 |
|
Fluoxetine After Weight Restoration in Anorexia Nervosa
A Randomized Controlled Trial
B. Timothy Walsh, MD;
Allan S. Kaplan, MD, FRCPC;
Evelyn Attia, MD;
Marion Olmsted, PhD;
Michael Parides, PhD;
Jacqueline C. Carter, PhD;
Kathleen M. Pike, PhD;
Michael J. Devlin, MD;
Blake Woodside, MD, FRCPC;
Christina A. Roberto, BA;
Wendi Rockert, MEd
JAMA. 2006;295:2605-2612.
ABSTRACT
| |
Context Antidepressant medication is frequently prescribed for patients with anorexia nervosa.
Objective To determine whether fluoxetine can promote recovery and prolong time-to-relapse among patients with anorexia nervosa following weight restoration.
Design, Setting, and Participants Randomized, double-blind, placebo-controlled trial. From January 2000 until May 2005, 93 patients with anorexia nervosa received intensive inpatient or day-program treatment at the New York State Psychiatric Institute or Toronto General Hospital. Participants regained weight to a minimum body mass index (calculated as weight in kilograms divided by the square of height in meters) of 19.0 and were then eligible to participate in the randomized phase of the trial.
Interventions Participants were randomly assigned to receive fluoxetine or placebo and were treated for up to 1 year as outpatients in double-blind fashion. All patients also received individual cognitive behavioral therapy.
Main Outcome Measures The primary outcome measures were time-to-relapse and the proportion of patients successfully completing 1 year of treatment.
Results Forty-nine patients were assigned to fluoxetine and 44 to placebo. Similar percentages of patients assigned to fluoxetine and to placebo maintained a body mass index of at least 18.5 and remained in the study for 52 weeks (fluoxetine, 26.5%; placebo, 31.5%; P = .57). In a Cox proportional hazards analysis, with prerandomization body mass index, site, and diagnostic subtype as covariates, there was no significant difference between fluoxetine and placebo in time-to-relapse (hazard ratio, 1.12; 95% CI, 0.65-2.01; P = .64).
Conclusions This study failed to demonstrate any benefit from fluoxetine in the treatment of patients with anorexia nervosa following weight restoration. Future efforts should focus on developing new models to understand the persistence of this illness and on exploring new psychological and pharmacological treatment approaches.
Trial Registration clinicaltrials.gov Identifier: NCT00288574
INTRODUCTION
Anorexia nervosa is a serious psychiatric illness with substantial morbidity and a lifetime mortality arguably as high as that associated with any psychiatric illness.1 A major contributor to the poor prognosis of this illness is the high rate of relapse following initial treatment. Despite successful weight restoration, 30% to 50% of patients require rehospitalization within 1 year of discharge.2-3 This discouraging experience has prompted interest in interventions aimed at preventing deterioration following weight restoration.
Patients with anorexia nervosa often exhibit symptoms of other psychiatric disorders,4 such as depression and obsessive-compulsive disorder, which are responsive to antidepressant medication, suggesting that pharmacological interventions might be of use. Surprisingly, virtually all of the controlled trials of medication (most of which have been conducted during the initial phase of treatment when patients are underweight) have shown no benefit of medication compared with placebo.5-6 Despite this lack of evidence of effectiveness, a substantial number of patients with anorexia nervosa receive antidepressant medications. In our sites, approximately 60% of patients with anorexia nervosa seeking treatment report having been treated with a selective serotonin reuptake inhibitor (SSRI). Since the ineffectiveness of medication at low weight might be attributable to the effects of starvation, researchers have more recently explored the utility of medication after initial treatment has reversed the acute effects of starvation, with differing results.
In a retrospective case-control analysis, Strober et al7 found no difference in the outcome of patients with anorexia nervosa taking fluoxetine following hospitalization compared with similar patients not taking fluoxetine. However, in a double-blind, placebo-controlled trial of 35 patients following significant weight restoration, Kaye et al8 found fluoxetine to be superior to placebo in preventing relapse.
