Anorexia Nervosa, Eating Disorders
Conditions
Keywords
Depression
Brief summary
This study will evaluate the effectiveness of fluoxetine versus placebo in reducing the rate of relapse of anorexia nervosa (AN) and enhancing the psychosocial and behavioral recovery of people who have been treated for AN.
Detailed description
Anorexia nervosa (AN), a type of eating disorder, is a serious psychiatric illness that is characterized by an extreme loss of appetite. People with AN view themselves as overweight and cannot bring themselves to eat, even though most are dangerously thin. Signs of the disorder include unusual eating habits, such as avoiding food and meals, picking out a few foods and eating them in small quantities, or carefully weighing and portioning food. Some people with AN fully recover after a single episode, some have a fluctuating pattern of weight gain and relapse, and others experience a chronic course of illness over many years. Effective drugs to treat the disorder are lacking. In addition, most past research has examined the effect of medications during the initial phase of treatment, a time when AN patients may not respond to medication because of the acute effects of starvation. Fluoxetine is a selective serotonin reuptake inhibitor (SSRI) that is commonly used to treat depression. This study will evaluate the effectiveness of fluoxetine versus placebo in reducing the rate of relapse of AN and enhancing the psychosocial and behavioral recovery of women who have already been treated for AN. Participants in this double-blind study will be recruited immediately following completion of a treatment program for AN, in which they maintained a body mass index (BMI) of at least 19 kg/m2 for two weeks. Upon study entry, participants will be randomly assigned to receive either fluoxetine or placebo for 12 months. Participants will begin receiving medication one week prior to discharge from the hospital in which they received care for AN. Medication doses will be increased up to a target dose of 60 mg per day, and will not exceed 80 mg per day. Participants will receive 50 sessions of cognitive-behavioral therapy, lasting approximately 45 minutes each and occurring twice weekly for the first month following discharge from the hospital. After the first month, therapy sessions will occur once weekly until Month 9 and then every other week until Month 12. Participants will also report to the study site to meet with a psychiatrist once a week for the first month following discharge and then every other week for the remainder of the study. General medical status, evidence of AN relapse, medication dose, and side effects will be assessed at these visits. Upon completing treatment, follow-up telephone calls will occur at Months 15 and 21, and follow-up visits will be held at Months 18 and 24. Psychopathology associated with AN, including concern with weight and shape, depressive symptoms, anxiety, and obsessive behavior, will be assessed.
Interventions
DRUGFluoxetine
DRUGPlacebo
Sponsors
National Institute of Mental Health (NIMH)
New York State Psychiatric Institute
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
DOUBLE (Subject, Investigator)
Eligibility
Sex/Gender
FEMALE
Age
16 Years to 45 Years
Healthy volunteers
No
Inclusion criteria
* Meets DSM-IV criteria for anorexia nervosa (except the requirement for amenorrhea) * Successfully completed treatment at one of the study sites in an inpatient or day-program setting immediately prior to study entry (BMI remained at least 19 kg/m2 for two weeks)
Exclusion criteria
* Currently taking any medications other than occasional lorazepam or zopiclone for anxiety or sleep disturbance * Previous serious adverse reactions to fluoxetine (e.g., allergy) * Currently at risk for suicide * Any medical condition requiring treatment with other psychotropic medication (except the occasional use of anti-anxiety medication) * Pregnant * Any serious medical illness besides the eating disorder * History of continuous illness (at a low weight with no periods of remission or return to normal functioning for more than 15 years)
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Proportion of Patients Remaining in Study at 1 Year | 12 months | The primary outcome measure was the proportion of patients with AN successfully completing 1 year of treatment and maintaining \> 85% Ideal Body Weight. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Anxiety Inventory (BAI) | 12 months | The Beck Anxiety Inventory is a 21 question self-report measure of anxiety symptoms during the past week. Possible scores range from 0 - 63, with higher scores indicating more severe symptoms. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Depression Inventory (BDI) | 12 months | The Beck Depression Inventory-II is a 21 question self-report measure of depressive symptoms. Possible scores range from 0 - 63, with higher scores indicating more severe symptoms.Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Rosenberg Self-Esteem Scale (RSES). | 12 months | The RSES is a 10 item self-report measure of self-esteem. Possible scores range from 0 - 30, with lower scores indicating more severe symptoms. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q). | 12 months | The Q-LES-Q is a 93 item self-report measure of enjoyment and satisfaction experienced by individuals in various areas of daily functioning. Each of the 93 items is scored on a five-point scale, and the total score is converted to a percentage of the maximum score possible. The range is therefore from 0 to 100, with a higher score indicating greater enjoyment or satisfaction. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change in Weight Per Month During Treatment | 12 months | — |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Bulimia Subscale. | 12 months | The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Bulimia subscale is comprised of seven items indicating the tendency towards episodes of uncontrollable overeating (binge eating). Possible scores range from 0 to 21, with higher scores indicating greater tendency. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Body Dissatisfaction Subscale. | 12 months | The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Body Dissatisfaction subscale is comprised of nine items indicating the belief that parts of the body are too large. Possible scores range from 0 to 27, with higher scores indicating greater dissatisfaction. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Perfectionism Subscale. | 12 months | The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Perfectionism subscale is comprised of six items Indicating excessive personal expectations for superior achievement. Possible scores range from 0 to 18, with higher scores indicating greater expectations. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Yale Brown Cornell Obsessive Compulsive Scale for Eating Disorders (YBC-EDS) | 12 months | The YBC-EDS is an eight item, clinician-rated instrument assessing eating related preoccupations and/or rituals. Possible scores range from 0 to 32, with higher scores indicating greater preoccupations. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
| Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Drive for Thinness Subscale. | 12 months | The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Drive for Thinness subscale is comprised of seven items indicating excessive concern with dieting, preoccupation with weight and entrenchment in an extreme pursuit of thinness. Possible scores range from 0 to 21, with higher scores indicating greater Drive for Thinness. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Fluoxetine fluoxetine up to 80 mg per day
Fluoxetine | 49 |
| Placebo Placebo group | 44 |
| Total | 93 |
Baseline characteristics
| Characteristic | Fluoxetine | Placebo | Total |
|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical >=65 years | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical Between 18 and 65 years | 49 Participants | 44 Participants | 93 Participants |
| Age, Continuous | 22.4 years STANDARD_DEVIATION 4.46 | 24.2 years STANDARD_DEVIATION 4.52 | 23 years STANDARD_DEVIATION 4.6 |
| Region of Enrollment Canada | 23 participants | 25 participants | 48 participants |
| Region of Enrollment United States | 26 participants | 19 participants | 45 participants |
| Sex: Female, Male Female | 49 Participants | 44 Participants | 93 Participants |
| Sex: Female, Male Male | 0 Participants | 0 Participants | 0 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | — / — | — / — |
| other Total, other adverse events | 0 / 49 | 0 / 44 |
| serious Total, serious adverse events | 1 / 49 | 0 / 44 |
Outcome results
Primary
Proportion of Patients Remaining in Study at 1 Year
The primary outcome measure was the proportion of patients with AN successfully completing 1 year of treatment and maintaining \> 85% Ideal Body Weight.
Time frame: 12 months
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Fluoxetine | Proportion of Patients Remaining in Study at 1 Year | 0.265 proportion of participants |
| Placebo | Proportion of Patients Remaining in Study at 1 Year | 0.315 proportion of participants |
Comparison: The proportion of patients successfully completing the trial in the fluoxetine and placebo groups was compared using the chi-squared statistic.p-value: 0.57Chi-squared
Secondary
Change in Weight Per Month During Treatment
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change in Weight Per Month During Treatment | -1.94 kg per month | Standard Error 0.51 |
| Placebo | Change in Weight Per Month During Treatment | -2.14 kg per month | Standard Error 0.33 |
Comparison: Random effects regression analysis of change during treatment in body weight, in fluoxetine vs placebo groups.p-value: 0.75Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Anxiety Inventory (BAI)
The Beck Anxiety Inventory is a 21 question self-report measure of anxiety symptoms during the past week. Possible scores range from 0 - 63, with higher scores indicating more severe symptoms. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Anxiety Inventory (BAI) | -0.70 units on a scale | Standard Error 0.12 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Anxiety Inventory (BAI) | -0.22 units on a scale | Standard Error 0.13 |
Comparison: Random effects regression analysis of change during treatment of BDI in fluoxetine versus placebo groupsp-value: 0.007Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Depression Inventory (BDI)
The Beck Depression Inventory-II is a 21 question self-report measure of depressive symptoms. Possible scores range from 0 - 63, with higher scores indicating more severe symptoms.Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Depression Inventory (BDI) | 0.12 units on a scale | Standard Error 0.25 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With Beck Depression Inventory (BDI) | 0.20 units on a scale | Standard Error 0.17 |
Comparison: Random effects regression analysis of change during treatment in BDI, in fluoxetine vs placebo groups.p-value: 0.79Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Body Dissatisfaction Subscale.
