Eating disorders in children: a review of clinical guidelines in Annals of Internal Medicine
Julie O’Toole, a Board certified pediatrician, has treated children, adolescents and young adults with eating disorders in a sub-specialty practice since 1998. She is the founder and medical director of Kartini Clinic, a pediatric clinic with inpatient, day treatment and outpatient care for children suffering from all conditions of disordered eating. I asked her to review the Annals of Internal Medicine In the Clinic feature on eating disorders. (She is also my aunt.)
This review article has many strong points and a few weak ones, particularly with regard to the pediatric population.
Risk. Females are at higher risk. In addition, the article highlights the importance of the family history. A first degree relative with anorexia nervosa confers a ten-fold increase in risk.
Weak points in the article’s discussion of screening are the recommendations for BMI screening. BMI can be deceptive. A child can start with a high BMI, lose a significant amount of weight, and wind up at a BMI that is still higher than average or well within the range of “normal,” falsely reassuring clinician and family.
Prevention. The authors dispel the illusion that anorexia nervosa, whose biophysiology hasn’t been elucidated and whose genetics is yet poorly defined, can be spoken of as being prevented. I agree however that early identification is critical, especially in children whose brains, bones and bodies are still being formed.
Diagnosis. This issue receives a great deal of attention. The new DSM-5 will bring substantial changes.
The older DSM-IV criteria are detrimental in pediatrics, where practitioners are tempted to either wait until age inappropriate DSM thresholds have been crossed or to use the unhelpful categorization “eating disorder not otherwise specified,” with insurance and triage repercussions. The most clinically useful suggestion the authors make, relevant to children, is to bypass psychological criteria (e.g. statements about being fat or being afraid to gain weight) in favor of “repeated failure to gain weight despite recommendation.” In childhood failure to gain weight can be as important as frank weight loss.
The physical exam, as the authors rightly emphasize, is more important than labs. Normal laboratory studies are little consolation to a clinician who has a cold, bradycardic, syncopal, cachetic patient. Not mentioned, however, is anosognosia or denial of the seriousness or even existence of the illness. The anosognosia of anorexia nervosa interferes more with successful treatment than medical compromise.
The American Academy of Pediatrics has hospitalization guidelines which have proven a boon to families who must access their insurance benefits for treatment of an eating disorder.
Where to treat. Anorexia nervosa has a mortality rate between 5-20%, with 10% a commonly given figure. This is comparable to the mortality from some types of childhood leukemia. Despite this, some practitioners are still reluctant to refer to specialists. Though the family may prefer that the generalist, whom they know and trust, manages their child’s eating disorder, this is not wise. These patients need time, which is lacking in a general practice! Even more than time, they need expertise and consistency of follow-up.
How to treat. Both cognitive behavioral therapy (CBT) and interpersonal therapy (IPT) have proven useful in the treatment of bulimia nervosa, but as the authors point out, there is little evidence for a robust effect of CBT or IPT in anorexia nervosa. In fact, a recommendation implied by the authors comes from pediatric eating disorder treatment: without weight restoration you will get nothing. Please do not follow a patient for more than a few weeks if they are not gaining weight consistently. At risk for the patient are the bones, the reproductive system, the brain, and meaningful social interaction.
Medication. Two very important points were made about medication, which should be shouted from the rafters: exogenous hormones do not help protect bones, and SSRI’s do not help in anorexia nervosa. Do not start the birth control pill to “jump start” periods, it doesn’t help their bones, and all it “jump starts” is cyclical withdrawal bleeding while imparting a false sense of normalcy.
Atypical antipsychotics have shown some use in anorexia nervosa, improving weight and cognition.
Pediatric treatment. A recommendation for family based interventions (FBT) is one I endorse. However, I am puzzled by the statement that a nutritionist should be included in any multidisciplinary team. Not only do we not use a nutritionist on our multi-disciplinary family-based treatment team, but the most widely known FBT programs in the country (Stanford University and the University of Chicago) do not use the services of a nutritionist either. The essential players in family-based interventions are the parents. Nutritional guidance can be provided in a family-based setting by specialist physicians and therapists working in collaboration with them.
Follow-up. The authors recommend frequent follow-up for any weight change. Many patients return to their home practice in good condition after a period of intense, often residential, treatment with everyone falsely reassured. Follow-up at monthly intervals is a disaster waiting to happen. Eating disorders in general and anorexia nervosa in particular are chronic illnesses; anosognosia means that external supervision and support is likely to be needed for an extended period of time.