The primary aim of the current study was to determine, in a large sample, whether fluoxetine, compared with placebo, reduced the rate of relapse and enhanced psychological and behavioral recovery following initial treatment for anorexia nervosa. After successfully completing a course of intensive treatment aimed at weight restoration, patients were randomly assigned to receive fluoxetine or placebo and treated for up to 1 year in double-blind fashion. In addition to medication or placebo, patients received individual cognitive behavioral therapy specifically designed to prevent relapse.9
METHODS
Patient Recruitment
A patient was eligible to participate if she: (1) was a female between the ages of 16 and 45 years; (2) met Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria for anorexia nervosa (except the requirement for amenorrhea; women who meet all DSM criteria except amenorrhea do not appear to differ substantially from those meeting all criteria10-12) prior to the initial phase of treatment; (3) immediately prior to randomization, had successfully completed treatment at 1 of the study sites in an inpatient or day-program setting during which body mass index (BMI [calculated as weight in kilograms divided by height in meters squared]) reached at least 19 and was maintained for 2 weeks; (4) was not at imminent risk for suicide; (5) did not have a serious medical illness aside from the eating disorder; and (6) was medication free (with the exception of occasional lorazepam, 0.5 mg, or zopiclone, 7.5 mg, for anxiety or sleep disturbance).
Recruitment methods included word of mouth, referrals, the Internet, and advertisements. This study was described either during the first evaluation of patients considering admission to the eating disorders programs or following admission. However, formal recruitment and signing of the informed consent document did not occur until the end of the initial phase of treatment. The institutional review boards of the New York State Psychiatric Institute and Toronto General Hospital approved this study.
Treatment Prior to Randomization
The inpatient/intensive outpatient treatment at both sites focused on nutritional rehabilitation through the provision of supervised meals and behavioral and cognitive interventions. Both programs provided a weight maintenance period of several weeks following weight normalization, during which patients spent increasing amounts of time outside of the hospital.
Treatment
After signing consent, patients were randomly assigned to receive either fluoxetine or placebo. Randomization lists, stratified by site and binge-purge subtype, were generated by a computer program utilizing a random number generator seeded by time of day. Randomization assignments were kept in sealed envelopes and all clinical staff involved in the care of the patient, as well as study coordinators, remained blind to medication assignment during the study. Program staff not involved in the study prepared medication and placebo, which were contained in identical bottles. A statistician performed primary analyses while the research staff remained blind to medication assignment. To determine the effectiveness of blinding procedures, patients, physicians, and therapists were asked upon study termination to guess each patient's assignment. Psychiatrists ( 2 = 8.14, P = .004) and patients (2 = 5.49, P = .02) guessed the correct medication assignment more frequently than chance, while therapists did not (2 = 1.94, P = .16).
Medication was initiated in double-blind fashion 1 week prior to discharge from the hospital. The dose increased from 20 mg/d (1 pill) to 60 mg/d (3 pills) over 1 week under close supervision for any adverse effects. Medication dose was maintained at 60 mg/d unless there were adverse effects, in which case the dose was lowered at the discretion of the psychiatrist; if the patient's clinical status was deteriorating, the dose could be raised to a maximum of 80 mg/d (4 pills). After the first week receiving study medication, patients were discharged to outpatient care and treated for 12 months or until they met criteria for premature study withdrawal or voluntarily withdrew.
In addition to medication, all patients received a form of manualized cognitive behavioral therapy found to be useful in reducing relapse in anorexia nervosa.9 Experienced psychologists and psychiatrists provided therapy, which consisted of 50 individual sessions of 45 minutes’ duration. Patients were able to have up to 5 supplemental sessions with family members and/or significant others. A psychiatrist saw patients to monitor the dose of medication, adverse effects, and general medical status. Therapy sessions were audiotaped and reviewed by 2 of the authors (K.M.P. and M.J.D.) to assess treatment fidelity and adherence. Blood samples were obtained every 3 months for determination of plasma levels of fluoxetine and nor-fluoxetine.
Study participation was terminated and the patient was considered to have relapsed if (1) the participant's BMI fell to or below 16.5 for 2 consecutive weeks; (2) severe medical complications (other than low weight) arose as a result of the eating disorder; (3) the participant was judged to be at imminent risk of suicide; or (4) the participant developed another severe psychiatric disorder requiring treatment. Thus, the definition of relapse captured severe worsening of symptoms of anorexia nervosa and/or the development of other major clinical problems.
Measures
To evaluate symptoms of disturbed eating behaviors and cognitions, depression, anxiety, and self-esteem, the following assessments were collected every 4 weeks: Eating Disorders Inventory (EDI),13 the Beck Depression Inventory (BDI),14 the Beck Anxiety Inventory (BAI),15 and the Rosenberg Self-Esteem Scale (RSE).16 Obsessions and quality of life were measured with the Yale Brown Cornell Obsessive Compulsive Scale for Eating Disorders (YBC-EDS)17 and the Quality of Life Enjoyment and Satisfaction Questionnaire (QlesQ)18 and were obtained before randomization, at 6 months, and again at study termination. Participants were weighed weekly.
