Eating disorders are serious psychiatric illnesses that often present during adolescence and young adulthood. They are associated with medical as well as psychological disturbances, and pediatricians play an important role in their identification, diagnosis, and management. There has been a paucity of treatment research that specifically focuses on children and adolescents with eating disorders. This article reviews the scientific evidence for the use of psychotropic medication in the treatment of children and adolescents with eating disorders.
Eating disorders are prevalent in adolescents and pediatricians play an important role in their identification, diagnosis, and management. Although some pediatricians may choose to refer their patients to specialized eating disorders programs, many, especially those who practice a long distance from major medical centers, may be called on to play a central role in the management of an adolescent with an eating disorder.
One of the most challenging aspects of treating a teenager with an eating disorder is that the wishes of the adolescent, driven by the eating disorder, are often in direct conflict with those of the treatment team and the family. It would be helpful if there were a safe psychopharmacologic agent that could alleviate the preoccupation with shape and weight or distortion in body image that makes treatment so challenging. The psychopharmacologic management of adolescents with eating disorders is further complicated by the high rates of psychiatric comorbidity and the paucity of data regarding the usefulness of psychiatric medications in younger populations. Recent black-box warnings about medications frequently prescribed to treat adolescents with eating disorders further add to the confusion. This article reviews the scientific evidence for use of psychotropic medication in the treatment of children and adolescents with eating disorders and briefly reviews the major types of eating disorders and the comorbidities associated with them. The efficacy and safety of psychotropic medications for the treatment of eating disorders in general are discussed, with a focus on data that are specific to children and adolescents.
Major categories of eating disorders
The spectrum of eating disturbances varies widely, ranging from mildly abnormal eating habits to life-threatening chronic disease. The most commonly used definitions of eating disorders are provided by the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV). These criteria are currently being revised in the soon-to-be-published DSM-5. In DSM-IV, 3 major groups of patients are identified: those who meet criteria for anorexia nervosa (AN), those who meet criteria for bulimia nervosa (BN), and those who have features of an eating disorder but who do not meet the diagnostic criteria for either AN or BN, referred to as an eating disorder not otherwise specified (EDNOS). In children and adolescents, most patients are in the EDNOS category. Changes to the diagnostic criteria being considered for DSM-5, including less-stringent frequency and severity criteria, make it likely that younger patients with weight and eating behavior disturbances will more likely meet DSM-5 criteria for the major eating disorder categories.
The core feature of AN is self-imposed weight loss, or failure to gain weight during a period of growth, leading to a weight that is less than that expected for height and age. In addition, patients have an intense fear of gaining weight, a distorted body image, and, for postmenarcheal girls, loss of at least 3 consecutive menstrual cycles. There are 2 subtypes: restricting type and binge-eating/purging type ( Box 1 ). The lifetime prevalence of AN in the United States and western Europe is reported to be 0.3% to 0.5%. Peak age of onset is 15 to 19 years, and 95% of sufferers develop the disorder before the age of 25 years. The female/male ratio is 9:1 with a higher ratio of boys reported in those less than 14 years old.
Anorexia nervosa
Refusal to maintain body weight at or greater than a minimally normal weight for age and height (eg, weight loss leading to maintenance of body weight less than 85% of that expected; or failure to make expected weight gain during period of growth, leading to body weight less than 85% of the expected)
Intense fear of gaining weight or becoming fat, even though underweight
Disturbance in the way in which one’s body weight or shape is experienced, undue influence of body weight or shape on self-evaluation, or denial of the seriousness of the current low body weight
In postmenarcheal women, amenorrhea, (ie, absence of at least 3 consecutive menstrual cycles). A woman is considered to have amenorrhea if her periods occur only following hormone (eg, estrogen) administration
Specify type
Restricting type: during the current episode of anorexia nervosa, the person has not regularly engaged in binge-eating or purging behavior (ie, self-induced vomiting or the misuse of laxatives, diuretics, or enemas)
Binge-eating/purging type: during the current episode of AN, the person has regularly engaged in binge-eating or purging behavior (ie, self-induced vomiting or the misuse of laxatives, diuretics, or enemas)
Bulimia nervosa
Recurrent episodes of binge eating. An episode of binge eating is characterized by both of the following:
Eating, in a discrete period of time (eg, within any 2-hour period), an amount of food that is definitely larger than most people would eat during a similar period of time and in similar circumstances
