Sie sind auf Seite 1von 20

Pharmacotherapy of Eating Disorders -- Jackson et al.

25 (2): 143 -- Nutrition in Clinical Practice

Frame contained PDF file, click here to view

Adv Search Browse Search History

My Marked Citations (0) My Tools

Quick Search this Journal

Go
Advanced Search

Journal Navigation
Journal Home

Subscriptions

Archive

Contact Us

Table of Contents

Version
PDF version of: Jackson et al. 25 (2):
143. (2010)

This Article
Abstract
Full Text Free to you

Alerts
Alert me when this article is cited
Alert me if a correction is posted

Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Add to Saved Citations
Download to citation manager
Request Permissions
Request Reprints
Add to My Marked Citations
Citing Articles via Scopus

Google Scholar

file:///C|/Documents%20and%20Settings/cjackson/Desktop...%20143%20--%20Nutrition%20in%20Clinical%20Practice.htm (1 of 2)5/6/2010 2:07:39 PM


Pharmacotherapy of Eating Disorders -- Jackson et al. 25 (2): 143 -- Nutrition in Clinical Practice

Articles by Jackson, C. W.
Articles by Lorenz, R.

PubMed
PubMed Citation
Articles by Jackson, C. W.
Articles by Lorenz, R.

Social Bookmarking

What's this?

Help
Adobe Acrobat plugin users: View
article in full window

Printing problems? To print a PDF using


the Acrobat plugin, use the printer
button in the plugin's toolbar, located
immediately above the document.
Download Adobe Reader.

Institution: AUBURN UNIV |


Sign In via User Name/Password

file:///C|/Documents%20and%20Settings/cjackson/Desktop...%20143%20--%20Nutrition%20in%20Clinical%20Practice.htm (2 of 2)5/6/2010 2:07:39 PM


Nutrition in Clinical Practice
http://ncp.sagepub.com

Pharmacotherapy of Eating Disorders


Cherry Wyant Jackson, Marshall Cates and Raymond Lorenz
Nutr Clin Pract 2010; 25; 143
DOI: 10.1177/0884533610362239

The online version of this article can be found at:


http://ncp.sagepub.com/cgi/content/abstract/25/2/143

Published by:

http://www.sagepublications.com

On behalf of:

The American Society for Parenteral & Enteral Nutrition

Additional services and information for Nutrition in Clinical Practice can be found at:

Email Alerts: http://ncp.sagepub.com/cgi/alerts

Subscriptions: http://ncp.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Invited Review Nutrition in Clinical Practice
Volume 25 Number 2
April 2010 143-159

Pharmacotherapy of Eating Disorders 2010 American Society for


Parenteral and Enteral Nutrition
10.1177/0884533610362239
http://ncp.sagepub.com
Cherry Wyant Jackson, PharmD1; Marshall Cates, PharmD1; hosted at
http://online.sagepub.com
and Raymond Lorenz, PharmD2

Eating disorders are complex, chronic disorders that are difficult date, and that is for fluoxetine in the treatment of bulimia. Clearly,
to treat. In addition, the 2 primary eating disorders, anorexia ner- new large-scale, independent studies using novel agents, and
vosa and bulimia nervosa, may have acute, life-threatening conse- studying the use of medications for both short- and long-term
quences. Medication trials for eating disorders have been outcomes, are needed. (Nutr Clin Pract. 2010;25:143-159)
hampered by high dropout rates, high placebo response, short trial
duration, insufficient doses, and difficult outcome measures. Only Keywords: eating disorders, anorexia, anorexia nervosa,
1 medication has been FDA approved for any eating disorder to bulimia, bulimia nervosa, medication therapy management

E
ating disorders are complex disorders that may have blinding, including whether the trial is single- or double-
life-threatening consequences. The 3 most com- blind. Also, some studies either use a nonplacebo control,
mon eating disorders are anorexia nervosa (AN), or they do not use a control in the study at all. We will
bulimia nervosa (BN), and binge-eating disorder (BED). review literature that is clearly described as both double-
There are varying estimates of the incidence and preva- blind and placebo-controlled, as we discuss the use of
lence of eating disorders. The lifetime estimate of AN in medications for the treatment of eating disorders.
women ranges from 0.3% to 3.7%, and the prevalence of
bulimia in women ranges from 1% to 4.2%.1-3 Based on
population-based research, BED prevalence has been esti- Anorexia Nervosa
mated at between 0.7% and 3%.4 Although eating disorders
occur more frequently in women, men may also have eat- AN is a disorder typically beginning in early to late adoles-
ing disorders at a prevalence ratio between 1:6 and 1:10.1 cence, although it has been reported in preadolescent
Antidepressants are the medications most frequently children as well.5 This disorder is often precipitated by
used for the treatment of eating disorders. Although the stressful life situations in females who are high achievers
mechanism of action of antidepressants in the treatment and feel that diet is an aspect of their life in which they
of eating disorders is unknown, it is thought that eating can exert some control when other areas of their lives
disorders may be caused by the same chemical imbalance seem out of control. AN is characterized by the refusal to
of serotonin and norepinephrine that causes depression. maintain ones body weight over a minimal normal weight
In addition, antidepressants decrease the anxiety, impul- for age and height.6 When this is the case in adolescents,
sivity, and obsessive behavior frequently seen in individu- it may also limit sexual maturation and bone growth.7,8
als with eating disorders. Initial adverse effects with Other characteristics include an intense fear of gaining
antidepressants, especially the second-generation agents weight, or of becoming fat, and a distorted body image.6
used for eating disorders, are nausea and headache. Currently the Diagnostic and Statistical Manual of Mental
Long-term adverse effects include sexual dysfunction and Disorders 4th edition Text Revised (DSM-IV-TR) lists 3
weight gain. cycles of amenorrhea as being part of the diagnostic crite-
Literature regarding the treatment of eating disorders ria; however, this is not always the case in patients with
with medications is of highly variable quality and can be anorexia and will probably not be listed as a criterion for
difficult to assess. Many studies have an ambiguous diagnosis in the DSM-V.9
design and do not adequately describe the method of Patients with AN may either restrict their total food
intake or binge/purge, typically by self-inducing vomiting,
From the 1University of AlabamaBirmingham; and the 2University using laxatives or diuretics, and/or extensively exercising.
of South Alabama, Mobile. As long as the patient continues to be 15% below ideal
body weight (IBW) and meets other criteria for AN, the
Address correspondence to: Cherry Wyant Jackson, PharmD,
University of Alabama, Birmingham, 127 Engel, 1700 7th diagnosis does not evolve to a diagnosis of BN.6
Avenue South, Birmingham, AL 35294-0018; e-mail: cwj0002@ AN has a high mortality, estimated at 10% within
auburn.edu. 10 years of diagnosis, and it is the primary cause of

143
Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010
144 Nutrition in Clinical Practice / Vol. 25, No. 2, April 2010