Statistical Methods
The study was designed to randomize 90 patients with equal allocation to fluoxetine and placebo. Ninety patients provide 90% power to detect a relative 50% reduction in relapse for fluoxetine compared with placebo based on a 2-sided .05 level log-rank test. A relapse rate of 70% was assumed for placebo and 35% for fluoxetine corresponding to a hazard ratio of approximately 0.34.
The primary outcome measure was time-to-relapse. Patients who voluntarily withdrew from treatment ("dropouts") were classified in 3 distinct ways. In the most conservative analysis, all patients who left the study prior to completing a full course of treatment (at least 38 psychotherapy sessions over 50 to 60 weeks) were considered to have relapsed. In a second least conservative analysis, only patients who met the predefined criteria described previously for study termination were considered to have relapsed; data from all other dropouts were censored. For the third analysis, dropouts were considered to have relapsed if they met any of the following criteria at the time they chose to end participation in the study: (1) had a BMI less than 17.5; (2) were binge eating and/or purging 2 or more times per week for the 4 weeks prior to dropping out; or (3) exhibited depressive and/or anxiety symptoms sufficient to interfere with functioning (eg, serious suicidal ideation). Cox proportional hazards regression was used to test for differences between randomization groups adjusting for site and diagnostic subtype (the 2 factors used to stratify the randomization) and by pretreatment BMI, which was found to differ significantly between randomization groups at the .05 level.
We also compared the proportion of patients in each group who failed to maintain normal weight (BMI 18.5) for 52 weeks (ie, relapsed) using a 2 test. Patients whose weight fell below this level for 4 consecutive weeks or who withdrew from the study prior to 52 weeks were considered to have relapsed. Also, at study termination, patient status was classified according to modified Morgan-Russell criteria,19-20 and the percentage of patients in fluoxetine and placebo groups classified as fair or better were compared using the 2 statistic. Additional secondary analyses examined differences between groups in change in symptoms over time using random-effects regression models. All outcome analyses adhered to the intention-to-treat principle; a P value of .05 using 2-sided tests was the criterion for significance for all analyses except the Cox proportional hazards regressions, for which, because of an interim analysis, a P value of .04 was used. Statistical analysis was completed using SAS version 9.1.3, SAS Institute Inc, Cary, NC, and SPSS version 13.0, SPSS Inc, Chicago, Ill.
RESULTS
Patient Characteristics
Patients were recruited between January 2000 and May 2004. The last patient completed the study in May 2005. One hundred fifty patients were evaluated for study participation (Figure 1). Twenty-three did not meet criteria for eligibility, 32 declined participation, and 2 were excluded for other reasons. The remaining 93 patients entered the trial, 45 in New York, 48 in Toronto. Forty-nine patients were assigned to fluoxetine and 44 to placebo. On average, patients were a mean (SD) of 23 (4.6) years old, had been ill for 56.5 (44.7) months, had been previously hospitalized 2.1 (2.5) times, and had a BMI of 15.4 (1.8) at the time of entry into the initial phase of treatment. Forty-five patients (48.4%) met DSM-IV criteria for binge-purge subtype. There were no statistically significant differences in these parameters between the patients assigned to fluoxetine or placebo, or between the New York or Toronto sites.
|
|
|
|
Figure 1. Flow of Patients Through the Study
Body mass index (BMI calculated as weight in kilograms divided by height in meters squared).
|
|
|
At the time of randomization, patients in the fluoxetine and placebo groups were quite similar (Table 1). However, there was a small but statistically significant difference in BMI with patients receiving placebo starting the trial at higher BMI. Only patients who had gained to a BMI of at least 19 were eligible to enter the randomized phase of the study. In practice, most patients gained little additional weight, leading to very limited variance in BMI at the time of randomization. The limited variation contributed to the statistical significance of the small difference in BMI between groups.
