2.1.1. Symptoms, presentation and patterns of illness

Anorexia nervosa is a syndrome in which the individual maintains a low weight as a result of a pre-occupation with body weight, construed either as a fear of fatness or pursuit of thinness. In anorexia nervosa, weight is maintained at least 15 per cent below that expected, or in adults body mass index (BMI) – calculated as weight in kilograms divided by height in metres squared – is below 17.5 kg/m². In younger people, the diagnosis may be made in those who fail to gain weight during the expected growth spurt of puberty, as they can become underweight without weight loss.

Weight loss in anorexia nervosa is induced by avoiding ‘fattening foods’, sometimes supported by excessive exercising or self-induced purging (by vomiting or misuse of laxatives). As a consequence of poor nutrition, a widespread endocrine disorder involving the hypothalamic-pituitary-gonadal axis develops, manifest in women by amenorrhoea and in men by a lack of sexual interest or potency. In prepubertal children, puberty is delayed and growth and physical development are usually stunted.

The subjective experience of anorexia nervosa is often at odds with the assessment of others. The conviction that weight control is desirable is usually strongly held, particularly when challenged and others are seen as mistaken in believing the person should gain weight, particularly where there is a marked disturbance of body image. Weight loss is experienced as a positive achievement and, therefore, may be strongly reinforcing to someone with low confidence and poor self-esteem. As a result, they will often deny the seriousness of the condition. The essential role of ‘weight phobia’ is increasingly being questioned however, and is believed by some to be culture specific.

The condition generally starts with dieting behaviour that may evoke no concern. Indeed, some will experience reinforcing compliments. After a while, however, the commitment to dieting increases, often with a number of secondary features such as social withdrawal, rigidity and obsessionality, particularly where these traits have previously been features of the person’s personality. A number of secondary difficulties may develop including physical adverse effects, social isolation, compromise of educational and employment plans and occupation in the areas of leisure, self-care, daily living and productivity of employment and/or education. A smaller number will enter anorexia nervosa through a pattern of purging behaviour without dieting, following a viral illness, which resulted in weight loss that then became positively valued, or in the context of a chronic illness such as diabetes or Crohn’s disease.

Typically individuals are persuaded to seek help by concerned family members, teaching staff or general practitioners with whom they consult about physical consequences. Sometimes, however, the person begins to appreciate the damaging effects of the disorder and may seek help in their own right. Children and adolescents are almost always brought to treatment, very rarely actively seeking help initially and can present more complex diagnostic challenges (Bryant-Waugh et al., 1992).

2.1.2. Diagnosis

The diagnosis of anorexia nervosa in its typical form is a relatively straightforward one in older adolescents and adults. The diagnosis has good validity and reliability, the main obstacle to diagnosis being the person’s own willingness or otherwise to disclose his or her motives, symptoms and behaviours. Thus, engagement in a supportive, empathic assessment interview is crucial in enabling the person to reveal fears around weight, dieting behaviour and any purging or other maladaptive behaviour such as excessive exercising. In the absence of this engagement, the individual may fail to reveal weight-controlling behaviours and collude with the doctor in pursuing physical investigations to explain the weight loss. In women, the presence of secondary amenorrhoea (i.e. cessation of menstruation after it has been established) or other physical features of starvation should always alert the physician to the possibility of this diagnosis. Diagnosis may be more problematic in children and younger adolescents, as the existing diagnostic criteria are insufficiently developmentally sensitive (Lask & Bryant-Waugh, 2000).

The diagnosis is made on the basis of the history, supported where possible by a corroborative account from a relative or friend. Physical examination, with measurement of weight and height and calculation of body mass index (BMI), can reveal the extent of emaciation. On occasion, clinical observation during a hospital assessment can enable characteristic behaviours to be observed. Physical investigations are less useful in making the diagnosis but are crucial in assessing the physical impact of the disorder and its complications. Depending on the results of the physical examination, these may include haematological tests, electrocardiography, radiological assessment and ultrasound (Royal College of Psychiatrists, 2002).

A diagnostic challenge occurs in those with comorbid physical disorders, such as diabetes, chronic bowel or thyroid disorder. In diabetes, the patient may be tempted to restrict insulin intake in order to lose calories, whilst on occasions the symptoms of organic intestinal disorder may mask the psychological condition.

The weight loss that occurs with the anorexia of depression can usually be distinguished from that resulting from the dietary control of anorexia nervosa, but the condition can sometimes be difficult to distinguish from post-viral and other chronic fatigue syndromes where food intake is poor. Weight loss and limited food intake secondary to a brain tumour are also known to have been mistaken for anorexia nervosa.

2.1.3. Physical and social consequences

Although in the acute stages of anorexia nervosa subjective distress may be limited, emotional disturbance is common, chiefly comprising anxiety and mood symptoms. With time, emotional difficulties usually increase along with a range of physical and social difficulties, including becoming unable to care for oneself adequately, reducing or stopping leisure activities, interrupting educational goals and losing personal autonomy. These affect the person’s quality of life and increase the reliance on and the importance of the eating disorder.

Depression is a common comorbid diagnosis, with rates of up to 63 per cent in some studies (Herzog et al., 1992), while obsessive-compulsive disorder (OCD) has been found to be present in 35 per cent of patients with anorexia nervosa (Rastam, 1992).

Physical problems can be classified as those due to the effects of starvation and the consequences of purging behaviour. Starvation affects every system in the body. In the musculo-skeletal system, this will be evident as weakness, loss of muscle strength (which also affects heart muscle), loss of bone density and impairment of linear growth. Young women with anorexia nervosa are at increased risk of bone fractures later in life (Lucas et al., 1999). The effects on the endocrine system have their impact on target organs, causing infertility, a risk of polycystic ovaries and loss of bone mineralisation. Where pubertal development has not been completed, incomplete development of secondary sexual characteristics may occur (Goldbloom & Kennedy, 1995) and permanent stunting of growth is common. Patients with anorexia nervosa have disorders in the reproductive hormones (low LH and FSH), suppressed TSH, growth hormone resistance and raised corticol levels. The effects of purging are described in Section 7.5.2, including long-term disabilities such as erosion of tooth enamel sometimes amounting to destruction of the whole dentition. Worn painful teeth can be a considerable concern to the patient in terms of comfort, appearance and, therefore, self-esteem.