Table 1. Medical Complications of Anorexia healthful nutrition and eating habits as well as having
Nervosa1,2,7,8,10-14 the patient incorporate the new habits into his or her
daily life, and helping him or her to change cognitive
Bradycardia
Orthostasis perceptions about eating, weight, and the underlying
Acrocyanosis feeling about perfection and control is imperative.
Cognitive impairment Finally, involving the family in the patients recovery
Anxious, depressed, or irritable mood and preventing relapse are all treatment goals.16
Possible seizures The use of medications in the treatment of AN has
Peripheral neuropathy largely been disappointing. AN was first defined in DSM
Low body temperature III in the early 1980s, and treatments have primarily used
Cold intolerance weight gain as a treatment outcome.17 Many of the first
Fatigue trials of medications conducted for AN were of too short
Electrolyte abnormalities
a duration or used doses of medication that would now be
Decreased serum T3; increase in reverse T3
considered too low to achieve a response. Studies were
Increased serum cortisol
Thiamine, folate, B12, niacin deficiency frequently completed without significant power, and gen-
Increased serum carotene der and ethnicity were not reported in all trials.
Delayed gastric emptying Table 2 includes patient demographics, study design,
Constipation duration, and outcome variables for 19 trials for the
Bloating medication treatment of AN completed between 1980
Cessation of menses and 2009.18-36 Of those trials, 8 used either selective sero-
Fertility problems tonin reuptake inhibitors (SSRIs) or venlafaxine, 2 used
Point tenderness tricyclic antidepressants (TCAs), 2 used cisapride, 3 used
Osteopenia or osteoporosis antipsychotics, 3 used hormones, and 1 used zinc. The
discussion below will focus on double-blind, placebo con-
trolled trials.
death in young women between the ages of 15 and
24.10 Death typically occurs as a result of suicide,
infection, or chronic starvation.11 Refeeding the patient Antidepressants in the Treatment of AN
may also be problematic and is associated with signifi-
Ten studies evaluated effects of antidepressants in indi-
cant morbidity and mortality.12,13 Medical evaluation
viduals with AN. Six studies involved fluoxetine; 1 citalo-
and stabilization of patients with AN is extremely
pram; 1 venlafaxine and fluoxetine; 1 fluoxetine,
important. Often, by the time family and friends real-
amitriptyline, and clomipramine; and 2 amitriptyline.
ize the patient has a problem or they are ready to seek
Again, we will only discuss the double-blind, placebo-
help, the patient is profoundly underweight and may
controlled trials here.
be experiencing a number of physical complications
like hypothermia, bradycardia, hypotension, edema,
lanugo, and a variety of metabolic changes (Table 1). Fluoxetine. Three studies compared fluoxetine vs placebo
In the case of severe weight loss (<75% of IBW), hos- in individuals with AN. In 1998, Attia and colleagues25
pitalization is necessary to prevent death secondary to performed a double-blind, placebo-controlled trial of
starvation.14 fluoxetine 60 mg (n = 15) vs placebo (n = 16) in 31 inpa-
Comorbid disorders are common in AN. In patients tients over a 7-week period. Patients entering the trial
who restrict eating, depression, anxiety, and obsessional weighed an average of 92 pounds (72.5% of IBW). The
behaviors such as intrusive or repetitive thoughts, perfec- primary outcome was for patients to gain 1.5 pounds per
tionistic traits, and rigid cognitive styles are common and week until 90% of IBW was achieved. Three percent of
may need treatment in addition to the eating disorder. In patients dropped out of the trial. Fluoxetine was well tol-
patients who purge, self-harm and suicidal behaviors may erated. There were no significant changes between fluox-
occur in addition to the comorbid disorders previously etine and placebo in terms of weight, anxiety, or depression
listed.15 scales.25
In 2001, Kaye and colleagues performed a 52-week
trial in which patients with AN were randomized to fluox-
Medication Treatment
etine (n = 19) or placebo (n = 20) prior to discharge from
The initial goal of treating AN is to restore the patient inpatient hospitalization. Fluoxetine was titrated up to a
to an IBW and treat any physical complications of the dose of 60 mg daily as tolerated. By the end of 52 weeks,
eating disorder. Following that, teaching the patient approximately 70% of patients had dropped out of

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Table 2. Anorexia Nervosa: Medication Trials
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
1984 Vandereycken18 18 0 13-33 Inpatient NR Double-blind, Sulpiride (300-500 mg) vs Yes No significant differences
crossover placebo
1985 Biederman et al19 25 0 11-27 Inpatient and 5 wk Double-blind, Amitriptyline (175 mg) vs No No significant differences
outpatient placebo placebo
controlled
1986 Halmi et al20 72 0 13-36 Inpatient 90 d Double-blind, Amitriptyline (160 mg) vs No Neither better than
placebo cyproheptadine (32 mg) vs placebo
controlled placebo
1994 Birmingham 54 0 >15 Inpatient NR Double-blind, Elemental zinc (14 mg) vs No Zinc improved BMI
et al21 placebo placebo (P = .03)
controlled
1995 Brambilla et al22 13 0 14-35 Outpatient 4 mo Not blinded Fluoxetine (60 mg) vs Yes No change in scores
amineptine (300 mg)
1995 Klibanski et al23 48 0 11-27 Outpatient 1.5 y No blind Estrogen (0.625 mg) progestin No No change in scores
reported, not (5 mg) vs nonmedication
placebo control
controlled
1995 Szmukler et al24 29 0 20-23 Inpatient NR Double-blind, Cisapride (10 mg tid) vs No No improvement
placebo placebo
controlled
1998 Attia et al25 33 0 16-45 Inpatient 7 wk Double-blind, Fluoxetine (60 mg) vs placebo No No change
placebo
controlled
1999 Ricca et al26 24 0 16-22 Outpatient 6 mo No blind Fluoxetine (40 mg) vs No Venlafaxine was as
reported, not venlafaxine (75 mg) effective as
placebo fluoxetine

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


controlled
2000 Hill et al27 14 1 12-18 Inpatient 28 d Double-blind, Growth hormone 15 mg/kg vs Yes Improved orthostasis only
placebo placebo no P value reported
controlled
2001 Kaye et al28 39 0 15-32 Inpatient and 52 wk Double-blind, Fluoxetine (20-60 mg) vs No No changes
outpatient placebo placebo
controlled
2001 Ruggiero et al29 35 0 19-30 Inpatient 3 mo Single-blind, Amisulpride (50 mg) vs No No significant difference
not placebo fluoxetine (28 mg) vs
controlled clomipramine (60 mg)
2002 Fassino et al30 39 0 18-32 Outpatient 12 wk No blind Citalopram (20 mg) vs waitlist No Improved depression,
reported, not obsessive-compulsive
placebo disorder, impulsiveness,
controlled and anger

(continued)

145
146
Table 2. (continued)
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
31
2004 Barbarich et al 26 0 17-29 NR 6 mo Double-blind, Fluoxetine (20-60 mg) vs No No significant difference
not placebo supplements
controlled
2005 Miller et al32 38 0 18-37 Outpatient 3 wk Double-blind, Testosterone (150-300 mcg No Improved mood and spatial
placebo transdermal) vs placebo cognition (P = .002)
controlled
2005 Mondraty et al33 15 0 18-32 Inpatient 70 d Open label, Olanzapine (20 mg) vs No Possible decrease in
no placebo chlorpromazine (200 mg) ruminations
control
2006 Walsh et al34 93 0 16-45 Inpatient 52 wk Double-blind, Fluoxetine (60 mg) vs placebo Yes No benefit
placebo
controlled
2007 Brambilla et al35 30 0 18-36 Outpatient 3 mo Double-blind, Olanzapine (5 mg) vs placebo Yes No statistically significant
placebo difference
controlled

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


2007 Brambilla et al36 20 0 19-28 Outpatient 3 mo Single-blind Olanzapine (5 mg) vs placebo Yes No statistically significant
difference

BMI, body mass index; NR, not reported.