|
|
|
|
Table 1. Demographic Information, Body Mass Index, and Psychological Assessments at Prerandomization and at Study Termination
|
|
|
Outcome
Forty of the 93 patients completed the entire course of treatment and 53 terminated prematurely (Figure 1 and Table 2). The major reasons for premature termination were meeting a priori criteria for relapse and patient dissatisfaction with treatment. Patients were also withdrawn from the study if they failed to attend a minimum number of psychotherapy sessions in a quarter (12 during the first quarter, 9 during the second and third quarters, and 8 during the last quarter) or if they took less than 10 mg of the medication for 2 weeks or more. A 17-year-old patient assigned to fluoxetine made a suicide attempt during the study; there were no other unexpected adverse events. There were no significant differences between fluoxetine and placebo groups in BMI or in measures of psychological state at the time of termination (Table 1). There were no significant differences between the fluoxetine and placebo groups in the percentages of patients completing the full study (2 = 0.001; P = .98), or in the average number of sessions of psychotherapy (fluoxetine, 38.1 [18.1]; placebo, 34.8 [16.8]; t = 0.48; df = 91; P = .63). However, there were significant differences in the rates of full-study completion between sites (New York, 28.9%; Toronto, 56.3%; 2 = 7.094; P = .008) and between subtypes (binge-purge, 31.1%; restricting, 54.2%; 2 = 5.037; P = .03).
|
|
|
|
Table 2. Reason for Study Termination and Status at Termination as Per Modified Morgan-Russell Criteria for Outcome
|
|
|
At the time of termination, similar percentages of patients in the fluoxetine and placebo groups met modified Morgan-Russell criteria for fair or better outcome (fluoxetine, 65.31%; placebo, 56.82%; 2 = 1.005; P = .32) (Table 2). By Morgan-Russell criteria at time of termination, there were no significant differences between sites or between binge-purge and restricting subtypes (data not shown).
Similar percentages of patients assigned to fluoxetine and to placebo maintained a BMI of at least 18.5 and remained in the study for 52 weeks (fluoxetine: 26.5%; placebo: 31.5%; 2 = 1.31, P = .57). A greater percentage of patients in Toronto (39.6%) met this outcome criterion compared with New York (17.8%; 2 = 5.36; P = .02). There was no statistically significant difference between binge-purge (22.2%) and restricting (35.4%) subtypes on this measure (2 = 1.96; P = .16).
Time-to-Relapse
The most conservative analysis of time-to-relapse, in which premature termination for any reason was considered to signify relapse, found no significant difference between the fluoxetine and placebo groups ( = –0.133; SE = 0.288; df = 1; P = .64) (Figure 2). At 52 weeks, 45% of the placebo group and 43% of the fluoxetine group had not relapsed. In this analysis, there were significant effects of site ( = 0.814; SE = 0.287; df = 1; P = .005), of subtype ( = 0.777; SE = 0.285; df = 1; P = .006), and of prerandomization BMI ( = –0.761; SE = 0.323; df = 1; P = .02). Patients in Toronto remained in the trial for a longer time prior to relapse than did those in New York, as did patients with the restricting subtype and patients with a higher BMI at prerandomization. Results from the analyses using the other methods of classifying dropouts were similar; neither indicated any difference between fluoxetine and placebo groups in time-to-relapse. At 52 weeks, using the least conservative analysis, in which patients who voluntarily withdrew were censored, 73% of the placebo group and 71% of the fluoxetine group had not relapsed. Using the analysis in which patients who dropped out were clinically classified, 51% of the placebo group and 49% of the fluoxetine group had not relapsed at 52 weeks.
|
|
|
|
Figure 2. Patients Receiving Placebo and Fluoxetine Remaining in Treatment vs Number of Weeks After Treatment Inception
Circles indicate when patients left the trial after a full course of treatment (minimum 38 sessions, maximum 61 sessions over 12 months). Twelve month/full course of treatment criterion was satisfied if 38-session minimum was met between 50 and 60 weeks after randomization.
|
|
|
As noted previously, the average prerandomization BMI of the placebo group was slightly but significantly higher than that of the fluoxetine group. Because a higher BMI was associated with a longer time-to-relapse, BMI was included in the analyses described above to control statistically for the difference between fluoxetine and placebo groups. Additional secondary analyses of time-to-relapse were conducted including only the 55 patients whose prerandomization BMIs were greater than 20.15. Among these patients, there was no statistically significant difference between the mean prerandomization BMIs of the fluoxetine (n = 24; 20.51 [0.42]) and placebo groups (n = 31; 20.69 [0.41]; t(53) = 1.61; P = .11). In none of these analyses did the difference between fluoxetine and placebo groups in time-to-relapse approach significance (P>.50 for all).
Because the study by Kaye et al8 had excluded patients with binge eating, these analyses were repeated using data from only the 48 patients with the restricting subtype of anorexia nervosa. There was no evidence of a significant difference between the fluoxetine and placebo groups (data not shown).