Brain volume is reduced in anorexia nervosa (Dolan, Mitchell & Wakeling, 1988; Kohn et al., 1997; Kingston, Szmukler, Andrewes, Tress & Desmond, 1996; Krieg, Pirke, Lauer & Backmund, 1988; Swayze et al., 1996). There are two small longitudinal studies, which have examined the structural changes in the brain of adolescents after full weight gain (Golden et al., 1996; Katzman et al., 1996). Both found persistent deficits in grey matter (cell bodies of neurons and glial cells) although there was recovery of white matter (mainly myelinated axons). This supports the finding of grey matter deficits in people who have made a full recovery from their eating disorder (Lambe, Katzman, Mikulis, Kennedy & Zipursky, 1997). One post-mortem study reported that there was a reduction in basal dendritic fields and dendritic spine density (Neumarker et al., 1997).

Many of the cognitive deficits in anorexia nervosa are restored after weight recovery. However, some abnormalities in executive function remain after weight restoration. For example, people with eating disorders have scores greater than one standard deviation from the norms on tests of perceptual rigidity, perseveration and set shifting and the neurological sign dysdiadokinesis (Tchanturia, Morris, Surguladze & Treasure, 2002).

Although little is known of the effects short or long term of extreme weight loss on brain development and function in children, it is possible that such weight loss may have both short and long-term effects on cognitive functioning.

Social difficulties may result in continued dependence on family of origin into adult life and often include difficulties engaging in intimate relationships. Employment prospects may be adversely affected either because of the limitations of the disorder or the disruption caused by lengthy hospitalisations.

2.1.4. Course and prognosis

The course of anorexia nervosa is very variable. There is no good evidence on the prognosis for people with anorexia nervosa who do not access formal medical care (Treasure & Schmidt, 2002). A summary of 68 treatment studies published before 1989 with a length of follow-up of one to 33 years, found that 43 per cent of people recover completely, 36 per cent improve, 20 per cent develop a chronic eating disorder and five per cent die from anorexia nervosa (Steinhausen, 1995). The overall mortality in these long-term studies ranged from 0–21 per cent from a combination of physical complications and suicide. The all-cause standardised mortality ratio anorexia nervosa has been estimated at 9.6 (95 per cent Confidence Interval 7.8 to 11.5) Nielsen (2001), about three times higher than other psychiatric illnesses. The average annual risk of mortality has been calculated at 0.59 per cent per year in females averaged from 10 samples, with a minimum follow-up of six years (Neilsen et al., 1998). The mortality rate appears to be higher for people with lower weight during their illness and those presenting between 20 and 29 years of age.

A number of those with anorexia nervosa progresses to other eating disorders, particularly bulimia nervosa, but also binge eating disorder, highlighting the relationship between the disorders. Movement in the other direction is less common, but a number of those with anorexia nervosa gives a premorbid history of obesity in childhood or adolescence.

2.1.5. Anorexia nervosa in children and adolescents

Although the essential psychological features are similar, children and younger adolescents may present with delayed puberty or stunted growth as well as weight loss. Parents or teachers are generally the ones who raise concern and the young person may resist medical attention. Some young people will voice anxieties around unwanted aspects of development, particularly if they have experienced early puberty or feel unable to engage with their peers’ increasing adolescent independence and social experimentation. In some, bullying or teasing about weight may have provoked this concern.

Although the principles of making the diagnosis are the same as in adults and are often straightforward, the greatest diagnostic difficulty occurs in the youngest cases. In children between the ages of around eight and 12, the condition is less common than in older individuals and should be distinguished from other types of eating disturbance seen in middle childhood, such as selective eating and food avoidance emotional disorder. By definition feeding disorder of infancy and childhood has onset below age six. In pubescent cases with primary amenorrhoea, it can sometimes be difficult to judge whether puberty has been delayed from the normal variation in timing of puberty.

Reference to height and weight centile charts is useful in evaluating weight in comparison to height. It is particularly helpful to compare presenting centiles for weight and height with historical values, as these may identify stunting of height (where the young person has crossed height centile lines). The result of such stunting is that the person may not appear unduly thin, though his or her weight may be considerably below the projected level as indicated by premorbid height and weight. It is also helpful to plot body mass index on BMI centile charts, as BMI norms are not stable over age. Average BMI increases with age during childhood and adolescence, a BMI of 17.5 kg/m² being close to the mean for a child at the age of 12 (Cole et al., 1995).

In children and adolescents with atypical presentations of an eating disorder, consideration should be given to the possibility of separate underlying physical pathology. In these circumstances the involvement of a paediatrician should be considered.

The prognosis for children and adolescents with anorexia nervosa is variable. Some (particularly those with a rapid and early onset) will make a full recovery from a first episode. This is most likely where early physical and psychosocial development has been healthy and where there is an identified precipitating negative life event such as bereavement (North et al., 1997). In such cases and where onset is pre-pubertal, physical consequences such as stunted growth and pubertal delay are usually fully reversible. Others with a more insidious onset, with earlier social difficulties or abnormal personality development, may go on to have a more chronic course into middle age (Gowers et al., 1991).

2.2.1. Symptoms, presentation and pattern of illness

Bulimia nervosa is characterised by recurrent episodes of binge eating and secondly by compensatory behaviour (vomiting, purging, fasting or exercising or a combination of these) in order to prevent weight gain. Binge eating is accompanied by a subjective feeling of loss of control over eating. Self-induced vomiting and excessive exercise, as well as the misuse of laxatives, diuretics, thyroxine, amphetamine or other medication, may occur. As in anorexia nervosa, self-evaluation is unduly influenced by body shape and weight, and there may indeed have been an earlier episode of anorexia nervosa. The diagnosis of anorexia nervosa is given precedence over bulimia nervosa; hence in bulimia nervosa BMI is maintained above 17.5 kg/m² in adults and the equivalent in children and adolescents (see Section 2.1.5). There is some controversy concerning whether those who binge eat but do not purge should be included within this diagnostic category. The ICD10 criteria (WHO, 1992) stress the importance of purging behaviour on the grounds that vomiting and laxative misuse are considered pathological behaviours in our society in comparison to dieting and exercise. The DSM-IV criteria (APA, 1994) agree about the importance of compensatory behaviour but distinguish between the purging type of bulimia nervosa in which the person regularly engages in self-induced vomiting or the misuse of laxatives, diuretics or enemas, from the non-purging type in which other inappropriate compensatory behaviours such as fasting or excessive exercise occur but not vomiting or laxative misuse.

People with bulimia nervosa tend to not disclose their behaviour nor to seek out treatment readily although may be more likely to do so than those with anorexia nervosa. The condition appears to be subjectively less ‘valued’ than anorexia nervosa; indeed binge eating and purging are commonly associated with extreme subjective guilt and shame. These emotions are sometimes reinforced by the pejorative language used by relatives and others including some clinicians, who may refer to ‘confessing’ or ‘admitting’ to purging behaviour. A person’s ambivalence towards treatment often arises from the fear that they will be stopped from vomiting and purging and then left to face the consequences of their binge eating, i.e. excessive weight gain.