Pharmacotherapy of Eating Disorders / Jackson et al 147

the trial. Three patients were dropped from the study with placebo in 30 patients. All patients also received
because the blind was opened early, and 1 dropped out cognitive behavioral therapy. Although olanzapine was
because of an automobile accident. Six patients dropped associated with an increase in weight, there were no sta-
out of the fluoxetine group: 3 were dropped because of tistically significant differences between the 2 groups.35
weight loss, 1 owing to persistent depression, and 2 as a
result of persistent eating disorder. Sixteen patients were
Hormones
dropped from the placebo group: 4 for substantial weight
loss, 2 for depressive symptoms, and 10 for persistent eat- Both growth hormone and testosterone were compared
ing disorder symptoms. In the group that remained, fluox- with placebo in the treatment of AN. In a 28-day random-
etine was associated with improvement in depression, ized, double-blind, placebo-controlled trial, Hill evaluated
obsessive-compulsive features, and eating-disorder related 15 mg/kg/day of growth hormone in adolescent (age
symptoms, although statistically it did not differ signifi- 12-18) inpatients with AN.27 The primary outcome in this
cantly from placebo.28 trial included time to medical stability, weight gain, and
Probably the largest and most important of the fluox- duration of hospitalization. No patients dropped out of
etine trials for AN was by Walsh et al34 in 2006. In this the study. Although there were no significant changes in
randomized, double-blind trial, 93 inpatient/day treat- weight, patients receiving growth hormone experienced
ment patients were randomly assigned to either fluoxet- improvement in medical stability more rapidly than
ine 60 mg (n = 49) or placebo (n = 44) over the course patients taking placebo (17 vs 37 hospital days, P = .02).
of 1 year. The primary outcome was time to relapse and In 2005, Miller and colleagues compared testosterone
to determine the proportion of patients completing 1 year transdermal patches with placebo in 38 women between
of treatment. Similar percentages of patients assigned to the ages of 18 and 37. They hypothesized that bone forma-
fluoxetine (26.5%) and placebo (31.5%) maintained a tion would increase and depressive symptoms and spatial
body mass index (BMI) of 18.5 and stayed in the study cognition would improve with short-term testosterone
for 52 weeks. There was no difference between groups in administration. Multiple studies of testosterone in humans
the time to relapse (P = .64).34 have found improvement in mood with increased testoster-
one levels. Five patients (13%) dropped out of the study,
Tricyclic Antidepressants. There are 2 double-blind, including 1 who became pregnant, 1 with multiple frac-
placebo-controlled trials of amitriptyline for the treatment tures from a car accident, 2 with life-threatening weight
of AN. The first trial, conducted by Biederman et al, was loss, and 1 with nausea. The researchers concluded that
a 5-week trial of patients with AN receiving either amitrip- patients using the transdermal patches had improvements
tyline (n = 11) in doses up to 175 mg daily or placebo in mood, but no improvements in weight over placebo.32
(n = 14). Eighteen additional patients served as a control
group and received only psychosocial treatment. No
patients dropped out of this trial; however, there was no Zinc
significant difference between the groups, although Open trials have suggested that zinc may cause weight
patients in the amitriptyline group complained of uncom- gain in patients with AN. In 1994, Birmingham and col-
fortable side effects.19 leagues compared the effects of elemental zinc (n = 16)
Halmi and colleagues randomly assigned 72 inpa- 14 mg daily with placebo (n = 19) in a double-blind,
tients with AN to amitriptyline 160 mg, cyproheptadine placebo-controlled trial of 35 inpatients with AN. Patients
32 mg, or placebo in a 90-day trial. Although weight gain remained in the hospital until they achieved a 10%
was higher in the cyproheptadine group and time to increase in their BMI. Patients in the zinc group were
achieve weight gain was shorter in both medication hospitalized for a mean of 24.5 days, whereas patients in
groups than in the placebo group, the results were not the placebo group were hospitalized for a mean of 31.5
statistically significant.20 days (P = .03). Nineteen patients (30%) did not complete
the study. Ten of the patients who dropped out of the
Antipsychotics study were in the zinc group, and 9 were in the placebo
group. The reasons for patients dropping out of the study
Although there are many case reports using antipsychot- were not reported.21
ics to treat eating disorders, few to date have been blinded
or compared with placebo. In 1984, Vandereycken com-
pared sulpiride with placebo in a small study of 18
Summary of Medication Trials for AN
patients with AN. Initially, sulpiride improved weight gain
in patients; however, in the crossover phase of the study, Eleven of 19 trials (Table 2) compared medication with
sulpiride did not differentiate from placebo.18 In a study placebo in the treatment of AN.18-21,24-25,28,3234-35 Many of
by Brambilla and colleagues, olanzapine was compared the trials involving medication treatment of AN have