Fluoxetine Levels
At the last visit, the mean (SD) dose of medication was 63.5 [15.8] mg/d for patients on fluoxetine and the equivalent of 71.4 [13.2] mg/d for patients receiving placebo (t = 2.6; df = 91; P = .01). Plasma samples for analysis of fluoxetine and nor-fluoxetine levels were available for 43 of the 49 patients assigned to receive fluoxetine. Plasma samples from 3 patients had no detectible fluoxetine or nor-fluoxetine. Exclusion of the data from these 3 patients did not substantially alter the results of the analyses of time-to-relapse. The average fluoxetine and nor-fluoxetine levels for the remaining 40 patients were 409 (225) and 297 (154) ng/mL, respectively; these levels are comparable with those reported in studies of obsessive-compulsive disorder21 and the study by Kaye et al.8 Among the patients with detectable fluoxetine or nor-fluoxetine levels, there was no significant relationship between the average total fluoxetine plus nor-fluoxetine plasma concentration and time-to-relapse using any of the methods for classifying dropouts (data not shown). Plasma samples were available for 38 of the 44 patients assigned to placebo; none had detectable levels of fluoxetine or nor-fluoxetine.
Prior SSRI Treatment
Twenty-four of the 93 patients had received 1 course of SSRI treatment (defined as 3 or more continuous weeks receiving an SSRI) prior to their participation in the current study, and 33 had received 2 or more previous courses of SSRI treatment. There were no significant differences between the fluoxetine and placebo groups in time-to-relapse using any of the criteria for classifying dropouts among patients with no prior SSRI treatment or among patients with no more than 1 course of prior treatment (data not shown).
Random-Effects Regression
Random-effects regression models were used to compare randomization groups over time using all available data from all patients (Table 3). The factors examined were weight, several measures of general (BDI, BAI, RSE) or specific (EDI-Drive for Thinness, Bulimia, Body Dissatisfaction and Perfectionism, and YBC-EDS) psychopathology, and a measure of quality of life (QlesQ). The only measure for which a significant fluoxetine-placebo difference was detected was the BAI; the mean (SD) change in BAI per month was –0.70 (0.12) in the fluoxetine group vs –0.22 (0.13) in the placebo group (P = .007; Bonferroni corrected: P = .07).
|
|
|
|
Table 3. Random-Effects Regression Analysis of Change During Treatment
|
|
|
COMMENT
The current study was designed to determine whether the antidepressant medication fluoxetine reduced relapse following successful initial treatment of anorexia nervosa. Ninety-three patients entered the double-blind phase, making this trial, to our knowledge, the largest controlled medication trial conducted to date in anorexia nervosa. All patients had reached a minimum BMI of 19.0, so that the acute effects of undernutrition were minimized. The average dose of fluoxetine was 63.5 mg/d similar to the dose routinely used in the treatment of bulimia nervosa. The study was conducted at 2 sites experienced in the treatment of anorexia nervosa, lending support to the generalizability of the findings. Despite these strengths, the current study found no convincing evidence that fluoxetine provided significant benefit compared to placebo. On virtually all measures, including time-to-relapse, rate of study completion, BMI, and clinical measures at time of termination, including depression, the group receiving fluoxetine was statistically indistinguishable from that receiving placebo. There was no indication that the effect of fluoxetine varied significantly by site, by subtype of anorexia nervosa, or by history of prior exposure to SSRIs. A single secondary analysis examining change in the BAI over time found a significant difference favoring the fluoxetine group; however, since the fluoxetine and placebo groups did not differ in similar analyses of 9 other measures, we cannot be confident that this difference between groups was not due to chance.
Both patients and their psychiatrists, but not their therapists, were more likely than chance to guess whether patients were receiving active medication. We speculate that this was due to the presence of minor adverse effects. However, since virtually no differences in outcome were detected between the fluoxetine and medication groups, it seems unlikely that this information substantially affected the assessment of clinical status.