The condition usually develops at a slightly older age than anorexia nervosa (the mean age of onset is 18 to 19, compared to 16 to 17 for anorexia nervosa). Bulimia nervosa sometimes arises from a pre-existing anorexic illness. Where this is not the case the development of the disturbance is often essentially similar to that of anorexia nervosa, arising from a background of attempts to restrain eating. In bulimia nervosa however, dietary restriction cannot be maintained and is broken by episodes of reactive binge eating, which result from a combination of physiological and psychological factors. Compensatory behaviours follow in order to counteract the effect of binge eating on weight. The person, therefore, maintains a weight, usually within the normal range despite overeating but commonly progresses into a vicious cycle of attempted dieting, binge eating and compensatory purging, frequently on a daily basis. As these behaviours dominate daily life, the person becomes preoccupied with thoughts of food and life may be re-organised around shopping, eating and purging behaviour. Initially, those with bulimia nervosa are generally secretive about their bulimic episodes, though some may leave obvious signs of their disorder such as empty food packaging and occasionally bags of vomit for other family members to discover.

Bulimic episodes are frequently planned, with food purchased or prepared in order to be consumed without interruption. The individual may also avoid situations in which they are likely to be exposed to food or will find it difficult to control their eating, such as when eating out with others. This avoidance behaviour tends to add to any social and relationship difficulties that may be present.

Mood disturbance is extremely common in bulimia nervosa and symptoms of anxiety and tension are frequently experienced. Self-denigratory thoughts may develop out of disgust at overeating or purging whilst low self-esteem and physical self-loathing may in some be rooted in the past experience of physical or sexual abuse. Self-harm, commonly by scratching or cutting, is common. A significant proportion of those with bulimia nervosa have a history of disturbed interpersonal relationships with poor impulse control. Some will abuse alcohol and drugs.

2.2.2. Diagnosis

As in anorexia nervosa, the diagnosis depends on obtaining a history supported, as appropriate, by the corroborative account of a parent or relative. This will require an empathic, supportive, non-judgemental interview style in which the person is enabled to reveal the extent of his or her symptoms and behaviours. Although those with bulimia nervosa generally have fewer serious physical complications than those with anorexia nervosa, they commonly report more physical complaints when first seen. They may complain of fatigue, lethargy, or feeling bloated, and they may suffer constipation, abdominal pain and on occasions swelling of the hands and feet or irregular menstruation. There may also be erosion of dental enamel in which the lingual surface of the upper teeth is mainly affected, and it has been argued that this is virtually pathognomonic of vomiting (Mitchell, 1995).

Physical examination is often normal, though the salivary glands (particularly the parotid glands) may be enlarged. Calluses on the back of the hand may be found; these result from the use of the hand to stimulate the gag reflex and induce vomiting. Oedema is common in those who have used laxatives or diuretics whilst these behaviours also lead to fluid loss and subsequent dehydration, which in turn may result in a metabolic alkalosis. This is generally accompanied by hypochloraemia and hypokalaemia. Overall, about 10 per cent of those with bulimia nervosa have electrolyte abnormalities detected on routine screening (Mitchell, 1995). Metabolic acidosis can also occur in patients who are abusing laxatives as a result of the loss of bicarbonate from the bowel. Less commonly hyponatraemia, hypocalcaemia, hypophosphataemia and hypomagnesaemia may develop.

In the gastrointestinal system, oesophagitis may occur. Gastric dilatation that poses the risk of gastric rupture and death occurs rarely but may be the most common cause of fatality (Mitchell, 1997). Constipation is extremely common, mainly due to dehydration; steatorrhoea and protein losing gastroenteropathy have also been reported secondary to laxative abuse. Severe use of laxatives has been reported to cause cardiomyopathy and other types of myopathic disorder as a result of storage of Ipecac in muscle tissue (Mitchell, 1995). Electrocardiography may indicate heart conduction abnormalities and possibly an increase in the risk of mitral valve prolapse.

Abnormal electroencephalographic (EEG) findings have been reported probably secondary to fluid and electrolyte abnormalities; there does not seem to be an association between bulimia nervosa and epilepsy.

Endocrine abnormalities are variable. These include abnormalities in the menstrual cycle and blunting of the thyroid stimulating hormone and growth hormone response to thyroid releasing hormone.

2.2.3. Impairment and disability

There is considerable overlap between the long-term disabling consequences of bulimia nervosa and those of anorexia nervosa. Mood and anxiety symptoms are very common. These symptoms, low self-esteem and body image disturbance can all have a negative effect on social relationships, which in turn may be damaged by a lifestyle that may be chaotic and characterised by impulsivity.

The adverse physical consequences of purging behaviour have been identified in Section 7.5.2. In addition, those with bulimia nervosa may be at risk from the effects of alternating weight loss and weight gain.

Lissner et al. (1991) in a series of 3130 participants found that both all-cause and coronary heart disease mortality were increased significantly in both men and women with high levels of weight variability. Morbidity from coronary heart disease was also increased.

Estimates of the prevalence of the diagnoses of personality disorder in people with bulimia nervosa have ranged from 21 per cent to 77 per cent. Obsessive-compulsive and avoidant personality disorders (Cluster C) have been described frequently (Braun et al., 1994). The relationship of borderline personality disorder to bulimia nervosa has been a source of considerable debate (Wonderlich, 1995) with reported rates ranging from two per cent to 47 per cent, these rates apparently influenced by subject and measurement variability across studies, thus BPD probably occurs no more frequently than other PDs in bulimia nervosa.

2.2.4. Course and prognosis

There have been few studies with a lengthy follow-up period of the course and outcome of bulimia nervosa in the community. Many people with bulimia nervosa are not receiving any form of help (Hsu, 1995). Of these, the majority will suffer chronicity or a relapsing course, maintained by over-valued belief in the importance of appearance and thinness in particular (Fairburn et al., 2000). With the most effective treatments about 50 per cent of people with bulimia nervosa can be expected to be asymptomatic two to 10 years after assessment. Twenty per cent are likely to continue with the full form of bulimia nervosa whilst the remainder (30 per cent) have a course of illness characterised either by remissions or relapses or persistent but subdiagnostic bulimia (Hsu, 1995). One 10-year follow-up study of 50 people with bulimia nervosa found that 52 per cent had fully recovered and only nine per cent continued to experience symptoms of bulimia nervosa (Collings & King, 1994). A larger study of 222 followed-up for a mean of 11 years revealed that 11 per cent still met criteria for bulimia nervosa whereas 70 per cent were in full or partial remission (Keel et al., 1999).