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


148 Nutrition in Clinical Practice / Vol. 25, No. 2, April 2010

significant limitations, including small trial size, lack of depression, substance abuse, anxiety disorders, and per-
adequate power, short duration, high dropout rates, and sonality disorders. People with BN have many of the same
lack of intention-to-treat analysis. Many of the trials did personality features as do patients with AN, including per-
not list the ethnicity of the patients involved in the trials, fectionism and rigid cognitive traits. They may also have
and only 1 male was involved in all of the AN trials. In low self-esteem, higher impulsivity, increased novelty seek-
addition, studies completed in inpatient units may not be ing, and lower levels of cooperativeness.15,36
generalized to the outpatient experience. Overall, no
medication was found to be significantly better than pla-
cebo for the treatment of AN. Medication Treatment for BN
The initial goal for treating BN is to decrease and eventu-
ally eliminate binging and purging behaviors and to treat
Bulimia Nervosa any medical complications. Then, as with AN, the goal is
to educate the patient on healthful eating habits and to
Bulimia nervosa (BN) usually begins in adolescence or help the patient incorporate those habits into his or her
early adulthood and is characterized by recurrent epi- daily routine and correct any misperceptions about eat-
sodes of binge eating marked by rapid consumption of ing, weight, and other thoughts and behaviors related to
large amounts of food over a distinct period of time. the eating disorder. It is important to treat any underlying
These patients feel that they have lost control over their psychiatric conditions, including impulse control, self-
eating behaviors during the binges and use purging meth- esteem issues, and unusual eating behaviors. Lastly, it is
ods, such as self-induced vomiting, laxatives or diuretics, important to get the family involved and prevent the
fasting, strict dieting, or excessive exercise to prevent patient from relapsing.16
associated weight gain. In addition, body shape and Although several antidepressants are effective for the
weight are constant concerns.6 treatment of BN, only fluoxetine is Food and Drug
To meet criteria for BN, patients must have an aver- Administration (FDA) approved for the treatment for
age of 2 binge-eating episodes per week for at least 3 BN.37,38 In fact, fluoxetine is the only medication to
months. Binging involves eating until the person experi- receive FDA approval for the treatment of any eating dis-
ences abdominal discomfort. Other reasons for terminat- order. There are 25 studies using medications to treat
ing the binge include stopping to sleep, social interruptions, bulimia,39-63 19 of which are double-blinded and placebo-
or induced vomiting. Vomiting often decreases the feeling controlled. Although including patient demographics,
of discomfort associated with abdominal distention, study design, duration, and outcomes for each of the 25
which may allow the person to continue with the binge or studies are listed in Table 3, we will review only the ones
to terminate it. Vomiting also relieves the anxiety fre- that are double-blinded and placebo controlled.
quently associated with binging, although feelings of
depression and self-criticism frequently remain.6
Antidepressants
BN can often remain hidden from friends and fami-
lies, since people with this eating disorder are within a Sixteen of the 20 studies reviewed are antidepressant tri-
normal weight range. People with BN may feel conflicted als.39-41,43,44,46-53,54,59,63 Seven of those trials involve fluox-
about eating, and fluctuations in weight are common etine, 3 desipramine, 1 trazodone, 1 brofaromine, 2
owing to repeated binges and fasts or purging. fluvoxamine, 1 moclobemide, and 1 citalopram.
There is a lower incidence of mortality with BN than
with AN. Approximately 1% of patients die within the first
Selective Serotonin Reuptake Inhibitors
10 years of diagnosis.10 This number may be artificially
low, considering that as many as 50% of people with AN Fluoxetine. Fluoxetine was studied in 7 trials. Three fluox-
evolve into BN during the course of their illness, but they etine trials of between 8 and 16 weeks in length by the
continue to carry the diagnosis of AN.10,11 Serious medical Fluoxetine Collaborative Group and Goldstein and col-
complications can occur from self-inducing vomiting and leagues showed improvement in binging, purging, and
use of laxatives and/or diuretics, including electrolyte bulimia symptoms.43,46,53 Eight-week trials by Kanerva et
imbalances and dehydration. Other rare complications al, Beaumont et al, and Romano et al also showed im-
include esophageal tears and gastric ruptures. Dental ero- provement in symptoms of bulimia.47,49,59 Sixty milligrams
sion is commonly seen as a result of the vomiting, as are of fluoxetine was better than 20 mg in these studies. In a
enlarged parotid glands.11 4-week trial by Fichter et al, fluoxetine 60 mg did not dif-
Comorbid psychiatric disorders are seen in as many as ferentiate from placebo, although this could have been a
80% of patients with BN. Common comorbidities include result of the short duration of the trial.40

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Table 3. Bulimia Nervosa: Medication Trials
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
39
1989 Pope et al 46 0 18-55 Outpatient 6 wks Double-blind, Trazodone (50 mg No Trazodone associated with a decrease in
placebo titrated to 400 binge and vomiting frequency (P < .05)
controlled mg) vs placebo
1991 Fichter et al40 39 1 25 mean Inpatient 35 d Double-blind, Fluoxetine (60 mg) Yes No differences on any measures
placebo vs placebo
controlled
1991 Walsh et al41 78 0 20-32 Outpatient 6 wk Double-blind, Desipramine (200- Yes Desipramine associated with fewer
placebo 300 mg) vs binging and vomiting episodes per
controlled placebo week; no P value reported
1992 Agras et al42 71 0 20-38 Outpatient 16 wk No blind Desipramine (300 Yes Desipramine and CBT better than CBT
reported, no mg) vs alone in reducing binge-and-purge
placebo desipramine (300 frequency; no P value reported
control mg) + CBT or
CBT alone
1992 Fluoxetine 387 0 19-35 Outpatient 8 wk Double-blind, Fluoxetine (20 mg) No Fluoxetine 60 mg associated with
Collaborative placebo vs fluoxetine (60 decreased binging, purging and
Study Group43 controlled mg) vs placebo bulimia; no P value reported
1993 Kennedy et al44 36 0 20-35 Outpatient 8 wk Double-blind, Brofaromine (175 No Brofaramine associated with a reduction
placebo mg) vs placebo in vomiting; no P value reported
controlled
1994 Agras et al45 71 0 20-38 Outpatient 16 wk No blind Desipramine (300 Yes See Agras et al42 1 year follow-up study;
reported, no mg) vs no P value reported
placebo desipramine (300
control mg) + CBT or

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


CBT alone
1995 Goldstein et al46 393 5 17-63 Outpatient 16 wk Double-blind, Fluoxetine (20 mg) No Fluoxetine associated with greater
placebo vs fluoxetine (60 decline in vomiting (P < .001) and
controlled mg) vs placebo binge eating (.002) and a decrease in
overall bulimia symptoms
1995 Kanerva et al47 50 0 >15 Outpatient 8 wk Double-blind, Fluoxetine (60 mg) No Fluoxetine associated with a greater
placebo vs placebo decline in symptoms of bulimia and
controlled preoccupation with food; no P value
reported
1996 Fichter et al48 257 0 19-30 Inpatient 3 wk Double-blind, Fluvoxamine (182 Yes Fluvoxamine associated with higher
placebo mg average) vs binge abstinence, lower relapse rates,
controlled placebo and a decreased urge to binge; no P
value reported
(continued)