The interpretation of treatment trials such as this one, in which a substantial number of patients did not complete the full course of treatment, can be problematic. However, since relapse soon after successful weight gain is a well-recognized clinical problem in the treatment of anorexia nervosa, we anticipated that many patients would not complete a full year of outpatient care and designed this study to examine the course of relapse during this time. The results of the survival analyses help clarify the reasons for premature termination and the impact of medication. Twenty of the 53 patients who did not complete a full course of treatment met our stringent, a priori criteria for full relapse and were therefore withdrawn from the study and referred for other treatment. An additional 24 of the 53 noncompleters had not deteriorated so severely as to require withdrawal but were clinically classified, blind to medication assignment, as doing sufficiently poorly to be considered as having relapsed at the time they left the study; many of these patients voluntarily withdrew from the study because of their lack of progress. Therefore, 44 (83%) of the 53 patients who did not complete treatment had deteriorated substantially, and this deterioration was a major contributor to their withdrawal. We had hoped that fluoxetine would alter the course of illness and thereby reduce such deterioration and the rate of withdrawal, as suggested by Kaye et al.8 Unfortunately, medication assignment had no discernible impact on time-to-relapse whether all dropouts were considered to have relapsed, whether no dropouts were considered to have relapsed, or whether dropouts were classified as having relapsed based on their status at time of study termination. In other words, because discontinuation of treatment by patients with anorexia nervosa is a likely sign of impending relapse, systematic examination of such withdrawal can provide information about treatment efficacy. In this study, fluoxetine was no more efficacious than placebo.
A limitation on the conclusions of the current study is that we examined the utility of fluoxetine at a particular stage of illness and in conjunction with a particular form of structured psychological treatment. We cannot exclude the possibility that antidepressants might have an effect at other stages of illness, for example, in patients who have had anorexia nervosa for only a short time or who maintained a normal weight for a longer period of time before receiving antidepressant treatment. Similarly, it is conceivable that fluoxetine might provide benefit if it were the sole treatment provided or given in association with some other intervention. However, the results of the current study are consistent with most previous controlled trials of medication in anorexia nervosa, which have failed to show significant benefit from antidepressant medication compared with placebo.5, 22
Unlike Kaye et al,8 we were unable to demonstrate a difference between fluoxetine and placebo on time-to-relapse. A direct comparison between that study and the current one is confounded by a number of differences. In the current study, all patients received a structured psychotherapy specifically aimed to reduce relapse, whereas in the study by Kaye et al,8 the psychosocial treatment received was not standardized; this difference may explain the high relapse rate in the placebo group in the study by Kaye et al.8 In addition, criteria for relapse were less specific in that study and patients were encouraged to withdraw if they felt they were doing poorly. Because of limited information available about the treatment response of individuals with the restricting subtype of anorexia nervosa and of indications that such patients might respond more favorably to fluoxetine,25 Kaye et al8 focused their analysis on 35 patients who denied binge eating. Additional patients with binge eating were randomized according to personal communication with Walter H. Kaye, MD, Western Psychiatric Institute and Clinic, University of Pittsburgh School of Medicine, and the restriction of the analysis to a subgroup may have inflated the possibility of a false-positive result.27
The current study has implications for both clinical practice and research. The present findings, coupled with those of previously published studies, indicate that the common practice of prescribing antidepressant medication is unlikely to provide substantial benefit for most patients with anorexia nervosa, either when they are underweight or immediately upon weight restoration. These data imply that therapeutic efforts would be better devoted to psychological and behavioral interventions for which there is some, albeit modest, evidence of efficacy.9, 23 Future research on the utility of novel psychological treatments24-25 and innovative psychotropic (eg, olanzapine)26-27 and nonpsychotropic (eg, cycloserine)28 medications is obviously needed. In addition, it may be important to examine the fundamental behavioral and cognitive disturbances of anorexia nervosa from fresh perspectives. Investigators have recently suggested that patients with anorexia nervosa might be viewed as having disturbances in extinguishing conditioned fears29 or in changing cognitive set.30-31 New frameworks for understanding the impressive persistence of anorexia nervosa may, eventually, lead to more effective treatments, which are sorely needed.
AUTHOR INFORMATION
Corresponding Author: B. Timothy Walsh, MD, Department of Psychiatry, College of Physicians and Surgeons of Columbia University/New York State Psychiatric Institute, 1051 Riverside Dr, Unit 98, New York, NY 10032 (btw1{at}columbia.edu).
Author Contributions: Dr Walsh had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Walsh, Kaplan, Attia, Olmsted, Parides, Carter, Pike, Devlin, Woodside, Rockert.
Acquisition of data: Walsh, Kaplan, Attia, Olmsted, Carter, Pike, Devlin, Woodside, Roberto, Rockert.
Analysis and interpretation of data: Walsh, Kaplan, Attia, Olmsted, Parides, Roberto, Rockert.
Drafting of the manuscript: Walsh, Kaplan, Parides, Roberto.
