Anorexia nervosa in children and adolescents

Kelly Y C Lai 黎以菁

HK Pract 2005;27:142-148

Summary

Anorexia nervosa is a complex and protracted illness that is associated with a high rate of morbidities and mortality. It is now affecting an increasing number of adolescent girls and young women in non-Western societies. Effective treatment has to address the underlying psychological issues. Family therapy has been shown to be more effective than individual therapy for children and adolescents

摘要

神經性厭食症是一種複雜而長期的疾病。它常會引起病態甚至死亡。 該病現正在西方社會以外的地方出現，累及越來越多的少女和年輕女性。 有效的治療需要解決內在的心理問題。對兒童和青少年來說，家庭療法已顯示比個人療法更有成效。

Introduction

Anorexia nervosa has traditionally been conceptualised as a socio-culturally determined illness that was confined to Western societies. The emergence of reports from non-Western cultures such as Hong Kong, Japan,¹ India,² and Middle East countries,^3,4 among others, led to the speculation that globalisation of Western culture, with its slim body ideal and gender issues, is the culprit. More recently, findings from twin and family studies have sparked new interests in the biological and genetic aspects of the illness. No doubt anorexia nervosa is a complex illness that reflects multifactorial contributions involving a dynamic interaction between the genes and the environment. The fact that the disorder peaks in adolescence and young adulthood underscores its developmental context. This article will focus on some of the cross-cultural issues in the diagnosis and management of this disorder, and highlight areas where data on children and adolescent populations is available.

Epidemiology

The illness typically affects adolescent girls and young women, but pre-pubertal girls can also be affected. Boys and men can be affected too but the rate is one tenth that of girls. In Western countries, around 10% of adolescents demonstrated disordered eating behaviours,⁵ and the prevalence of the disorder is estimated to be about 0.2-0.9% for young females.6 Whether the incidence rate is on the rise is still in debate, but it is clear that Western figures have documented an increasing number seeking medical help.⁶ Equivalent figures in Hong Kong showed that a desire to be thinner was prevalent. In a survey of 357 adolescent female students between 15-21 years old, 85% wanted to weigh less, although only 4.8% were actually overweight.⁷ There has not been any epidemiological study to indicate a prevalence rate of the disorder per se.

Diagnosis

According to DSM-IV,⁸ the diagnosis is based on 1) a refusal to maintain body weight at or above a minimally normal weight for age and height, which is suggested to be less than 85% of the expected weight (defined as BMI 17.5kg/m² in ICD-10⁹) usually achieved through severe dieting and may be accompanied by compensatory behaviours such as excessive exercising, self-induced vomiting, laxative abuse, use of slimming agents and bingeing episodes; 2) a morbid fear of fatness; 3) disturbance in perception of body weight or shape, or denial of the seriousness of current low body weight, and 4) secondary amenorrhoea.

Though the criteria is widely accepted, cross-cultural research has particularly questioned the notion of fear of fatness being a central component of the illness. Lee et al,¹⁰ in his examination of 70 Hong Kong Chinese women with the disorder in the 1980s, reported that fear of fatness was present in less than 50% of his cases. Reasons given for not eating included abdominal bloating, nausea, loss of appetite, and difficulty in swallowing. He proposed that fear of fatness was a plastoplastic manifestation of the disorder. Yet another case series reported by Lai¹¹ of a group of 16 Hong Kong Chinese adolescents with the disorder in the 1990s found that fear of fatness was present in 87%. She suggested that the difference between the two series could be because as Western ideals about weight and shape became more widely accepted over time, the pursuit of thinness became legitimised and presented as a justified reason for food refusal among adolescents in the 1990s. The debate remains unresolved.