149
150
Table 3. (continued)
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
49
1997 Beaumont et al 67 0 19-31 Outpatient 8 wk Double-blind, Fluoxetine (20 mg Yes Improvement in concern about shape
placebo tid) vs placebo and weight at week 8
controlled
1997 Fichter et al50 257 0 19-30 Outpatient 12 wk Double-blind, Fluvoxamine (182 Yes See Fichter et al42; continuation of
placebo mg) vs placebo 3-week study; no P value reported
controlled
1997 Goldbloom et al51 76 0 20-31 Outpatient 6 wk No blind Fluvoxamine (300 Yes CBT alone and CBT + fluvoxamine
reported, no mg) vs CBT vs associated with a greater reduction
placebo Fluvoxamine (300 than fluvoxamine alone; no P value
control mg) + CBT reported
1997 Walsh et al52 120 0 20-33 Outpatient 16 wk Double-blind, CBT + DMI (300 Yes CBT better than supportive therapy;
placebo mg) vs CBT + CBT + meds was better than CBT
controlled placebo vs DMI alone; meds alone was better than
(300 mg) vs CBT alone for decreasing BMI and
supportive care + weight; meds alone was better for
DMI (300 mg) vs binging and vomiting frequency
supportive care + (P = .03)
placebo
1999 Goldstein et al53 393 5 17-63 Outpatient 16 wk Double-blind, Fluoxetine (20 mg) Yes See Goldstein et al40; same study,
placebo vs fluoxetine (60 reported additional measures
controlled mg) vs placebo (P = .05)
1999 Wilson et al54 120 0 20-33 Outpatient 16 wk Double-blind, CBT + DMI (300 Yes See Walsh et al52; same study, reported
placebo mg) vs CBT + additional domains, therapeutic
controlled placebo vs DMI alliance (P = .03), predictive factors

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


(300 mg) vs (P = .03), and time course to change
supportive care + (P = .03)
(DMI 300 mg) vs
supportive care +
placebo
2000 Faris et al55 29 0 23-35 Outpatient 4 wk Double-blind, Ondansetron (4 mg Yes Ondansetron associated with lower
placebo up to 6 doses a binge/purge frequency at week 4
controlled day) vs placebo (P < .001)
2001 Carruba et al56 78 0 19-40 Inpatient 6 wk Double-blind, Moclobemide (600 No Moclobemide is not effective
placebo mg) vs placebo
controlled

(continued)
Table 3. (continued)
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
57
2001 Mitchell et al 48 0 18-46 Outpatient 16 wk Single-blind Placebo vs self- Yes Fluoxetine alone and fluoxetine with
help vs fluoxetine self-help is better than self-help or
(60 mg) vs placebo as far as decrease in vomiting;
fluoxetine (60 no P value reported
mg) + self-help
2002 Mitchell et al58 62 0 20-35 Outpatient 16 wk No blind Fluoxetine (60 mg) Yes No difference
reported, no vs IPT
placebo
control
2002 Romano et al59 155 5 21-40 Outpatient 8 wk Double-blind, Fluoxetine (60 mg) Yes Fluoxetine is associated with smaller
placebo vs placebo increase in eating disorder behavior
controlled (P < .02)
2003 Hoopes et al60 68 1 19-40 Outpatient 10 wk Double-blind, Topiramate (100 Yes Topiramate associated with greater
placebo mg) vs placebo percent reduction in binging and
controlled purging days, carb craving, 50%-75%
improvement in binging and purging
(P = .015)
2003 Hedges et al61 68 1 19-40 Outpatient 10 wk Double-blind, Topiramate (100 Yes Same as Hoopes e al60; reported
placebo mg) vs placebo psychiatric measures
controlled
2004 Walsh et al62 91 0 22-39 Outpatient 4 mo No blind Fluoxetine (60 mg) Yes Fluoxetine and placebo are better than
reported, vs placebo vs self- self-help; no P value reported

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


placebo help vs fluoxetine
controlled (60 mg) + self-
help
2005 Sundblad et al63 46 0 NR Outpatient 12 wk Double-blind, Flutamide (500 Yes Decrease binding with flutamide but not
placebo mg) vs citalopram citalopram, increase in transaminases
controlled (40 mg) vs in flutamide patients
placebo or
flutamide (500
mg) + citalopram
(40 mg)

BMI, body mass index; NR, not reported; CBT, Cognitive Behavioral Therapy; DMI, desipramine.

151
152 Nutrition in Clinical Practice / Vol. 25, No. 2, April 2010

Citalopram. Sundblad and colleagues (2005) compared 600 mg or placebo for 6 weeks and found that moclobe-
flutamide 500 mg to citalopram 40 mg and to placebo vs mide was not effective for symptoms of bulimia.56 Patients
a combination of flutamide 500 mg vs citalopram 40 mg did not have difficulty adhering to the MAOI diet in either
in a double-blind, placebo-controlled trial of 46 female trial.
outpatients over a period of 12 weeks. Flutamide appeared
to decrease binging, but also to increased serum liver
Anticonvulsants
transaminases. At a dose of 40 mg, citalopram did not
have any significant effect on symptoms of bulimia.63 Topiramate. In 2003, Hoopes performed a 10-week trial
comparing effects of topiramate 100 mg with placebo.
Compared with placebo, topiramate was associated with
Fluvoxamine. Fichter et al (1996, 1997) compared fluvox- a greater decline in binging and purging days, a decline
amine 182 mg (mean dose per day) with placebo in 257 in patient carbohydrate craving, and a 50%-75% improve-
inpatients for 3 weeks and found no improvements in ment in binge purge days.60 Hedges reported the results
symptoms of bulimia, but they then extended the trial to of the psychological measures (Eating Disorder Inven-
12 weeks post-discharge to evaluate the role of fluvox- tory [EDI], Eating Attitudes Test [EAT], Hamilton Rating
amine in relapse prevention. They found that patients Scale for Anxiety [HAM-A], Hamilton Rating Scale for
who received fluvoxamine had fewer binge episodes, less Depression [HAM-D] and Patient Global Improvement
of an urge to binge, and fewer episodes of vomiting than [PGI]) from this same study.61
patients receiving placebo.48,50
5HT 3 Antagonist
Ondansetron. Faris and colleagues performed a 4-week,
Tricyclic Antidepressants double-blind, placebo-controlled trial in 29 female pa-
Walsh (1991) and coauthors published results of a trial tients comparing ondansetron (n = 14) to placebo (n =
using desipramine vs placebo for 6 weeks and found 12). Ondansetron was dosed 4 mg up to 6 times daily dur-
desipramine was associated with reduced binging and ing the 4 weeks. The authors found that ondansetron was
vomiting episodes each week.41 Another study by Walsh, associated with a lower binge/purge frequency (6.5 times
Wilson, and colleagues in 1997 randomized patients to 1 per week) vs placebo (13.2 times per week) at the comple-
of 4 groups: cognitive behavioral therapy (CBT) and tion of the study.55
desipramine, CBT and placebo, desipramine and support-
ive therapy, or supportive therapy and placebo. They Summary of Medication Trials for BN
found that CBT was better than supportive therapy, and
CBT and desipramine was better than CBT alone. Antidepressant medications have shown significant ben-
According to the authors, desipramine was significantly efit in the treatment of BN. To date, fluoxetine is the only
superior for reducing BMI as well as binging and vomit- medication that is FDA-approved for this indication.
ing frequency.52 Wilson and Walsh published the results Studies have shown that at a dose of 60 mg daily, fluox-
of the 1997 trial again in 1999 along with additional etine administered for at least 6 weeks can improve symp-
domains not reported in the original trial.54 toms of BN, as well as prevent relapse for up to 52 weeks.
Trazodone, fluvoxamine, desipramine, brofaromine, topi-
Trazodone. In a 6-week double-blind, placebo-controlled ramate, and ondansetron have shown similar benefits to
trial of 42 females, Pope and colleagues found that trazo- fluoxetine, but larger studies are needed to replicate these
done, in doses titrated up to 400 mg daily, was associated findings.
with a decrease in frequency of binging and purging (P <
.05). There were no improvements in patient depression
and anxiety scores at the end of the trial.39 Binge-eating Disorder