A sense of lack of control of eating during the episode (eg, a feeling that one cannot stop eating or control what or how much one is eating)
Recurrent inappropriate compensatory behavior to prevent weight gain, such as self-induced vomiting; misuse of laxatives, diuretics, enemas, or other medications; fasting; or excessive exercise
Binge eating and inappropriate compensatory behaviors both occur on average at least twice a week for 3 months
Self-evaluation is unduly influenced by body shape and weight
The disturbance does not occur exclusively during episodes of AN
BN is characterized by recurrent episodes of binge eating accompanied by episodes of inappropriate compensatory behaviors, such as self-induced vomiting or misuse of laxatives, diuretics, or other medications. These behaviors are accompanied by cognitions that reflect influence of body shape and weight on self-evaluation. Diagnosis of BN requires that the behavioral disturbance not occur in the context of AN; therefore, individuals with BN are generally at normal or above normal weight. The lifetime prevalence of BN has been estimated at 1.5% to 2.0% in adult women. The prevalence of DSM-IV BN is lower in adolescent samples because BN develops in older adolescence and young adulthood. Prevalence of BN was found to be 0.3% in a nationwide study of 12 to 23 year olds in Portugal and 0.75% in a study of girls in Spain. The gender ratio is slightly more balanced in BN than AN, with 20% to 30% of those affected being male. DSM-IV defines BN as requiring that binge and purge episodes occur at least twice weekly for 3 months, but the criteria for BN being proposed for DSM-5 describe that once-weekly episodes for a period of 3 months is sufficient to make the diagnosis. It is likely that adolescents who had previously received the nonspecific EDNOS diagnosis for subthreshold BN will more likely meet full diagnostic criteria for BN using DSM-5.
EDNOS is a heterogeneous diagnostic category that includes patients presenting with symptoms similar to AN or BN but who do not meet all of the diagnostic criteria for those disorders. Included in this group are those who have started weight loss, but may not have reached markedly low weight for height, those who have low weight but have not lost menstrual activity, or those with binge and purge behavior but with frequency less than twice weekly or duration less than 3 months, or those who purge but do not binge, to maintain a weight they perceive to be thin and healthy. EDNOS also includes some other conditions such as binge-eating disorder (BED) and night-eating syndrome (NES).
In BED, patients binge eat but do not engage in compensatory behaviors such as purging. In DSM-IV, BED is categorized within the EDNOS umbrella, although it is being proposed as a formal diagnostic category for DSM-5. Among adults, BED is the most common eating disorder, with lifetime prevalence estimates in the community of 3.5% among women and 2.0% among men. BED is considerably more prevalent in adults than it is in children and adolescents, although it has been described in younger groups, especially because overweight and obesity are increasingly present in pediatric samples. BED is more difficult to assess in children because of the challenges inherent in self-reported dietary recall in younger samples, as well as the difficulty children may have in understanding the concept of loss of control during eating episodes. Developmentally appropriate language may be necessary to assess accurately the concept of loss-of-control eating. Up to 30% of overweight children and adolescents report episodes of loss of control during eating episodes. Binge eating equally affects boys and girls, although girls may be more willing to report disordered eating than boys.
NES is characterized by morning anorexia, evening hyperphagia, and insomnia with awakenings followed by nocturnal ingestions believed to be explained by a delay in the usual circadian timing of food intake relative to the sleep-wake cycle. The female athlete triad is a syndrome occurring in female athletes consisting of 3 interrelated conditions: low energy availability, menstrual dysfunction, and low bone mineral density. Although some patients with the triad have an eating disorder, many do not, but are consuming inadequate calories for their activity level. The inadequate energy consumption may be caused by body image concerns, but may also be the result of a desire to improve athletic performance. Some athletes may have a lack of knowledge regarding the true metabolic demands of their athletic training. Inadequate energy intake results in an energy imbalance, leading to suppression of the hypothalamic-pituitary-ovarian axis. The athlete develops menstrual irregularity, amenorrhea, and a low estrogen state that contributes to low bone mass and increased fracture risk.
In children and younger adolescents, there are some other distinct clinical conditions related to disturbances with eating that are sometimes grouped in the EDNOS classification. The Great Ormond Street Children’s Hospital Classification includes eating disturbances in children and young adolescents in which there is no distortion in body image or preoccupation with shape or weight. Some of these conditions may be accompanied by loss of weight and growth retardation ( Box 2 ). DSM-5 is considering the inclusion of a new category of eating disorder called avoidant/restrictive food intake disorder (ARFID), which describes many of these eating-related disturbances ( Box 3 ).
