This study will examine the effectiveness of inpatient treatment for anorexia nervosa versus a sequenced treatment that transitions from higher intensity inpatient care to lower intensity outpatient care.

Anorexia nervosa is an eating disorder in which a person does not eat enough to maintain a healthy body weight. This disorder can lead to starvation, which may, in turn, lead to heart failure or brain damage. Anorexia can also lead to a host of other health problems, including mild anemia, swollen joints, and brittle bones. Treatment for anorexia begins by attempting to bring the anorexic person to a healthy body weight, while also conducting therapy aimed at developing habits and ways of thinking that promote maintenance of a healthy weight. Traditionally, the first part of this treatment-restoring healthy body weight-has taken place in inpatient hospital care. Recently, however, some patients have begun with high intensity inpatient care and then shifted to lower intensity outpatient care over the course of treatment. This study will compare the effectiveness of these two different methods of delivering treatment.

Length of participation in this study will vary, depending on the effectiveness of treatment. All participants will have follow-up assessments for 18 months after study entry. Every participant who passes the screening examination will be admitted to inpatient care in the study hospital. On Day 3 of their stay, participants will be randomly assigned to receive either continued inpatient treatment or sequenced treatment.

Those who receive continued inpatient treatment will stay in the hospital until they reach a healthy body weight and maintain it for 10 to 14 days. The average length of inpatient treatment will be 2 to 3 months. Those who receive sequenced treatment will transition to day patient status once they have met the following health benchmarks: an increase in body mass index (BMI) greater than 15.5 kg/m2; normal electrocardiograms, lab test results, and vital signs; eating all meals for more than 3 days; no weight loss for at least 1 week or weight gain for at least 1 week if the weight gain protocol has begun; and moderate improvement on a clinical global impression (CGI) rating of eating disorder behavior. If participants do not meet these benchmarks and are not at medical risk or suicidal, they will be automatically transitioned to day patient care after 28 days. Once participants reach the day patient status, they will be in the hospital Monday through Friday from 8:00 a.m. until 6:00 p.m. If these participants then reach and maintain a healthy body weight for 10 to 14 days, they will be transitioned to outpatient care, during which they will receive 50-minute therapy sessions once or twice weekly.

The aims and methods of treating anorexia will be the same for both groups. All participants will be prescribed meals, a snack, and dietary supplements in order to reach a healthy weight. The amount of exercise they engage in will be limited to ensure weight gain. Participants will also attend individual, family, and group therapy sessions aimed at supporting healthy thoughts and beliefs about eating behavior. Measurements of weight gain will be made every Monday, Wednesday, and Friday morning. Every 3 months participants will undergo blood tests and assessments of eating behaviors, associated psychopathology, and health service utilization. Clinical interviews will be conducted twice: once at study entry and once after 18 months.

Inclusion Criteria:

• Diagnosis of anorexia nervosa, according to DSM-IV criteria
• Body mass index (BMI) less than 18.5 kg/m2

Exclusion Criteria:

• Presence of psychosis or the axis I disorders of schizophrenia, bipolar disorder, or active substance abuse or dependence requiring specialized treatment
• Requires specialized treatment that is for major depression or anxiety disorders and is inconsistent with study treatments
• Any major medical illness requiring immediate medical treatment or that would affect clinical treatment associated with the study
• Significant risk of suicide