Currently, binge-eating disorder (BED) belongs to the


DSM-IV-TR diagnostic criteria of Eating Disorders Not
Monoamine Oxidase Inhibitors (MAOIs)
Otherwise Specified (EDNOS). Eating disorders in the
In 1993, Kennedy et al conducted an 8-week trial of bro- diagnostic category of EDNOS may be similar to the pre-
faromine at a mean dose of 175 mg daily vs placebo and viously discussed eating disorders, but they do not meet
found that it was associated with a reduction in vomiting.44 the full diagnostic criteria. EDNOS are treated in much
In 2001, Carruba et al randomized patients to moclobemide the same way as the disorders that meet full criteria. Of

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Pharmacotherapy of Eating Disorders / Jackson et al 153

the eating disorders in the EDNOS category, provisional of CBT and placebo. The authors found that CBT was su-
diagnostic criteria were established for BED, and it will perior to fluoxetine alone or placebo alone in decreasing
likely be a stand-alone eating disorder when DSM V is eating behaviors and binging and in decreasing concerns
published in 2012.9 over shape. Although the combination of CBT and fluox-
Like BN, BED is marked by binge eating, but without etine was better than placebo alone, the combination was
the compensatory behaviors such as purging, excessive not found to be better than CBT alone.75
exercise, or laxative or diuretic abuse. BED appears to
have an older age at onset, similar male-to-female ratio of Fluvoxamine. In 1998, Hudson and coworkers66 com-
occurrence, and more ethnic diversity. BED is character- pared fluvoxamine, at an average dose of 300 mg, with
ized by low self-esteem, depression, and dissatisfaction placebo in a 9-week trial involving 85 patients. The inves-
with body shape and image. In addition to depression, tigators found that fluvoxamine decreased binge fre-
another common comorbidity of BED includes obesity.6 quency, decreased binge severity, and improved BMI over
the course of the trial.66 In a second 12-week study of 20
patients in which fluvoxamine was compared with pla-
Medication Treatment for BED
cebo,74 no differences emerged for binge frequency, shape
With BED, there are typically 2 primary goals of treat- and weight concerns, or weight changes. This study by
ment. The first is to decrease binging behaviors with a Pearlstein et al, completed in 2003, used an average flu-
goal of eliminating them, and the other goal is to work voxamine dose of 239 mg daily.74 It is possible that the
with the patient to help him or her lose weight.16 lower average fluvoxamine doses used in the Pearlstein
Antidepressants used at higher doses than are typi- trial are responsible for the lack of efficacy, but because
cally used for depression are effective in decreasing bing- there is no long-term follow-up, findings of both studies
ing behaviors, but they are not effective for any significant should be viewed with caution.
weight loss. Appetite suppressants and topiramate have
also been studied in BED and may have promising roles.
Thirteen medication-related studies were conducted for Sertraline. In a 6-week trial of sertraline at an average dose
the treatment of BED (see Table 4).64-76 Eleven of these of 187 mg/day, McElroy and colleagues (2000)68 found
are double-blind, placebo-controlled trials.65-68,70-76 that compared with placebo, sertraline decreased binge
frequency and improved body weight. The authors also
Antidepressants suggested that illness severity declined, but there were no
changes in depressed mood compared with placebo.
Seven of the 11 trials reviewed are antidepressant tri- Twenty-eight percent of the patients in the sertraline group
als.66-68,70,72,74-75 Two involved fluoxetine, 2 involve fluvox- and 19% of the patients in the placebo group dropped out
amine, and 1 each for sertraline, citalopram, and of the study.68
imipramine.

Selective Serotonin Reuptake Inhibitors Citalopram. McElroys group also compared citalopram
Fluoxetine. Fluoxetine was studied in 2 trials for BED. 40 mg with 60 mg and with placebo in 38 patients in
2003. Citalopram caused a decline in binge eating and
The first trial, by Arnold and colleagues in 2002, was a
binge-eating days, and also caused a decrease in BMI. In
6-week outpatient trial of fluoxetine (n = 30) in doses of
this 6-week study, patients lost 2.7 kg, whereas patients in
60-80 mg daily compared with placebo (n = 30).70 In this
the placebo group gained 5.2 kg. The dropout rate in this
trial, fluoxetine was associated with a decrease in the se-
study was 16% in the citalopram group (1 for worsening
verity of BED, reduced weight gain, and increased con-
depression, 1 for sexual dysfunction, 1 for a personal
trol over the patients BMI over the course of 6 weeks.
conflict) and 21% for the placebo group (3 for worsening
In a second 16-week trial of 108 patients by Grilo and
depression, 1 for noncompliance).72
coworkers in 2005, patients received fluoxetine 60 mg
or placebo alone, fluoxetine 60 mg plus CBT, or placebo
plus CBT. The primary outcome was based on remission
rates, which were defined as no binges occurring within a Tricyclic Antidepressants
28-day period. Eighty-six patients (80%) completed treat-
ment. Remission, as defined, occurred in 29% of fluox- Imipramine. In a trial by Laederach-Hoffman and colleagues
etine patients, 30% of placebo patients, 55% of patients on in 1999, imipramine 25 mg three times daily or placebo
CBT and fluoxetine, and 73% of patients on a combination was added to a treatment regimen of diet counseling and

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


154
Table 4. Binge Eating Disorder: Medication Trials
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
1994 Agras et al64 109 0 22-65 Outpatient 9 mos Additive design, Weight loss No CBT was better than weight loss
no blind treatment vs weight alone in decreasing binge
reported, not loss + CBT vs frequency; CBT + desipramine
placebo- weight loss + CBT was better at decreasing weight;
controlled + desipramine (300 no p value reported
mg)
1996 Stunkard et al65 28 0 NR Outpatient 8 weeks Double-blind, d-fenfluramine (15 NR d-fenfluramine decreased the
placebo- mg) vs placebo frequency of binge eating
controlled
1998 Hudson et al66 77 8 31-52 Outpatient 9 weeks Double-blind, Fluvoxamine (300 Yes Fluvoxamine decreased binge
placebo- mg) vs placebo frequency (P < .006), severity
controlled (p < .04) and BMI (P< .002)
over 9 weeks
1999 Laederach- 27 4 25-50 Outpatient 32 weeks Double-blind, Imipramine (25 mg No Imipramine decreased binge
Hoffman et al67 placebo- tid) vs placebo frequency, illness severity, BMI
controlled and weight over 14 weeks
(P< .001)
2000 McElroy et al68 32 2 29-53 Outpatient 6 weeks Double-blind, Sertraline (187 mg Yes Sertraline was superior in
placebo- mean) vs placebo decreasing binge frequency (P<
controlled .008), illness severity (P< 0.01),
and BMI (P< .01)
2001 Ricca et al69 64 44 19-33 Outpatient 18 mos Parallel series, CBT vs CBT + Yes CBT was more effective than
open label, not fluoxetine (60 mg) fluoxetine or fluvoxamine;