There are few consistent predictors of longer-term outcome, though a number of studies have shown strong associations between weight fluctuation (which commonly occurs in bulimia nervosa) and negative health outcomes. Patients with the uni-impulsive form of the illness without additional control difficulties are also thought to do better. The mortality rate associated with bulimia is uncertain but may be higher than in the matched general population (Hsu, 1995).

2.2.5. Bulimia nervosa in children and adolescents

The full syndrome of bulimia nervosa is rarely seen in young people under the age of 14 (Bryant-Waugh & Lask, 1995). Indeed in these authors’ uniquely specialised clinic fewer than five per cent of children under the age of 14 presenting with eating disorders were diagnosed with bulimia nervosa. Where these presentations occurred the clinical features were the same as found in the older age group. Young people with this disorder were also, in common with adults with the same diagnosis, depressed and suffered poor self image (Bryant-Waugh & Lask, 1995). There is no substantial literature on treatment or outcome in this age group. Adolescents with bulimia nervosa may sometimes be considered to be suffering incipient personality disorder, though caution should be exercised in diagnosing personality disorder before development is completed. In these cases it can be difficult to judge whether the eating disorder is contributing to abnormal personality development or conversely, if the personality difficulties have acted as a risk factor for the development of bulimia nervosa.

Since bulimia nervosa is very rarely seen in children and uncommonly in younger adolescents, and there has been no research on the treatment of adolescents with bulimia nervosa, in line with much current clinical practice and taking into account the above issues, the GDG took the view that, subject to adaptation to their age, circumstances and level of development, adolescent patients with bulimia nervosa should receive the same type of treatment as adults with the disorder along with appropriate family involvement.

2.3.1. Symptoms, presentation and pattern of illness

A number of people suffer from eating disorders that closely resemble anorexia nervosa and bulimia nervosa, but which are considered atypical, as they do not meet the precise diagnostic criteria for these conditions (Fairburn & Harrison, 2003; Turner & Bryant-Waugh, 2003; Ricca et al., 2001). In Europe, these are often termed ‘atypical eating disorders’ (Fairburn & Harrison, 2003), the equivalent American term being ‘eating disorders not otherwise specified’ (American Psychiatric Association, 1994). For example, the patient’s weight might be just above the diagnostic threshold for anorexia nervosa or she might still be menstruating. Binge eating and purging may occur less frequently than specified for a diagnosis of bulimia nervosa. Over concern with weight and shape is generally present in these disorders, although in some the primary focus is on maintaining strict control over eating. Although the diagnostic criteria may not be met, many atypical eating disorders are as severe and long lasting as anorexia nervosa and bulimia nervosa. Patients’ treatment needs and prognosis may be virtually identical.

Binge eating disorder (BED) is a recently described condition, first defined as a research category in DSM-IV (APA, 1994), though there is some overlap with the ICD 10 category ‘Overeating associated with other psychiatric disturbance’ (F50.4) (WHO, 1992). In BED, individuals engage in uncontrollable episodes of binge eating but do not use compensatory purging behaviours. These binge eating episodes are associated with three or more of the following:

• Eating much more rapidly than normal
• Eating until feeling uncomfortably full
• Eating large amounts of food when not physically hungry
• Eating alone through embarrassment at the amount one is eating
• Feeling disgust or extreme guilt after overeating.

Marked distress regarding binge eating is present and social avoidance is common.

2.3.2. Diagnosis

Atypical eating disorders are conditions of clinical severity that do not conform to the diagnostic criteria for anorexia nervosa or bulimia nervosa. An example would be someone with extreme dietary restraint, who exercised excessively to control weight, which was maintained in the low normal range. This condition would also include those with the features of anorexia nervosa at low weight who are still menstruating. Many people with atypical eating disorders have suffered with anorexia nervosa or bulimia nervosa in the past.

In comparison with anorexia nervosa and bulimia nervosa, far less is known about binge eating disorder (Fairburn & Harrison, 2003). Apart from binge eating, its systematic profile overlaps little with the other eating disorders. Defining binge eating can be problematic and there may be a discrepancy between the subjective experience and clinical assessment of a binge. The onset of binge eating disorder is usually in the teenage years or early 20s, but people tend to present later, typically in their 30s or 40s, when they have become overweight or obese. The sex ratio is more even and binge eating generally occurs against a background of a tendency to overeat rather than of dietary restraint. Many people with binge eating disorder are obese. By definition self-induced vomiting and laxative misuse are not present or only occasionally present. Depressive features and dissatisfaction with shape is common, though over-evaluation of the importance of weight and shape is less marked than in bulimia nervosa.

As with the other eating disorders, the diagnosis of binge eating disorder is made on the basis of the history, with physical investigations being used to assess any physical consequences. Observation and assessment in hospital is rarely indicated, though where available as part of a day programme, assessment of meal preparation and eating may be a useful adjunct to treatment planning.

2.3.3. Impairment and disability

Where atypical eating disorders are similar to the full syndromes of anorexia and bulimia nervosa, the physical dangers and psychosocial impairments closely resemble those of the diagnostic conditions. Many of those with BED will suffer similar physical complications of binge eating to those with bulimia nervosa, though purging carries greater physical risk than binge eating. Those who are obese are at risk of the psychological and physical disabilities associated with this condition, namely low self-esteem, diabetes, heart disease, hypertension and stroke.

2.3.4. Course and prognosis

The prognosis for those disorders that resemble anorexia nervosa and bulimia nervosa depends on the severity of the associated physical and psychological features. Those with binge eating disorder typically give long histories of proneness to binge eating but these may alternate with extended periods free from binge eating. The spontaneous remission rate may be high (Fairburn et al., 2000). Short-term response to treatment appears better than for anorexia nervosa and bulimia nervosa but there are as yet no studies of long-term course or outcome. There is a certain amount of movement from one diagnosis to another, thus those with atypical eating disorders may go on to develop bulimia nervosa, or more rarely anorexia nervosa.