Diagnosing the disorder in pre-pubertal children poses other difficulties. Children present with a wider range of eating problems than adults, such as selective eating and eating disorders related to emotional disturbances. Yet, because of their relatively low body fat levels, children become emaciated much more quickly than adults. A weight of 85% expected weight may have much more severe consequences in children than in adults. On the other hand, a BMI of 17.5 may be normal in children. Lask et al¹² emphasised that a failure to gain weight in children should be a signal for serious concern. Secondly, when pre-pubertal children are affected, the amenorrhoea is primary and is not a useful indicator. Thirdly, not all children demonstrate the cognitive distortion of the disorder and some may not volunteer fear of fatness as a reason for food refusal.¹³ Hence the assessment of anorexia nervosa in the younger age group must include a comprehensive assessment from a developmental perspective.

Physical complications

In the acute stage, patients look emaciated and dehydrated, their skin is dry and sallow. There may be hair loss while fine, downy lanugo hairs may be present on their back and arms. The palms and soles may have an orange discolouration, which may denote liver dysfunction. In severe cases proximal muscle weakness may be present. On cardiovascular examination, the extremities are cold, pulse weak, and a bluish tinge may be present. There may be bradycardia and hypotension, associated with complaints of dizziness. In severe cases, arrhythmias and heart failure may be evident.

Laboratory investigations usually reveal a variety of abnormalities including electrolyte imbalance (noticeably hypokalaemia, hypochloraemia, and metabolic alkalosis through repeated vomiting and purging), deranged liver enzymes (especially mildly elevated alkaline phosphatase), hypoglycaemia, mild iron-deficiency anaemia, and leucopaenia. Many of these normalise with weight gain. Radiological imaging of the brain has shown enlarged lateral ventricles and cortical sulci, which correlates with the degree of weight loss, and may not reverse with weight gain.

Endocrine disturbance includes a widespread suppression of the hypothalamic-pituitary-gonadal axis, leading to arrest in secondary sexual development and amenorrhoea. Restoration of weight does not immediately revert the situation, and a prolonged period of amenorrhoea (6 months) is associated with osteopaenia and osteoporosis.¹⁴ The return of menses is variable and achieved only when the weight and body composition has stabilised at its optimum level. An optimum weight can be predicted by serial pelvic ultrasound scans, which are used to track the progress of the ovaries during weight gain. This optimum weight has been shown to be 95% of the "ideal body weight" - which is the weight-for-height ratio in children, and pre-morbid weight in adults.^15-17

There are relatively fewer studies on the effect on linear growth. In patients whose illness had an onset after menarche, linear growth did not appear to be affected, perhaps because these patients had achieved their expected height by the time the illness started. However, for pre-pubertal cases, there was evidence of growth stunting.¹⁸

Psychiatric morbidities

The illness is strongly associated with a range of psychiatric diagnoses, noticeably depression, anxiety and obsessive-compulsive disorder. The onset of these disorders may precede that of anorexia nervosa. The lifetime prevalence of psychiatric comorbidities has been documented to be as high as 80%.

Depressed and irritable affect commonly accompanies weight loss, and tends to improve with weight gain. Nevertheless, concurrent depressive disorders have been found in about 50% of adolescent anorexic patients, while the lifetime prevalence of depressive disorders has been found in as many as 70%.¹⁹ Among adolescent patients, Ivarsson et al found that up to 80% had at least one episode of depressive disorder within 10 years after the onset of anorexia nervosa.²⁰ Depression is closely associated with suicide, which accounts for a significant proportion of deaths associated with anorexia nervosa. The relationship between depression and anorexia nervosa is complex. There is evidence of shared familial causal factors, perhaps genetic in nature.^21,22 It is also possible that caloric deprivation from starvation causes a depressed affect through elevations in corticotrophin-releasing hormone and lowering of 5HT functioning.¹⁹ Whether this exacerbates a pre-existing predisposition or a sole cause is still unclear.

Anxiety disorders are another common co-morbidity, with a lifetime prevalence of 33% to 72%, 2-4 times higher than in community populations.^23,24 However, because anxiety disorders are highly prevalent in the community, and because of variations in study methodology, elevation in rates has not been consistently found. A more robust finding relates to obsessive-compulsive disorder, whose lifetime prevalence is clearly higher among anorexics.

Aetiology

Anorexia nervosa is a multi-factorial illness that involves genetic, psychological, familial and socio-cultural factors, which predispose, precipitate and maintain the illness. It is increasingly being recognised that the mechanisms behind the interactions of these factors have to be understood within a developmental context.