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


placebo- vs CBT vs fluoxetine did not enhance the
controlled fluvoxamine (300 CBT while fluvoxamine did
mg) vs fluoxetine seem to enhance CBT
(60 mg) vs (P< .01)
fluovoxamine (300
mg)
2002 Arnold et al70 56 4 31-55 Outpatient 6 weeks Double-blind, Fluoxetine Yes Fluoxetine was associated with a
placebo- (60-80 mg) vs decrease in illness severity (P=
controlled placebo 0.32), less weight gain (P<
.001), and controlled BMI over
6 weeks
(continued)
Table 4. (continued)
Intention to
Year Source Females Males Age (y) Location Length Design Treatment Treat Outcome
2003 Appolinario 53 7 26-46 Outpatient 12 weeks Double-blind, Sibutramine (15 mg) Yes Sibutramine decreased binge
et al71 placebo- vs placebo frequency and severity (P< .03),
controlled weight decline in sibutramine
group (P< .001)
2003 McElroy et al72 36 2 38-52 Outpatient 6 weeks Double-blind, Citalopram (60 mg) Yes Citalopram decreased binge days
placebo- vs placebo (P< .001), BMI (P< .001) and
controlled weight (P< .005)
2003 McElroy et al73 61 0 31-50 Outpatient 16 weeks Double-blind, Topiramate (212 mg) Yes Topiramate superior in decreasing
placebo- vs placebo binge frequency (P< .0004),
controlled illness severity (P< .02), BMI,
and weight over the length of the
study (P< .005)
2003 Pearlstein et al74 17 3 41 mean Outpatient 12 weeks Double-blind, Fluvoxamine (239 NR No difference between
placebo- mg) vs placebo fluvoxamine and placebo; no
controlled weight change over time
2005 Grilo et al75 84 24 21-59 Outpatient 16 weeks Double-blind, Fluoxetine (60 mg) Yes CBT superior to fluoxetine alone
placebo- vs placebo vs and placebo alone in decreasing
controlled CBT + placebo vs eating shape concerns and
CBT + fluoxetine binging; CBT + fluoxetine better

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


60 mg) than placebo alone; CBT and
fluoxetine not better than CBT
alone; no P values given
2005 Grilo et al76 44 6 35-58 Outpatient 12 weeks Double-blind, CBT + orlistat (120 Yes CBT + orlistat was superior in
placebo- mg) vs CBT + total weight loss over 12 weeks;
controlled placebo no P values given

BMI, body mass index; NR, not reported; CBT, cognitive behavioral therapy.

155
156 Nutrition in Clinical Practice / Vol. 25, No. 2, April 2010

sychological support in 31 patients. Patients were assessed


p percent of patients in the orlistat group and 20% in the
at 8 weeks and again at 32 weeks. Patients in the imipramine placebo group dropped out of the study.76
group experienced reductions in depressed mood, binges,
and body weight at both 8 and 32 weeks, whereas those in
the placebo group had an increase in weight. Dropout rates D-fenfluramine. In an 8-week trial of 28 patients by
were relatively low: 6% for imipramine and 7% for placebo.67 Stunkard and colleagues in 1996, d-fenfluramine at a
dose of 15 mg daily was found to decrease the frequency
Anticonvulsants of binge eating when compared with treatment with pla-
cebo.65 Because of concerns about pulmonary hyperten-
Topiramate. Topiramate, at an average dose of 212 mg sion and deformities in heart valves when taken in
daily, was studied in 61 females in a 16-week trial by combination with phenteramine, d-fenfluramine is no
McElroy and colleagues in 2003. Patients receiving topi- longer available in the United States or in global markets.
ramate had a greater decline in binge frequency, illness se-
verity, and BMI vs patients receiving placebo. The dropout
rate in this trial was high (topiramate 47%, placebo 39%).
Reasons for dropping out of the study included patient Summary of Medication Trials for BED
choice (11 topiramate, 21 placebo), adverse effects (28 There is good evidence that antidepressants, especially
topiramate, 15 placebo), lack of efficacy (1 topiramate, 3 SSRIs, can improve BED in terms of reducing binge-eat-
placebo), lost to follow-up (13 topiramate, 17 placebo), ing behaviors, at least in the short term. The antidepres-
and other reasons (2 topiramate, 3 placebo). The high sants do not appear to be associated with long-term
dropout rate makes it difficult to evaluate the significance significant weight loss. The case for the appetite suppres-
of the results.73 sant sibutramine is better in terms of both reducing
binge-eating behaviors and causing weight loss.
Appetite Suppressants Topiramate may lead to improvements with binging and
weight loss as well, although more studies are necessary
Sibutramine. Sibutramine (n = 23) 15 mg daily, was com- to confirm these results.
pared with placebo (n = 30) in a 12-week trial by Appoli-
nario in 2003.71 The primary outcome was binge frequen-
cy. In this trial, sibutramine caused a significant decrease Conclusion
in binge frequency and severity. Patients also reported im-
provement in feelings of depression. At the end of the trial, The eating disorders AN, BN, and BED are important,
the sibutramine group lost 7.4 kg, whereas the placebo complex, and potentially life-threatening disorders. The
group gained weight (1.4 kg). Twelve (52%) of 23 sibutra- first of these, AN, was not recognized as an official disor-
mine and 8 (32%) of 25 placebo patients attained remis- der until the mid-1980s. BED has not yet been classified
sion at week 12. Eighteen patients (78%) of 23 sibutra- as an official disorder, but there is a high likelihood that
mine and 13 (52%) of 25 placebo patients showed a 50% it will be in the near future.
reduction in binge frequency. The dropout rate in this trial Medication trials for the treatment of these disorders
was 23% for sibutramine and 17% for placebo. Four pa- are highly variable, and the literature is especially weak in
tients withdrew from the study because of adverse effects. the area of medication use in the treatment of AN.
In the sibutramine group, 1 patient withdrew for insomnia Managing patients with AN using medications alone is
and 1 for appendicitis. In the placebo group, 1 withdrew not appropriate. No study to date has shown efficacy for
because of nausea, and 1 because of headache.71 medications alone, and the medical complications of AN
and the mortality associated with it demand more effica-
Orlistat. In 2005, Grilo and colleagues compared CBT cious treatment. There is some evidence that the use of
and orlistat with CBT and placebo in 50 obese BED CBT in AN appears to decrease the incidence of relapse
patients in a 12-week trial. Orlistat was used at a dose of after weight restoration, and should be considered in
120 mg three times daily. The primary outcome was for addition to medical stabilization. In addition, the use of
no binges to occur over the course of 28 days. In this medications in the treatment of AN is associated with a
study, the authors reported that a combination of CBT high dropout rate, suggesting that patients do not find
and orlistat was associated with a more significant weight medications acceptable.
loss (64%) than CBT and placebo (36%) at the end of the Binging and purging, the core physiologic symptoms
study; however, at a 3-month follow-up, there was no dif- of BN, respond to higher doses of fluoxetine (60 mg/
ference in weight loss in either group (52%). Twenty-four day). Lower doses do not appear to work, and in fact they