Critical revision of the manuscript for important intellectual content: Attia, Olmsted, Carter, Pike, Devlin, Woodside, Rockert.
Statistical analysis: Walsh, Parides, Roberto.
Obtained funding: Walsh, Kaplan.
Administrative, technical, or material support: Walsh, Kaplan, Attia, Olmsted, Roberto, Rockert.
Study supervision: Walsh, Kaplan, Pike.
Financial Disclosures: Dr Walsh has reported receiving research support from Eli Lilly and Co, Abbott Laboratories, Ortho-McNeil Pharmaceuticals, and GlaxoSmithKline. Drs Kaplan and Woodside have reported receiving research support from Eli Lilly and Co. Dr Devlin has reported receiving research support from Ortho-McNeil Pharmaceuticals and Eli Lilly and Co. Dr Attia has reported receiving research support from Eli Lilly and Co and Pfizer Inc. No other authors reported disclosures.
Funding/Support: This study was supported in part by grants MH060271 and MH60336 from National Institutes of Health. Eli Lilly supplied fluoxetine and placebo.
Role of the Sponsor: The National Institutes of Health and Eli Lilly had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript.
Previous Presentation: Preliminary results of this study were presented at the meeting of the Eating Disorders Research Society; September 29, 2005; Toronto, Ontario.
Acknowledgment: We thank the staffs of the General Clinical Research Unit at New York State Psychiatric Institute/Columbia University Medical Center, New York, NY and of Toronto General Hospital, Toronto, Ontario. The members of the data and safety monitoring board: Sidney Kennedy, MD, University of Toronto, Toronto, Ontario; Andrew Leon, PhD, Weill Medical College of Cornell University, New York, NY; Steven Roose, MD, New York State Psychiatric Institute/Columbia University Medical Center; Jane Pearson, PhD, program officer, National Institute of Mental Health, Bethesda, Md. The members of the Eating Disorders Research Unit at New York State Psychiatric Institute/Columbia University Medical Center: Lauren Escott, MA, and Dana Satir, BA, Boston University, Boston, Mass, for data collection; Juli Goldfein, PhD, Diane Klein, MD, Lisa Kotler, MD, Laurel Mayer, MD, Pamela Raizman, PhD, Joanna Steinglass, MD, and Sara Wolk, PhD, New York State Psychiatric Institute/Columbia University Medical Center, for clinical care of study participants; Katharine Loeb, PhD, Mt Sinai School of Medicine, New York, NY; Sally Woodring, RN, New York State Psychiatric Institute/Columbia University Medical Center, for medication supply management; Jonathan Cohen, MS, New York State Psychiatric Institute/Columbia University Medical Center, for database creation and support. The members of the Toronto General Hospital research team: Traci McFarlane, PhD, and Randy Staab, MD, FRCP(C), Toronto General Hospital/University of Toronto; Pavla Reznicek, PhD, private practice, Toronto, for clinical care of study participants; Anita Federici, MSc, York University, Toronto, for data collection; and Nancy Lipson, MSW, RSW, Concurrent Disorders Service, Center for Addiction and Mental Health, Toronto, for medication supply management.
The following individuals included in the acknowledgments reported having received compensation from the NIH grant for their involvement in the study: Jonathan Cohen, MS; Lauren Escott, MA; Anita Federici, MSc; Juli Goldfein, PhD; Lisa Kotler, MD; Nancy Lipson, MSW; Pamela Raizman, PhD; Pavla Reznicek, PhD; Randy Staab, MD, FRCP(C); Dana Satir, BA; and Sara Wolk, PhD.
Author Affiliations: New York State Psychiatric Institute/Columbia University Medical Center, New York, NY (Drs Walsh, Attia, Parides, Pike, and Devlin and Ms Roberto); and University of Toronto and Toronto General Hospital, Toronto, Ontario (Drs Kaplan, Olmsted, Carter, and Woodside and Ms Rockert).
REFERENCES
| |
1. Sullivan PF. Mortality in anorexia nervosa. Am J Psychiatry. 1995;152:1073-1074.
FREE FULL TEXT
2. Pike KM. Long-term course of anorexia nervosa: response, relapse, remission, and recovery. Clin Psychol Rev. 1998;18:447-475.
FULL TEXT
|
ISI
| PUBMED
3. Jones LM, Halford WK, Dooley RT. Long term outcome of anorexia nervosa. Behav Change. 1993;10:93-102.
4. Halmi KA, Eckert E, Marchi P, Sampugnaro V, Apple R, Cohen J. Comorbidity of psychiatric diagnoses in anorexia nervosa. Arch Gen Psychiatry. 1991;48:712-718.