Genetic factors

Evidence from familial aggregation and twin studies has consistently supported a genetic loading,^21,22 although the magnitude is uncertain. The search for the candidate genes is preliminary, with recent studies focusing on serotonin-related genes, because abnormalities in central serotonin function appear to persist after recovery, suggesting that it could be a traits marker.²⁵ Other areas of investigation include the role of hormones that control weight and appetite, such as gastrointestinal hormones and leptin.²⁶ The personality of anorexic patients also shows heritable traits. They are typically described as rigid, controlled, and perfectionistic.^27,28 Cluster C personality traits (avoidant, dependent, obsessive-compulsive traits) have been found to differentiate between the anorectic and their non-affected sisters.²⁴ They are often described as "perfect children" before the onset of the illness.

Psychological factors

The fact that the onset of the disorder peaks at adolescence suggests a developmental context. Bruch described the anorexic to have failed to develop a sense of self and who has "a paralysing sense of ineffectiveness".²⁹ How this has occurred remains hypothetical. However, in keeping with this observation, studies have confirmed that despite being successful in many ways, patients with anorexia nervosa have low self-esteem, poor self-concept and are self-critical.³⁰ One common theme that emerges is the feeling that life is out of control, which is displaced onto a control over eating. Crisp³¹ suggested that the illness reflects the adolescent's fear of growing up, so that it allows the patient to "regress to a simpler existence, without the conflicts of growth, sexuality and personal independence". Our clinical experience concurs with this suggestion.¹¹

Family factors

The speculation about family dynamics in the aetiology has a long history. Famous family therapists such as Minuchin and Palazzoli each described characteristic patterns of interaction in families with an anorexic daughter. Minuchin described enmeshment, over-protection, rigidity, and lack of conflict resolution,³² while Palazzoli highlighted the supreme loyalty and rigidity in family beliefs.³³ It is clear that these are neither specific to the disorder, nor necessarily typical of these families. Even in the presence of family dysfunction the direction of causality cannot be determined, as the evidence is of association between the illness and family interactions. Clinical experience attests to the fact that families can be functioning satisfactorily before the onset of the illness and conflicts have only emerged in an effort to handle the illness. Family factors therefore are best construed as perpetuating factors and not necessarily precipitating factors. Nonetheless, the effect the illness has on family relationships is real and needs to be addressed in the course of treatment.

Socio-cultural factors

The endorsement of a thin body ideal is now widely considered a predisposing factor. Those whose occupation or interests necessitate an emphasis on weight and shape, such as ballet dancers, gymnasts and jockeys are high-risk groups. The spread of the illness from Western to westernising societies is thought to be a result of the globalisation of Western culture and the acceptance of its slim body ideal.³⁴ The increasing concern among non-Western girls of their weight and shape, and their desire to be thin reflect this shift in the emphasis on the physical appearance.

Precipitating event

Against this array of vulnerability factors, the stresses that trigger the onset of eating problem are often varied and non-specific. It may be a chance remark about the girl's weight or shape, or severe trauma such as sexual abuse.³⁵ The essence of these stresses is to further jeopardise a precarious psychological balance. On a psychodynamic level, the adolescent's focus on eating is hypothesized to be her attempt to exert control over life's circumstances.³⁶

A link between child sexual abuse and anorexia nervosa has been documented since the late 1980s. Both clinic samples and community random samples have now confirmed a higher rate of child sexual abuse among women with anorexia nervosa when compared with healthy controls, but the rate is similar when compared with psychiatric controls.^37,38 Child sexual abuse is therefore considered a general risk factor for psychiatric disorders that include, but is not specific to, eating disorders.

Prognosis

Anorexia nervosa is a protracted illness. Outcome data differ according to the parameters studied, such as weight and physical recovery (which includes a return of menses), eating attitudes, mental status, psychosocial functioning and mortality rate. The duration of follow up reveals different recovery rates. In a review of 31 outcome studies involving 941 child and adolescent patients, Steinhausen³⁹ found that an overall 52% had achieved complete recovery, while 29% had improved and 19% remained chronically ill.