2.3.5. Atypical eating disorders in children and adolescents

Atypical eating disorders are relatively commonly diagnosed in childhood, in part because of the difficulty in strictly applying existing diagnostic criteria for anorexia nervosa and bulimia nervosa, and in part because of an ongoing lack of clarity about the classification of eating disturbances in this age group. Children and younger adolescents may present with a range of other types of clinical eating disturbance, which may be different in terms of psychopathology to anorexia nervosa/bulimia nervosa presentations (Cooper et al., 2002). They frequently develop in those who have suffered feeding disorders of childhood and sometimes the distinction (particularly in middle childhood) can be a difficult one. Some have suggested that atypical eating disorders probably occur more commonly (Bryant-Waugh, 2000) and some have suggested modification to the ICD10 criteria for the diagnosis of eating disorders in children is required. Binge eating disorder has not been systematically investigated in this age group.

2.5.1. Genetic factors

The majority of family studies have shown that eating disorders run in families. In a large case-control family study (Strober et al., 2000), female relatives of those with anorexia nervosa were 11.4 times as likely to suffer the disorder than relatives of control subjects, while female relatives of those with bulimia nervosa were 3.7 times as likely to suffer with bulimia. Some family studies have also reported familial aggregation of milder sub-diagnostic eating disorders or related concerns (Strober, 2000). The prevalence of full and partial bulimia nervosa has been shown to be more common in female relatives of those with anorexia nervosa than in relatives of control subjects and vice versa. These findings suggest that eating disorders form a spectrum of clinical severity in which there is a continuum of familial liability. The aggregation of full and subclinical eating disorders suggests that genetic factors are likely to be involved in causation. A twin study of anorexia nervosa has estimated the heritability to be 58 per cent (95 per cent Confidence Interval 33 per cent to 77 per cent) with the remaining variance explained by non-shared environment (Wade et al., 2000). Genetic factors are also purported to play a role in the aetiology of bulimia nervosa (Kendler et al., 1991). Fifty-four per cent to 83 per cent of the variance in liability has been thought to be due to common genetic factors (Bulik et al., 2000), but again the confidence intervals around the estimation of heritability are broad, thus the relative contribution of genetic to other factors is unclear. The contribution of twin studies to the estimates of heritability of eating disorders has been criticised (e.g. Fairburn et al., 1999) on the grounds of ascertainment bias, small sample sizes or violation of the equal environment assumption, which assumes that both identical and non-identical twins are equally exposed to aetiological environmental factors.

2.5.2. Physical risk factors

A history of premorbid obesity has been documented in series of those with both anorexia nervosa (seven to 20 per cent) and bulimia nervosa (18 to 40 per cent) (Cooper, 1995) as has constitutional leanness/thinness. There is prospective evidence that this experience leads to a propensity to an increase in body dissatisfaction and likelihood of dieting behaviour (Stice, 2002) as well as greater self-evaluation in terms of weight. However, there is little prospective evidence that high body mass leads to an increase in eating pathology.

Early feeding difficulties in infancy or early childhood are often documented but it is unclear whether this constitutes a physical risk factor or is related to the mother child relationship.

Although a range of neuro-endocrine and metabolic disturbances occurs in those with eating disorders, the evidence suggests that these disturbances are secondary rather than primary to the disorder.

Early menarche has long been considered a risk factor for eating disorder through a putative relationship with adiposity and body dissatisfaction.

The dietary restraint model suggests that calorie restriction increases the risk for binge eating and bulimia nervosa. Although dieting appears to increase negative affect and may contribute to eating difficulties, dieting has a small effect size in the contribution to the development of eating pathology (Stice, 2002).

2.5.3. Adverse life events and difficulties

Severe life stresses have been implicated in the aetiology of both anorexia nervosa and bulimia nervosa, with approximately 70 per cent of cases being triggered by severe life events or difficulties. These stresses most commonly occur in the area of close relationships with family or friends (Schmidt et al., 1997, Welch et al., 1997). One controlled study of anorexia nervosa in adolescents (North et al., 1997) suggested that a severe independent negative life event had occurred in the year before onset of approximately one quarter of an adolescent series and this was associated with a good prognosis. Particular attention has focused on the experience of childhood sexual abuse as a potential predisposing factor. There is little persuasive evidence that either sexual abuse or other stressful life events are specific predisposing factors for eating disorders rather than psychiatric disorder per se. Childhood sexual abuse did not emerge as a significant predictor of the onset of binge eating in the only prospective study to date (Vogelantz-Holm et al., 2000)

2.5.4. Family factors

Mounting evidence from family studies suggests increased rates of affective disorder among first and second degree relatives of people with both anorexia nervosa and bulimia nervosa. This is three times greater than for relatives of normal control participants (Cooper, 1995). Several family studies have revealed higher rates of alcohol abuse among first degree relatives of those with bulimia nervosa but again, the specificity to eating disorder as opposed to psychiatric disorder in general is unclear. A family history of substance abuse may however be a specific risk factor for bulimia nervosa (Cooper, 1995).

A number of studies have looked at family environment and functioning. There are difficulties extrapolating from disturbance within the family after the disorder has developed and assuming that it predated the onset of disorder. Prospective studies have to date failed to provide support for the causative role of dysfunctional family systems (Stice, 2002). Familial faddy eating and undue concern about the importance of weight and shape have been described in family members of those with eating disorders and may contribute to weight and shape concern in vulnerable adolescents (Gowers & Shore, 2001). There is also growing evidence of disturbance of eating in the children of mothers with eating disorders (Stein, 1995) but it is as yet unclear whether these children will be at risk of anorexia and bulimia nervosa in their own adolescence. One controlled study demonstrated a style of ‘high concern’ parenting in a number of adolescent cases of anorexia nervosa, which arose in infancy and long predated the onset of eating difficulties. In a quarter of cases, this followed an earlier severe obstetric tragedy prior to the subject’s birth. These rates were significantly higher than in a matched control group (Shoebridge & Gowers, 2000) and challenge the notion that observed parenting styles are solely a response to the development of an eating disorder in adolescents.

2.5.5. Socio-cultural factors

A number of socio-cultural theories have been put forward to explain the aetiology of eating disorders. Such theories include the meaning of weight and shape for women in different cultures and the impact of advertising and other media. It is argued that societal pressure to be thin fosters an internalisation of a thin ideal and body dissatisfaction, which in turn leads to dieting behaviour and places the person at risk for eating pathology (Striegel-Moore et al., 1986). This perceived pressure does appear to predict dieting and eating pathology (Stice, 2002). Furthermore, in prospective studies perceived pressure to be thin predicted the onset of binge eating, bulimic symptoms and increases in eating pathology (Stice, 2002).

2.5.6. Perfectionism

This personality trait has long been considered a risk factor for eating pathology as it may promote the relentless pursuit of the ‘thin ideal’. Fairburn (1997) has implicated perfectionism as a maintenance factor for bulimic pathology as it encourages rigid dieting thought to underlie the binge-purge cycle. Meta-analysis of prospective studies provides support for the notion of perfectionism as a risk factor for bulimic pathology and a maintenance factor for more general eating pathology (Stice, 2002).