Food avoidance emotional disorder (FAED)
Food avoidance or difficulty
Weight loss
Mood disturbance
No cognitive distortions regarding weight or shape
No organic brain disease, psychosis, or drug-related cause
Selective eating disorder (SED)
Limited food choices for at least 2 years
Unwilling to try new foods
No cognitive distortions regarding weight or shape
No fear of choking or vomiting
Weight and height are usually appropriate for age
Functional dysphagia
Food avoidance
Fear of choking or vomiting
Often there is a history of an episode of choking
No cognitive distortions regarding weight or shape
No organic brain disease or psychosis
Pervasive food refusal
A refusal to eat, drink, walk, talk, or care for self
Resistant to others’ efforts to help
Anorexia nervosa
Bulimia nervosa
ARFID
Eating or feeding disturbance as manifested by persistent failure to meet appropriate nutritional and/or energy needs associated with 1 or more of the following:
- 1.
Significant weight loss (or failure to achieve expected weight gain or faltering growth in children)
- 2.
Significant nutritional deficiency
- 3.
Dependence on enteral feeding
- 4.
Marked interference with psychosocial functioning
- 1.
There is no evidence that lack of available food, an associated culturally sanctioned practice, or a general medical condition or other mental disorder alone is sufficient to account for the disorder
The eating disturbance does not occur exclusively during the course of AN or BN
If the eating disturbance occurs exclusively in the context of another mental disorder (eg, mental retardation or a pervasive developmental disorder), it is sufficiently severe to warrant independent clinical attention
Comorbidity
Patients with eating disorders have high rates of psychiatric comorbidity, either concurrent with the eating disorder, or manifesting earlier or later. For AN, the lifetime prevalence of affective disorders, especially depression, is 50% to 68% and the lifetime rate of anxiety disorders (especially obsessive-compulsive disorder and social phobia) is 30% to 65%. These symptoms are worsened by malnutrition, and weight restoration will contribute to improvement in mood and anxiety symptoms, in addition to improving the core eating disorder. Anxiety disorders often predate the onset of AN, and depression and anxiety may persist after recovery from the eating disorder. For BN, approximately 80% of patients report a lifetime prevalence of psychiatric comorbidity with high rates of mood disorders (50%–70%), anxiety disorders (13%–65%), substance abuse disorders (approximately 25%), and personality disorders (20%–80%). Patients with BN tend to be impulsive and may engage in high-risk activities such as sexual promiscuity, stealing, and self-injury. An anxiety disorder may predate the onset of BN, and substance abuse disorder frequently follows its onset.
Comorbidity
Patients with eating disorders have high rates of psychiatric comorbidity, either concurrent with the eating disorder, or manifesting earlier or later. For AN, the lifetime prevalence of affective disorders, especially depression, is 50% to 68% and the lifetime rate of anxiety disorders (especially obsessive-compulsive disorder and social phobia) is 30% to 65%. These symptoms are worsened by malnutrition, and weight restoration will contribute to improvement in mood and anxiety symptoms, in addition to improving the core eating disorder. Anxiety disorders often predate the onset of AN, and depression and anxiety may persist after recovery from the eating disorder. For BN, approximately 80% of patients report a lifetime prevalence of psychiatric comorbidity with high rates of mood disorders (50%–70%), anxiety disorders (13%–65%), substance abuse disorders (approximately 25%), and personality disorders (20%–80%). Patients with BN tend to be impulsive and may engage in high-risk activities such as sexual promiscuity, stealing, and self-injury. An anxiety disorder may predate the onset of BN, and substance abuse disorder frequently follows its onset.
Pharmacologic treatment of AN
The difficulties in conducting randomized controlled trials (RCTs) in AN have recently been reviewed. Difficulties with subject recruitment, accompanied by high dropout rates, result in small sample sizes that limit statistical power and make interpretation of existing studies difficult. Medication treatments may be particularly difficult for AN patients to accept compared with psychological interventions. Furthermore, although some RCTs of medication treatments have been conducted in adults with AN, there have been no published RCTs specifically conducted in children and adolescents.