The mean crude mortality rate was 2%. The outcome for core symptoms was slightly better: 68% achieved weight normalisation, 64% a return of menstruation and 52% a resumption of normal eating behaviour. Development of binge eating behaviours or frank bulimia nervosa is found in 20-30% of cases, and most often occurs during the first few years of the recovery phase.^40,41 This switch to binge eating appears to be predicted by family relationship problems such as hostility and lack of parental affection for the child.⁴¹ Additional psychiatric disorders were found in a large proportion of patients at follow up, primarily depressive disorders, anxiety disorders and obsessive compulsive disorders.³⁹ Local findings evidenced similar patterns.⁴² Importantly, studies find that when patients have achieved complete recovery from the eating disorder, they have a good chance of overcoming other psychiatric disorders too.^20,40

The search for prognostic indicators consistently found that the outcome continues to improve over many years (although the mortality rate also increases).^39,43 Vigorous treatment and arrangement for long follow-up is recommended. Patients whose illness has an onset in adolescence appear to do better,^40,43 although it is less clear for pre-pubertal cases. Symptoms such as vomiting, purging, bingeing, chronicity and obsessive-compulsive personality trait point to a less favourable outcome.

Management

As the state of thinness is treasured by the patient, the patiant is often reluctant to receive help. The first step in treatment is therefore to help the patient understand that she needs help, and to help the family persevere in the face of strong resistance from the patient. In the initial phase of management, weight restoration is the primary goal. This can be achieved in the out-patient, day-patient or in-patient settings. Indications for hospitalisation include 1) low or rapidly dropping body weight, such as when the weight is below 80% of the ideal body weight, 2) evidence of physical complications, such as electrolyte imbalance and cardiovascular decompensation, 3) persistent self-induced vomiting or laxative abuse, 4) family is unable to supervise the patient's diet or have severe conflicts, 5) presence of psychiatric co-morbidities and /or suicide risk, and 6) failed out-patient treatment. It is not uncommon for patients and/or parents to resist hospitalisation. Empathetic explanations to both the patient and parents about the risks of the illness and the benefits of hospitalisation may help to overcome the resistance. Developing a therapeutic alliance with the family and patient is essential. The use of compulsory admission, though legal, is still controversial and must be carefully deliberated.^44,45

A diet that is normally taken by the patient should be gradually introduced to achieve a weight gain of around 0.5kg per week. Dietary advice is often necessary at this stage, but the patient should not be allowed to dictate the range of food that she is willing to accept, as this will inadvertently put the focus on her eating symptoms once more. Close supervision during and after mealtime is necessary to ensure that food is actually eaten and chance of vomiting minimised. Restriction of daily activities may be necessary to facilitate weight gain. Because of delayed gastric emptying, patients often complain of gastric bloating in the initial stages of re-feeding, and reassurance is needed. A "target weight" should only be used as an indication of the approximate weight at which menses is likely to return, and should not be rigidly set.

When the weight gain has achieved a satisfactory level, the focus of treatment must shift to address the underlying psychological difficulties. Individual exploration with the patient about her experiences to make a link with her eating problem is the beginning of psychological treatment. The meaning of her self-starvation has to be understood. For children and adolescents, family therapy has been shown to be more effective than individual psychotherapy, especially among patients whose illness began before the age of 19 and lasted less than 3 years.^46,47 The premise of family therapy is that family relationships have been altered by the illness and this is perpetuating the illness. No causal relationship is implied. Our experience with Hong Kong Chinese adolescents has revealed four common themes regarding the meaning of self-starvation in the family context: i) self-starvation to gain love and control; ii) coalition of the anorexic daughter with her powerless and helpless mother to fight for a better position in the family by triangulating herself in the parents' marital conflicts; iii) family loyalty, sacrificing oneself for the good of the family, and iv) the anorexic daughter acting as emotional support for her helpless and powerless mother.⁴⁸ For older patients, there is still a paucity of treatment research data, and family therapy has not been found to be more effective than other forms of psychotherapy.