2.5.7. Impulsivity

General problems in the area of impulse control have been proposed as risk factors for eating pathology as they may render the individual vulnerable to episodes of uncontrollable binge eating (Hawkins & Clement, 1984). The empirical support for this association is weak (Stice, 2002), however some indirect support is offered by the finding of an association between substance use and bulimic pathology (Stice & Agras, 1998)

2.6.1. Primary care

Studies from the US and continental Europe suggest that only a fraction of people with eating disorders receive specialised treatment for their eating disorders. For example, in a large sample of female primary care and obstetric gynecology attendees in the US fewer than one in 10 cases with bulimia nervosa or BED were recognised by the physician (Johnson et al., 2001). In the Netherlands, on average, only 40 per cent of the community cases of anorexia nervosa are detected by general practitioners and 79 per cent of these patients are referred on for mental health care (Hoek, 2003). In the case of bulimia nervosa, only a small subgroup, an average 11 per cent of the community cases, are detected in primary care and 51 per cent of these cases are referred to secondary care. A general practitioner with a list of 2000 people could expect to have one or two patients with anorexia nervosa and 18 patients with bulimia nervosa (Hoek, 1991). Importantly, the poor detection rates for eating disorders do not suggest that these patients avoid their general practitioners. Indeed, over the five years prior to the diagnosis of the eating disorder, these individuals consult their general practitioners significantly more frequently than do people without an eating disorder (Ogg et al., 1997). In these earlier consultations, patients typically present with psychological, gastro-intestinal or gynaecological complaints prompted by symptoms of the eating disorder or its complications. These findings suggest that there are high levels of hidden eating disorder morbidity at primary care level leading to a considerable underestimate of costs, and greatly emphasise the importance of primary care in the detection and early treatment of eating disorders.

In the UK, an epidemiological study using the general practice research database found similar figures for referrals, 80 per cent of cases of anorexia nervosa and 60 per cent of cases of bulimia nervosa were referred on to secondary care, the majority to psychiatrists. The study also revealed that general practitioners tend to prescribe laxatives or diuretics for 27 of 100 people with the diagnosis of bulimia nervosa and psychotropic medication in 45 per cent of the eating disorders cases (Turnbull et al., 1996) It should also be noted that many GPs may fail to consider an eating disorder as possible diagnosis in children presenting with typical eating disorder features (Bryant-Waugh et al., 1992). The only calculation of the total cost of eating disorders in primary care was done based on the Third National Survey of Morbidity in General Practice which found that 1/1000 females and 0.6/1000 males consulted their general practitioner for anorexia nervosa (no data for bulimia nervosa, EDNOS or BED were available) in one year (Office of Health Economics, 1994). Applying these rates to the UK populations of 1990 suggest that approximately 47,000 individuals consult their GP with anorexia nervosa each year, which yields an estimated cost of £580,000 for anorexia nervosa consultations in primary care in 1991 prices (Office of Health Economics, 1994).

2.6.2. Secondary and tertiary care

A recent North American study analysing a national insurance claim database found that specialist eating disorder treatment provision is rare and eating disorders usually remain undetected or undertreated especially in the case of men (Striegel-Moore, 2000). A UK study of patients with anorexia nervosa revealed that under-diagnosis is common. Some patients may spend months or years in non-specialist units posing enormous extra costs for the NHS (Howlett, 1995). The NICAPS study (O’Herlihy et al., 2001) has shown that more beds were occupied by young people with eating disorders than any other diagnostic group, with the same number of patients with eating disorders in specialist eating disorders services being found as in non-specialist psychiatric units. In all, there was a total of 22.2 per cent of the inpatient population with a diagnosis of an eating disorder.

In the US as in the UK, outpatient treatment is the norm with low hospitalisation rates for eating disorders. Striegel-Moore (2000) reviewed costs for the disorders in the US. The average numbers of outpatient episodes are 17, 15.6 and 13.7 for anorexia nervosa, bulimia nervosa and EDNOS, respectively. Only 21.5 per cent of women with anorexia nervosa are hospitalised per year with an average length of stay in the US of 26 days, and very much lower hospitalisation rates were found for bulimia nervosa and EDNOS. Mean annual treatment (outpatient and inpatient) costs for female patients were $6,045 for anorexia nervosa, $2,962 for bulimia nervosa and $3,207 for EDNOS. These costs of treating eating disorders were compared with treatment costs of schizophrenia and obsessive-compulsive disorder (OCD). The mean treatment cost for anorexia nervosa was significantly higher than the mean cost for schizophrenia, but mean treatment cost for bulimia nervosa was significantly lower than for schizophrenia. Treatment costs for any eating disorder were much greater than that of OCD. These cost estimates are likely to represent an underestimate of the true costs of these illnesses since costs were calculated based on diagnostic codes (Crow & Peterson, 2003).

In the UK in 1985 an estimated 25,748 bed days were allocated to inpatient treatment of women and girls with anorexia nervosa (Hospital Inpatient Enquiry) with the average length of 21.5 days per episode. Only a fraction of the medication costs were estimated in the same study. There are no more recent UK estimates about inpatient treatment, other medication, outpatient treatment and psychotherapy costs, and no data are available for bulimia nervosa, EDNOS or BED. When looking at the organisation of eating disorders care in the UK, it is important to mention the significant role of private providers even though they are acknowledged as being a more expensive way to contract with providers (Brown, 1997).

The most recent estimate of the health care costs of eating disorders comes from Germany. Krauth et al. (2002) calculated that the health care costs of anorexia nervosa for the year 1998 amounts to 64.9 million euros and that of bulimia nervosa to 9.8 million euros without the costs of primary care, outpatient care (including psychotherapy) and pharmaceuticals. The estimated hospitalisation cost of 12,800 euros per patient with anorexia nervosa was found to be much higher than the hospitalisation cost of 3600 euros for an average patient with bulimia nervosa (Krauth et al., 2002). However, it is worth mentioning that in Germany the threshold for inpatient admission for anorexia nervosa is probably much lower than in the NHS in the UK if not in the private sector.

The cost of managing physical and dental complications (such as infertility, osteoporosis and reconstructive dentistry) in secondary care has not been quantified but may be considerable.

