994). A family study of obsessive compulsive, eating and mood disorders. European Psychiatry, 9, 33–8.

62. Klerman, G.L., Weissman, M.M., Rounsaville, B.J., and Chevron, E.S. (1984). Interpersonal psychotherapy of depression. Basic Books, New York.

63. Fairburn, C.G. (1997). Interpersonal psychotherapy for bulimia nervosa. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 278–94. Guilford Press, New York.

64. Fairburn, C.G., Jones, R., Peveler, R.C., Hope, R.A., and O'Connor, M. (1993). Psychotherapy and bulimia nervosa: the longer-term effects of interpersonal psychotherapy, behaviour therapy and cognitive behaviour therapy. Archives of General Psychiatry, 50, 419–28.

65. Bulik, C.M., Sullivan, P.F., Carter, F.A., McIntosh, V.V., and Joyce, P.R. (1998). The role of exposure with response prevention in the cognitive-behavioural therapy for bulimia nervosa. Psychological Medicine, 28, 611–23.

66. Cooper, P.J. (1995). Bulimia nervosa and binge eating: a guide to recovery. Robinson, London.

67. Fairburn, C.G. (1995). Overcoming binge eating. Guilford Press, New York.

68. Schmidt, U. and Treasure, J. (1993). Getting better bit(e) by bit(e). Erlbaum, Hove.

69. Carter, J.C. and Fairburn, C.G. (1998). Cognitive-behavioral self-help for binge eating disorder: a controlled effectiveness study. Journal of Clinical and Consulting Psychology, 66, 616–23.

70. Rosen, J.C. (1997). Cognitive-behavioural body image therapy. In Handbook of treatment for eating disorders (ed. D.M. Garner and P.E. Garfinkel), pp. 188–201. Guilford Press, New York.

71. Fennell, M.J.V. (1997). Low self-esteem: a cognitive perspective. Behavioural and Cognitive Psychotherapy, 25, 1–25.

72. Keel, P.K. and Mitchell, J.E. (1997). Outcome in bulimia nervosa. American Journal of Psychiatry, 153, 313–21.

73. Collings, S. and King, M. (1994). Ten-year follow-up of 50 patients with bulimia nervosa. British Journal of Psychiatry, 164, 80–7.

74. Keel, P.K., Mitchell, J.E., Miller, K.B., Davis, T.L., and Crow, S.J. (1999). Long-term outcome of bulimia nervosa. Archives of General Psychiatry, 56, 63–9.

75. Fairburn, C.G., Stein, A., and Jones, R. (1992). Eating habits and eating disorders during pregnancy. Psychosomatic Medicine, 54, 665–72.

76. Morgan, J.F., Lacey, H.J., and Sedgwick, P.M. (1999). Impact of pregnancy on bulimia nervosa. British Journal of Psychiatry, 174, 135–40.