There is little convincing evidence that drug treatment is effective in the management of anorexia nervosa per se. Relevant literature to date consists mostly of case reports and open medication trials. Methodological problems also preclude any consensus conclusions. However, in the presence of psychiatric co-morbidities such as depressive or anxiety disorders, the use of selective serotonin re-uptake inhibitors may be indicated. Atypical antipsychotics such as olanzepine has been reported to be a useful adjunct if the psychopathology takes on a delusional quality.^49,50

It is essential that the approach to management is multidisciplinary. Different professional expertise will be required during the course of treatment. When the physical state is a primary concern, a paediatrician or physician may need to be involved. Subsequently close collaboration involving, in the least, psychiatrist, psychologist and family therapist is appropriate. Good communication is crucial among professionals and between the treatment team and the family, not least because the complex nature of the disorder and the patients' resistance towards treatment need a concerted and coordinated effort on all fronts. The following case description will illustrate the point.

Case example

M, a 13 years old girl, had a weight of 26.5kg (height 146cm, BMI 12.4, 72% of expected weight-for-height) when she was admitted into hospital with a four months history of anorexia nervosa. Despite involvement of a dietician and complete bed rest, her weight remained static throughout the first month. She was resistive, and refused to have any discussion with her attending doctor except to bargain over food intake. She refused to talk to her parents except to demand to be taken out of hospital. She kept a distance from the nursing staff and other patients on the ward, and occupied herself with studying. When she failed to make any noticeable progress, her parents' determination about keeping her in hospital wavered. Several joint meetings between the clinical team and the parents were held to consolidate the treatment alliance. When M's repeated demands to be discharged were firmly refused by her parents as well as her attending doctor, she finally began to externalise her feelings by crying. At this time, her intake improved and her weight began to rise.

However, she was still not talking to anyone. Her preoccupation with studying provided a window for engagement and the Unit's teachers involved her in classroom activities as a means of helping her verbalise her feelings. Family therapy was started when her body weight rose to 85% of ideal body weight. One of the goals was to help M express her feelings in an age-appropriate manner. The onus was on her to discuss her wish to be discharged with her attending doctor. The doctor reinforced the expectation of age-appropriate communication by having M discuss the details with her family. The effect of this timely collaboration between the family therapist and clinical team was clear - M started to talk to her parents and a discharge plan acceptable to all was gradually formulated. She was discharged after four months of in-patient treatment. Her weight at discharge was 37kg, 100% ideal body weight. In the last follow up nearly two years later, she has remained well.

The success of this case was rooted in 1) the clear treatment alliance between the parents and the clinical team, 2) the use of different expertise within a multidisciplinary team, 3) close collaboration between the clinical team and family therapist, and 4) timely intervention strategies. This case clearly illustrates the essence of multidisciplinary approach in the treatment of such a complex disorder.

Conclusion

Anorexia nervosa is no longer confined to the Western world. It now exists in many different cultures and across different social classes. Not only does it affect adolescent girls and young women, but pre-pubertal children may be affected too. Management must encompass a careful assessment of the multitude of contributing factors within a developmental framework. Treatment cannot focus on physical recovery alone. A multidisciplinary approach to address the underlying psychological factors is a pre-requisite for successful treatment.

Key messages

1. Anorexia nervosa is a complex, multi-factorial illness. There is a high rate of co-morbidities with other psychiatric disorders such as depression and obsessive-compulsive disorder.
2. Food refusal is only a symptom - its underlying psychological meaning has to be understood and managed.
3. It is essential to manage anorexia nervosa using a multidisciplinary approach targeting both physical and psychological recovery.
4. Western studies found that family therapy is more effective than individual therapy for children and adolescents with the disorder. Local experience confirms the usefulness of family therapy.
5. Recovery from anorexia nervosa is protracted. Non-recovery is associated with considerable physical and psychological sequelae. Vigorous treatment is warranted.
6. There is as yet no conclusive evidence for the effectiveness of drug therapy in the management of the disorder.