2.6.3. Non-health care burden

Eating disorders have a substantial impact on social functioning, including occupational and educational impairment. These disorders are the most common at an age when people are in secondary school, in higher education or at the beginning of their working careers. Eating disorders often result in lost productivity due to the inability of people to work or premature death. The lost productivity costs of anorexia nervosa were estimated to be 130.5 million euros (inability to work: eight million euros, death: 122.5 million euros) for the 82 million German population in the year 1998. The same figure for the indirect economic burden of bulimia nervosa was 113.9 million euros (inability to work: 1.7 million euros, death: 112.2 million euros) (Krauth et al., 2002).

The average duration of an episode of anorexia nervosa is six years (Herzog et al., 1997) and the family are usually the main carers for most of the duration of the illness. A recent UK pilot study found that carers of anorexic patients reported similar experiences in terms of the difficulties experienced to those of carers of adults with psychosis, but had significantly higher level of psychological distress (Treasure et al., 2001). The burden of caregiving has never been examined in economic terms.

These estimates suggest that the broader social costs of eating disorders may be even more substantial than their health care costs, and that early diagnosis, prevention of chronicity and appropriate and cost-effective treatments may greatly reduce the broad economic burden of eating disorders.

2.7.1. Pharmacological treatment

Pharmacotherapy has not been the treatment of first choice for eating disorders, but it has been used as an adjunct to psychological therapies or to treat physical or co-morbid psychological problems. In anorexia nervosa, medication has to date been disappointing in influencing the core symptoms of the disorder, promoting weight gain or reducing associated mood disturbance (Treasure & Schmidt, 2001). Where modest improvements have been reported, consideration of unwanted effects (e.g. in prolonging the QT cardiac interval), have led researchers to conclude that drugs confer little advantage when added to standard treatment (e.g. Vandereycken, 1984). In bulimia nervosa and binge eating disorder, whilst drugs are not the first option, there is some evidence that antidepressants, particularly selective serotonin reuptake inhibitors (SSRIs) contribute to the cessation of binge eating and purging. As endogenous opiates play a role in the hypothalamic regulation of hunger and satiety, a few trials have been conducted with opiate antagonists.

2.7.2. Psychological interventions

Choosing the most effective treatment approach for someone with an eating disorder should take account of a number of variables, including physical and psychological risk, motivation, social support, comorbidity, and age. Often and particularly in anorexia nervosa, treatment planning will require co-ordinated, multidisciplinary, physical, psychological and service interventions. Psychological interventions are, however, considered to be crucial in addressing the core attitudes that underlie these disorders and in influencing the longer-term outcomes. The nature of the psychological therapies chosen will be influenced by patient preference, their motivation, the nature of associated psychological features and their age or stage of development. Some will prefer a non-verbal projective therapy, using art, drama or music. Younger patients in particular or those who are dependent on relatives or carers are seen as often requiring family or systemic therapy. Of the individual therapies, cognitive behaviour therapy (CBT) is the best researched, though the evidence base in bulimia nervosa and binge eating disorder (BED) is far larger than in anorexia nervosa. The evidence base for other psychological therapies such as supportive psychotherapy and interpersonal psychotherapy (IPT) is growing, while more specialised therapies such as cognitive analytic therapy (CAT) and dialectical behaviour therapy (DBT) have been considered in complex cases, and for DBT those complicated by repeated self-harm (Treasure & Schmidt, 2003).

2.7.3. Service-level interventions

Decisions about the right treatment setting in which to manage a person with an eating disorder currently depend on the nature of the disorder, the level of risk, physical and psychological complications and patient preference. Within the NHS currently, the availability of different models of service is a major consideration with patients and referrers sometimes having to weigh up the advantages of a locally accessible service with those of a more comprehensive service at some geographical distance. In general, those with bulimia nervosa and binge eating disorder will be treated on an outpatient basis either in primary care or secondary services. Those with anorexia nervosa will generally be treated in secondary care, the choice of in, out, or day patient provision depending on the above considerations. Most will have a trial of outpatient intervention first. A stepped-care model in which patients move up from secondary to tertiary care subject to locally agreed protocols makes clinical sense, but there is at present little evidence to guide decisions about service setting. There is also considerable debate over the requirement for some or all of people with eating disorders to be treated in specialist eating disorder services. In general the current practice would seem to favour the treatment of the large majority of adult patients with anorexia within specialist eating disorder services, whereas such specialist services may only be needed for patients with severe forms of bulimia nervosa. However, irrespective of the service setting (specialist secondary or tertiary, or generalist) the key aim of this guideline will be to ensure that as far the evidence allows, effective treatments should be provided. Whatever the service, the setting should be age-appropriate, with the social, and educational needs of children and adolescents requiring particular attention in order to avoid the development of secondary handicaps. Inpatient management will require comprehensive physical and psychological management; currently a number of models exist from medical admission with psychiatric/psychological liaison to psychiatric admission with medical supervision. Children, in particular are often admitted to general medical (paediatric) services.

2.7.4. Primary–secondary care interface

Many patients with eating disorders will require treatment for a considerable time, often for a number of years. During this time, they may require treatment in primary care, secondary and on occasions, tertiary care services. The boundary between primary and secondary care is one of a number of interfaces that need to be managed effectively. Others include the interface between medical and psychiatric services, between inpatient and outpatient services, and for adolescents, between Child and Adolescent Mental Health Services (CAMHS) and adult services. It is particularly important for effective management that communication is good and that areas of responsibility are clear. Sometimes, for example a patient may be receiving psychological therapy from a secondary care service, but responsibility for physical monitoring may remain with primary care. Clear treatment protocols will ensure smooth management across these interfaces.

2.7.5. Physical health care

Patients with eating disorders always require assessment of their physical health and any associated risk as part of a comprehensive assessment. This should also be reviewed as treatment progresses. Physical complications can be classified as follows:

• The complications of starvation (chiefly in anorexia nervosa)
• The consequences of purging behaviours (including dental complications)
• The complications of associated physical conditions (such as diabetes or pregnancy)
• The complications of excessive exercise
• Physical aspects of stunted or incomplete development (in younger people); the main effects being on skeletal growth and the reproductive system
• Physical complications of dietary imbalance (e.g. a high fibre, low fat diet)
• Physical complications of refeeding or restoration of normal diet.

Many of these may have effects on a number of systems within the body. In any treatment plan, it must be clear who is taking responsibility for physical assessment and how any risk identified is to be managed. This often involves effective communication between primary and secondary care services (see Section 2.7.4).