Antidepressants and AN
Patients with AN often exhibit symptoms of depression and early studies examined the use of antidepressants in this disorder. There have been 4 published RCTs on the use of antidepressants in adults with AN. Lacey and Crisp conducted an RCT on the use of clomipramine, a tricyclic agent with antidepressant and antiobsessional properties, in 16 adult patients with AN who were hospitalized for refeeding. Clomipramine was associated with increased hunger and appetite but a reduced rate of weight gain. Halmi and colleagues conducted an RCT on the use of the tricyclic antidepressant amitriptyline, the appetite stimulant cyproheptadine, and placebo in 72 subjects with AN, and found that amitriptyline did not increase the rate of weight gain or improve depression in patients with AN. Cyproheptadine increased rate of weight gain and reduced depressive symptoms in those who were not purging, but impaired treatment efficacy for those who were purging. In a 5-week double-blind RCT conducted by Biederman and colleagues, 11 subjects were treated with amitriptyline and 14 received placebo. Amitriptyline was associated with significant side effects and did not improve weight gain, eating disorder symptoms, or symptoms of depression. Tricyclic antidepressants are associated with several cardiac side effects including tachycardia, hypotension, and cardiac arrhythmias, in particular prolongation of the QT interval, and should only be used with extreme caution in malnourished patients with AN.
The selective serotonin reuptake inhibitors (SSRIs) have a better safety profile than the tricyclic antidepressants and also have proven efficacy for obsessive-compulsive disorder in children and adolescents. The findings of altered levels of serotonin metabolites in the cerebrospinal fluid of low-weight and weight-restored subjects with AN suggest alterations in the serotonergic pathways in this disorder. The SSRIs increase serotonin levels available to the postsynaptic receptor and, on a theoretic basis, might be useful as an adjunct in the treatment of AN. Attia and colleagues randomized 31 adult women with AN, who were hospitalized on an inpatient clinical research unit, to receive 7 weeks of fluoxetine at a target daily dose of 60 mg, or placebo in a double-blind fashion. However, although both groups improved in weight and psychological symptoms, there was no additional benefit associated with fluoxetine compared with placebo. A retrospective study of adolescents with AN, who were assessed during their inpatient stay and at 3 and 6 months following hospitalization, compared 19 subjects treated with an SSRI during inpatient and/or outpatient periods with 13 who did not receive an SSRI, and found no difference between the 2 groups in body mass index (BMI), eating disorder symptoms, or depressive symptoms at discharge from the inpatient program or at follow-up 6 months after discharge.
Two studies have examined the use of fluoxetine in relapse prevention in adults with AN. After an inpatient stay intended to restore weight, Kaye and colleagues assigned 16 subjects to receive fluoxetine and 19 to receive placebo in a randomized double-blind fashion. Subjects were then followed for a year without controlling for other treatments the subjects may have received. Compared with those receiving placebo, the subjects on fluoxetine had improved weight gain and a reduction in core eating disorder symptoms and symptoms of depression and anxiety, suggesting that fluoxetine may play a role in relapse prevention in AN. Walsh and colleagues were not able to replicate these findings in a large, carefully designed multisite study of 96 subjects with AN from the New York State Psychiatric Institute and Toronto General Hospital. Subjects were weight restored in inpatient or partial hospital programs and then randomly assigned to fluoxetine or placebo, together with cognitive behavioral therapy (CBT), and followed as outpatients for 1 year in a double-blind manner. Time to relapse, BMI, and clinical measures of depression at follow-up did not differ between the groups. Moreover, there was no difference in response to fluoxetine between clinical sites, suggesting that the findings of the study can be generalized.
Atypical Antipsychotic Medication and AN
Isolated case reports, retrospective studies, and some uncontrolled case series have generated considerable interest in the possible role of the atypical antipsychotic medications in the management of AN. Atypical antipsychotic medications block both the serotonergic and dopaminergic receptors and reduce anxiety, agitation, depression, and obsessional thinking in clinical populations with psychotic disorders for whom these medications are generally prescribed. An added potential advantage in the management of patients with AN is that, both in laboratory animals and in humans, the atypical antipsychotic medications are associated with weight gain. Although most of the studies on the atypical antipsychotics have been conducted in adults, both olanzapine and quetiapine have been used safely in children and adolescents with AN. The atypical antipsychotic most studied is olanzapine, but other atypicals with published reports include quetiapine and risperidone. In a series of 5 hospitalized children with AN described by Boachie and colleagues, administration of olanzapine was associated with significant increases in rate of weight gain and a reduction in agitation and anxiety before and after meals. In Mehler and colleagues’ series of 5 adolescents with AN treated with olanzapine, preoccupation with body image and anxiety related to gaining weight also improved but, in contrast to the findings of Boachie and colleagues, the rate of weight gain in adolescents treated with olanzapine did not exceed the rate of weight gain before initiating treatment with olanzapine. Uncontrolled open-label trials in larger numbers of subjects with AN have documented more significant weight gain in short periods of time in adults, as well as in adolescents. In the adolescent study, 13 subjects aged 9.6 to 16.3 years (mean age 13.7 ± 2.3 years) were treated with olanzapine for 6 months. BMI increased significantly after 1 month of treatment and continued to increase until the end of the study. Eating attitudes and global functioning scores improved, and there was a reduction in anxiety, hyperactivity, and symptoms of depression. However, none of these studies included a control group, and it is not clear how much of the response to treatment was secondary to the multidisciplinary and multimodal treatment that subjects were receiving.