ABSTRACT
5. Zhu AJ, Walsh BT. Pharmacologic treatment of eating disorders. Can J Psychiatry. 2002;47:227-234.
ISI
| PUBMED
6. Attia E, Haiman C, Walsh BT, Flater SR. Does fluoxetine augment the inpatient treatment of anorexia nervosa? Am J Psychiatry. 1998;155:548-551.
FREE FULL TEXT
7. Strober M, Freeman R, DeAntonio M, Lampert C, Diamond J. Does adjunctive fluoxetine influence the post-hospital course of anorexia nervosa? a 24-month prospective, longitudinal followup and comparison with historical controls. Psychopharmacol Bull. 1997;33:425-431.
ISI
| PUBMED
8. Kaye WH, Nagata T, Weltzin TE, et al. Double-blind placebo controlled administration of fluoxetine in restricting- and restricting-purging type anorexia nervosa. Biol Psychiatry. 2001;49:644-652.
FULL TEXT
|
ISI
| PUBMED
9. Pike KM, Walsh BT, Vitousek K, Wilson GT, Bauer J. Cognitive behavior therapy in the posthospitalization treatment of anorexia nervosa. Am J Psychiatry. 2003;160:2046-2049.
FREE FULL TEXT
10. Cachelin FM, Maher BA. Is amenorrhea a critical criterion for anorexia nervosa? J Psychosom Res. 1998;44:435-440.
FULL TEXT
|
ISI
| PUBMED
11. Garfinkel PE, Lin E, Goering P, et al. Should amenorrhoea be necessary for the diagnosis of anorexia nervosa? evidence from a Canadian community sample. Br J Psychiatry. 1996;168:500-506.
FREE FULL TEXT
12. Watson TL, Andersen AE. A critical examination of the amenorrhea and weight criteria for diagnosing anorexia nervosa. Acta Psychiatr Scand. 2003;108:175-182.
FULL TEXT
|
ISI
| PUBMED
13. Garner DM, Olmsted MP. The Eating Disorder Inventory Manual. Odessa, Fla: Psychological Assessment Resources; 1984.14. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:561-571.
ISI
| PUBMED
15. Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol. 1988;56:893-897.
FULL TEXT
|
ISI
| PUBMED
16. Rosenberg M. Conceiving the Self. New York, NY: Basic Books; 1979.17. Mazure CM, Halmi KA, Sunday SR, Romano SJ, Einhorn AM. The Yale-Brown-Cornell eating disorder scale: development, use, reliability and validity. J Psychiatr Res. 1994;28:425-445.
FULL TEXT
|
ISI
| PUBMED
18. Endicott J, Nee J, Harrison W, Blumenthal R. Quality of life enjoyment and satisfaction questionnaire: a new measure. Psychopharmacol Bull. 1993;29:321-326.
ISI
| PUBMED
19. Eckert ED, Halmi KA, Marchi P, Grove W, Crosby R. Ten-year follow-up of anorexia nervosa: clinical course and outcome. Psychol Med. 1995;25:143-156.
ISI
| PUBMED
20. Russell GFM, Szmukler GI, Dare C, Eisler I. An evaluation of family therapy in anorexia nervosa and bulimia nervosa. Arch Gen Psychiatry. 1987;44:1047-1056.
ABSTRACT
21. Koran LM, Cain JW, Dominguez RA, Rush AJ, Thiemann S. Are fluoxetine plasma levels related to outcome in obsessive-compulsive disorder? Am J Psychiatry. 1996;153:1450-1454.
FREE FULL TEXT
22. Holtkamp K, Konrad R, Kaiser N, et al. A retrospective study of SSRI treatment in adolescent anorexia nervosa: insufficient evidence for efficacy. J Psychiatr Res. 2005;39:303-310.
FULL TEXT
|
ISI
| PUBMED
23. Lock J, le Grange D. Family-based treatment of eating disorders. Int J Eat Disord. 2005;37:S64-S67.24. McIntosh VV, Jordan J, Carter FA, et al. Three psychotherapies for anorexia nervosa: a randomized, controlled trial. Am J Psychiatry. 2005;162:741-747.
FREE FULL TEXT
25. Kaye WH, Weltzin TE, Hsu LK, Bulik CM. An open trial of fluoxetine in patients wit |