The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Body Dissatisfaction subscale is comprised of nine items indicating the belief that parts of the body are too large. Possible scores range from 0 to 27, with higher scores indicating greater dissatisfaction. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Body Dissatisfaction Subscale. | -0.24 units on a scale | Standard Error 0.09 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Body Dissatisfaction Subscale. | -0.26 units on a scale | Standard Error 0.09 |
Comparison: Random effects regression analysis of change during treatment on Body Dissatisfaction subscale, in fluoxetine vs placebo groups.p-value: 0.86Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Bulimia Subscale.
The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Bulimia subscale is comprised of seven items indicating the tendency towards episodes of uncontrollable overeating (binge eating). Possible scores range from 0 to 21, with higher scores indicating greater tendency. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Bulimia Subscale. | -0.11 units on a scale | Standard Error 0.07 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Bulimia Subscale. | 0.035 units on a scale | Standard Error 0.08 |
Comparison: Random effects regression analysis of change during treatment in Bulimia subscale, in fluoxetine vs placebo groups.p-value: 0.46Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Drive for Thinness Subscale.
The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Drive for Thinness subscale is comprised of seven items indicating excessive concern with dieting, preoccupation with weight and entrenchment in an extreme pursuit of thinness. Possible scores range from 0 to 21, with higher scores indicating greater Drive for Thinness. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Drive for Thinness Subscale. | -0.24 units on a scale | Standard Error 0.09 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Drive for Thinness Subscale. | -0.81 units on a scale | Standard Error 0.09 |
Comparison: Random effects regression analysis of change during treatment in Drive for Thinness subscale, in fluoxetine vs placebo groups.p-value: 0.19Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Perfectionism Subscale.
The EDI is a 64 item self-report measure of psychological and behavioral characteristics of eating disorders. The Perfectionism subscale is comprised of six items Indicating excessive personal expectations for superior achievement. Possible scores range from 0 to 18, with higher scores indicating greater expectations. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Perfectionism Subscale. | -0.037 units on a scale | Standard Error 0.03 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Eating Disorders Inventory (EDI), Perfectionism Subscale. | 0.05 units on a scale | Standard Error 0.06 |
Comparison: Random effects regression analysis of change during treatment on Perfectionism subscale, in fluoxetine vs placebo groups.p-value: 0.25Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q).
The Q-LES-Q is a 93 item self-report measure of enjoyment and satisfaction experienced by individuals in various areas of daily functioning. Each of the 93 items is scored on a five-point scale, and the total score is converted to a percentage of the maximum score possible. The range is therefore from 0 to 100, with a higher score indicating greater enjoyment or satisfaction. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q). | 0.23 percentage of maximum possible score | Standard Error 0.23 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q). | 0.31 percentage of maximum possible score | Standard Error 0.18 |
Comparison: Random effects regression analysis of change during treatment in Q-LES-Q, in fluoxetine vs placebo groups.p-value: 0.78Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Rosenberg Self-Esteem Scale (RSES).
The RSES is a 10 item self-report measure of self-esteem. Possible scores range from 0 - 30, with lower scores indicating more severe symptoms. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Rosenberg Self-Esteem Scale (RSES). | 0.12 units on a scale | Standard Error 0.08 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Rosenberg Self-Esteem Scale (RSES). | 0.07 units on a scale | Standard Error 0.08 |
Comparison: Random effects regression analysis of change during treatment in RSES, in fluoxetine vs placebo groups.p-value: 0.69Mixed Models Analysis
Secondary
Change Per Month in Psychological Symptoms During Treatment, Assessed With the Yale Brown Cornell Obsessive Compulsive Scale for Eating Disorders (YBC-EDS)
The YBC-EDS is an eight item, clinician-rated instrument assessing eating related preoccupations and/or rituals. Possible scores range from 0 to 32, with higher scores indicating greater preoccupations. Random effects regression models were used to compare fluoxetine vs placebo groups over time, using data from all patients.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Fluoxetine | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Yale Brown Cornell Obsessive Compulsive Scale for Eating Disorders (YBC-EDS) | -0.18 units on a scale | Standard Error 0.13 |
| Placebo | Change Per Month in Psychological Symptoms During Treatment, Assessed With the Yale Brown Cornell Obsessive Compulsive Scale for Eating Disorders (YBC-EDS) | 0.028 units on a scale | Standard Error 0.13 |
Comparison: Random effects regression analysis of change during treatment on the YBC-EDS, in fluoxetine vs placebo groups.p-value: 0.26Mixed Models Analysis