There have been 2 published placebo-controlled RCTs examining the use of olanzapine in AN, both conducted in adults. One compared olanzapine with placebo, and both groups received CBT. The other compared olanzapine with placebo in patients who were attending a specialized eating disorders day program. Both studies found that treatment as usual resulted in increases in BMI as well as improvement in eating disorder symptoms, obsessionality, and depression. Both studies also found that olanzapine improved obsessionality more than standard treatment plus placebo. Brambilla and colleagues’ study of 30 subjects, 15 of whom were treated with olanzapine and 15 who received placebo, found no difference in the rate of change of BMI between the 2 groups. Bissada and colleagues’ study of 34 subjects, 16 assigned to olanzapine and 18 assigned to placebo, found that treatment with olanzapine resulted in greater and more rapid weight gain than placebo. The Bissada and colleagues used a higher target dose of olanzapine (10 mg) than did Brambilla and colleagues (5 mg). Although we are aware of ongoing studies examining the use of olanzapine in children and adolescents with eating disorders, to our knowledge, there have been no published RCTs on the use of olanzapine in children and adolescents with AN.
Quetiapine is an atypical antipsychotic that is less likely to cause dramatic weight gain and may be more acceptable to patents with AN who fear gaining weight. In 2 open-label studies, adults treated with quetiapine showed improvement in anxiety, depression, and eating disorders scores, but weight gain was modest. Quetiapine has been used safely in a series of 3 treatment-resistant children and adolescents with AN. The 3 subjects, aged 11, 14, and 15 years reported improvement in the delusional quality of their body image distortion when treated with quetiapine. As with olanzapine, we are not aware of any RCTs on the use of quetiapine for children and adolescents with AN.
There have been 3 published case reports on the use of risperidone in adolescents with AN. The first described a young girl who was diagnosed with an autistic disorder at age 4 years and then at age 13 years developed AN. Treatment with risperidone was associated with a decrease in agitation and aggression, lessening of her rigidity, and loss of paranoid ideation. The 2 cases described by Newman-Toker were adolescents, aged 12 and 19 years, with restrictive AN but no other psychopathologic disorder. Within a week of adding risperidone to an SSRI, both patients reported improvement in anxiety and obsessions about food and both patients began to gain weight. Both patients believed that the risperidone contributed to the improvement of their cognitive distortions around eating and allowed them to feel more in control of their thinking.
Common side effects of the atypical antipsychotic medications include somnolence and dry mouth, but more serious side effects include metabolic abnormalities such as insulin resistance, hyperlipidemia and diabetes, and the emergence of extrapyramidal symptoms, the most serious of which is tardive dyskinesia. Further study is needed to determine whether the side effect profile for this class of medications is the same among individuals with AN as it is in other clinical populations.
In summary, none of the RCTs conducted in adults with AN found significant benefit associated with antidepressant medication, either in promoting weight gain or in alleviating the core eating disorder symptoms. It has been postulated that, with malnutrition, depletion of tryptophan, a precursor of serotonin, may interfere with the action of antidepressants, and in particular the SSRIs. However, in the malnourished state, even after supplementation with tryptophan, fluoxetine has not been shown to improve weight gain or eating disorder symptoms. In addition, even after partial weight restoration and nutritional rehabilitation, the SSRIs have not proved to prevent relapse. Case reports and open-label studies suggest that olanzapine may be useful in reducing anxiety related to eating, obsessive thinking, depression, and some eating disorder symptoms, but the effect on weight again, beyond that achieved by standard treatment, is not clear. A recent systematic review concluded that, based on the results of existing studies, “there is insufficient evidence to confirm that the atypical antidepressants enhance weight gain in AN.” There may be a role for psychopharmacology in individual patients with AN, for example, in those with overwhelming anxiety or obsessionality, or in those with comorbid depression or obsessive-compulsive disorder, especially when the depressive and obsessional symptoms predated the onset of the eating disorder, but the sample sizes in the existing studies have been too small to identify those patients most likely to benefit from the addition of psychotropic medication. At present, the results regarding the role of the atypical antipsychotics remain inconclusive. What is clear is that psychopharmacology should not be the first line of treatment of AN.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