2.7.6. The relationship of the evidence base for adults to that for children and adolescents

In the absence of a strong body of treatment research in the child and adolescent literature, it is tempting to extrapolate from adult findings, but the validity and limitations of doing so requires careful consideration. A key argument in favour of extrapolating from adult findings is that adolescence is a developmental stage that is not defined merely by age. It can be argued that many young adults with eating disorders are still in the throes of addressing the challenges of adolescence and indeed developmental difficulties have been thought to underlie the aetiology of anorexia nervosa in particular (e.g. Crisp, 1995). A second argument is that the essential features of anorexia nervosa and bulimia nervosa are consistent across the age spectrum – both in terms of characteristic behaviours (dieting, binge eating, purging), specific psychopathology (over-evaluation of the self in terms of weight and shape) and non-specific features (low self-esteem, perfectionism, poor interpersonal confidence). Thirdly some of the literature reports combined adolescent/adult case series without separate analysis. Finally some of the treatments that have been found to be effective in adult eating disorders (e.g. CBT and antidepressants) are effective in the treatment of adolescents with other conditions (particularly depression).

There are, however, a number of reservations about extrapolating from adult findings. In younger patients, the disorders less commonly fall neatly into the ICD10 or DSM-IV categories; that is to say, atypical forms are more commonly diagnosed. Some clinical eating disturbances seen in childhood and early adolescence currently classified as atypical eating disorders may not have much in common with the adult category of EDNOS. As treatments for EDNOS are poorly developed in adults, there are few findings in any case to draw on. The treatment aims, particularly in anorexia nervosa are often different in childhood and adolescence, because of the different physical issues involved, i.e. where the onset is before growth and development are complete, treatment needs to address the completion of puberty and growth. Unlike the treatment of adults with anorexia nervosa where recovery usually involves returning to a premorbid healthy physical state, in younger patients it is more a question of coming to terms with a new physical state. In terms of weight targets this involves ‘constantly moving the goal posts’ as healthy weight is recalculated with the attainment of greater height. All this might indicate a need for a longer duration of treatment for younger cases, whilst a shorter duration of illness before treatment might argue for the opposite.

When considering the literature on pharmacotherapy, one should be aware of differences in the pharmacodynamics and pharmacokinetics in children. In general the latter means that children and adolescents require higher doses of drugs per kg body weight to attain similar blood levels and therapeutic effect, owing to children’s more rapid metabolism by the liver and clearance by the kidney.

Irrespective of any consideration of aetiological variables, parents will usually need to be involved in the management of younger patients. This is especially so if they are at risk and parental involvement is likely to reduce the risk. Both anorexia nervosa and bulimia nervosa require aspects of behavioural management and parents will need to be involved if management of these is to be effective; at a practical level, parents usually have a role in shopping for food, meal planning, and managing meal times. The involvement of siblings is generally regarded as beneficial, for the sibling if not for the patient, as this provides an opportunity for them to express fears or guilt and to dispel any false ideas about the nature of the condition, its likely causes and prognosis.

Additional attention will need to be given to the different social and educational needs of this age group in treatment, particularly when treated in hospital.

In addition, carers and parents of adolescents have identified a lack of and need for support, involvement and education about eating disorders for themselves (Kopec-Schrader et al., 1993; Haigh & Treasure, 2003).

2.8.1. Clinical practice recommendations

2.8.1.1.

Assessment of people with eating disorders should be comprehensive and include physical, psychological and social needs, and a comprehensive assessment of risk to self. [C]

2.8.1.2.

The level of risk to the patient’s mental and physical health should be monitored as treatment progresses because it may increase – for example following weight change or at times of transition between services in cases of anorexia nervosa. [C]

2.8.1.3.

Health care professionals assessing children and adolescents with eating disorders should be alert to indicators of abuse (emotional, physical and sexual) and should remain so throughout treatment. [C]

2.8.1.4.

The right to confidentiality of children and adolescents with eating disorders should be respected. [C]

2.8.1.5.

Health care professionals working with children and adolescents with eating disorders should familiarise themselves with national guidelines and their employers’ policies in the area of confidentiality. [C]

2.10.1. Clinical practice recommendations

2.10.1.1.

Health care professionals should acknowledge that many people with eating disorders are ambivalent about treatment. Health care professionals should also recognise the consequent demands and challenges this presents. [C]

2.10.1.2.

Patients and, where appropriate, carers should be provided with education and information on the nature, course and treatment of eating disorders. [C]

2.11.1. Involving carers

Eating disorders have a social as well as an individual context. Once ‘out in the open’, family members and carers are inevitably drawn into the illness and many will want to offer whatever help they can. The UK government has recognised the important role and contribution that carers provide for people with a range of mental health problems and information and guidance is available through a website (www.carers.gov.uk).

2.11.2. Confidentiality

Some patients will be concerned that the involvement of their relatives may breach their rights of confidentiality. Consultations between those with eating disorders and health care professionals are bound by generic rules regarding confidentiality, i.e. that this should only be breached if the patient or others are at significant risk and that a breach of confidentiality is likely to reduce that risk (Department of Health, 2001; GMC, 2003). Furthermore the patient should be informed of any necessary breach of confidentiality. However issues of good practice with regard to confidentiality should not be accepted as an excuse for not listening to or communicating effectively with carers. Carers should be given sufficient information by medical and mental health services in a way that they can readily understand, to help them provide care effectively (Department of Health, 2003). Information from carers is also subject to the same rules of confidentiality as those applied to the individual with the illness.

2.11.3. Information for carers

Carers should be given the opportunity to ask health care professionals about the illness in general and specific risks. The following information might reasonably be expected to be available to them:

2.11.4. Carer involvement in treatment

The extent to which the family and/or next of kin need to be involved in treatment relates to age and developmental issues, the severity of the illness and the risk of harm. In general, parents and other family members will be included in the treatment of children and adolescents and in the treatment of adults depending on their wishes and the assessment of risk or severity. In the rare cases where compulsory treatment is necessary, there is a statutory role for the next of kin enshrined in the 1983 Mental Health Act legislation.

Therefore, although services should aim to keep treatment confidential they also have to ensure that they practice safely and consider the needs of carers. They have a statutory obligation to let other people know if there are health and safety issues that need to be considered.

2.11.5. Carers’ experiences and wishes

A survey of carers (Haigh & Treasure, 2003) and a focus group convened to inform the GDG, revealed a number of common themes:

• Carers are concerned about a perceived variable range of experience and expertise in the management of eating disorders amongst health service professionals.
• They express a need for quality information and effective communication.
• They often express a wish to be included in their relative’s ‘treatment team’.
• They are concerned about a perceived lack (in many areas) of adequate service provision and many worry about delays in accessing effective services and smooth transition between tiers of care.

The issues raised can be classified as engaging with primary care, experiences of treatment, and issues in aftercare. The carers focus group also made the following observations.