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Pharmacotherapy of Eating Disorders / Jackson et al 157

may contribute to a relapse in symptomatology. There is addition, studies do not give enough attention to medical
some evidence that other antidepressants (trazodone, issues and the long-term sequelae of the eating disorders.
fluvoxamine, desipramine, MAOIs) and the anticonvul- It is clear that more research is needed in this area.
sant topiramate may also decrease the physiological Large-scale, independent studies using appropriate doses
symptoms of BN. To date, only 1 medication has been and duration of therapy, in addition to delineation of sex,
FDA approved to treat an eating disorder, and that is age, and ethnicity, need to be undertaken. Studies using
fluoxetine for the treatment of BN. appropriate statistical analysis, power, research design, and
There is good evidence that CBT is effective for BN. outcome measures are needed. For all 3 disorders, use of
There is some preliminary evidence that the combination novel therapeutic agents and comparison of medications to
of CBT and medication may be effective, but more psychotherapy would be important.
research is needed in this area. Compliance with CBT in There is a paucity of outcome studies in the area of
BN tends to be greater than with medications. Many eating disorders. Recovery, remission, and relapse need to
patients treated with medications for BN are noncompli- be defined and adequately researched to give eating dis-
ant in the long term, owing to adverse effects such as order patients and their caregivers information on treat-
sexual dysfunction and weight gain. ment for these life-long illnesses.
To date, most of the trials for BED have been short-
term trials. There is initial evidence that SSRIs, tricyclic
antidepressants, and sibutramine may be effective. CBT References
appears to decrease the primary target symptom of BED,
binging, but it does not help with weight loss. Initial evi- 1. Hoek HW, van Hoeken D. Review of the prevalence and incidence
of eating disorders. Int J Eat Disord. 2003;34:383-339.
dence using combination treatments of CBT and medica- 2. Walters EE, Kendler KS. Anorexia nervosa and anorexic-like syn-
tion appear to improve both binge eating and weight loss. dromes in a population-based female twin sample. Am J Psychiatry.
Patients with BED tend to have lower study dropout rates 1995;152:64-71.
and better compliance with medications than do the 3. Garfinkel PE, Lin E, Goering P, et al. Bulimia nervosa in a
other 2 eating disorders. Canadian community sample: prevalence and comparison of sub-
groups. Am J Psychiatry. 1995;152:1052-1058.
The studies that have been completed for the treat- 4. Reichborn-Kjennerud T, Bulik CM, Sullivan P, et al. Psychiatric
ment of the eating disorders have multiple limitations. and medical symptoms in binge eating in the absence of compen-
Studies do not distinguish between what is statistically satory behaviors. Obes Res. 2004;12:1445-1454.
and clinically significant. In addition, multiple assess- 5. Arnow B, Sanders M, Steiner H. Pre-pubertal vs. post-pubertal
ment tools make comparing trials difficult. Not having a anorexia nervosa: psychological characteristics. Clin Child Psychol
Psychiatry. 1999;4:403-416.
consistent definition of stage of illness, remission, recov- 6. American Psychiatric Association. Diagnostic and Statistical
ery, and relapse also limit the ability to compare outcome Manual of Mental Disorders. 4th ed. Washington, DC: American
data from trials. Psychiatric Association; 2000.
There is a tendency for lack of rigor in the scientific 7. Golden NH, Katzman DK, Kreipe RE, et al. Eating disorders in
design and statistical analysis in these trials. Lack of adolescents: position paper of the Society for Adolescent Medicine.
J Adolesc Health. 2003;33:496-503.
attention to baseline characteristics of subgroups, poorly 8. Katzman DK, Zipursky RB. Adolescents with anorexia nervosa: the
designed and/or reported randomization designs, insuffi- impact of the disorder on bones and brains. Ann N Y Acad Sci.
cient sample sizes, lack of sufficient power and reporting 1997;817:127-137.
of degrees of freedom, as well as lack of intention-to-treat 9. Wilfley DE, Bishop ME, Wilson GT, Agras WS. Classification of
analysis are all problems with many of the current studies eating disorders: toward DSM-V. Int J Eat Disord.
2007;40(Suppl):S123-S129.
that may bias results. 10. Sullivan P. Course and outcome of anorexia nervosa and bulimia
The high dropout rate inherent in psychiatric research nervosa. In: Fairburn CG, Brownell KD, eds. Eating Disorders and
also has a negative effect on study outcomes. Initial Obesity. 2nd ed. New York, NY: Guilford Press; 2002:226-232.
power analyses often did not take the high dropout rates 11. Sullivan PF. Mortality in anorexia nervosa. Am J Psychiatry.
into account. There were multiple reasons for dropouts in 1995;152:1073-1074.
12. Crook MA, Hally V, Panteli JV. The importance of the refeeding
these trials. The highest dropout rates occurred in the AN syndrome. Nutrition. 2001;17:632-637.
trials. Patients with AN often fear taking medication and 13. Brooks MJ, Melnik G. The refeeding syndrome: an approach to
the weight gain associated with it. These dropouts have understanding its complications and preventing its occurrence.
an effect on the integrity of outcome data. Pharmacotherapy. 1995;15:713-726.
Eating disorders create a life-long challenge for 14. Katzman DK. Medical complications in adolescents with anorexia
nervosa: a review of the literature. Int J Eat Disord. 2005;37:S52-S59.
patients, and long-term follow-up is important in eating 15. Strober M, Freeman R, Lampert C, et al. Controlled family study
disorder trials. Most trials that have been completed are of anorexia nervosa and bulimia nervosa: evidence of shared
of short duration, and it is inappropriate to expect that liability and transmission of partial syndromes. Am J Psychiatry.
these results can be extrapolated to long-term results. In 2000;157:393-401.

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


158 Nutrition in Clinical Practice / Vol. 25, No. 2, April 2010

16. Joel Yager J, Devlin M. J, Halmi KA. Practice Guideline for the 38. Crow SJ, Mitchell J, Roerig JD, et al. What potential role is there
Treatment of Patients with Eating Disorders. 3rd ed. June 2006. for medication treatment in anorexia nervosa. Int J Eat Disord.
Available at: http://www.psychiatryonline.com/pracGuide/loadGuide 2009;42:1-8.
linePdf. Accessed September 10, 2009. 39. Pope HG Jr, Keck PE Jr, McElroy SL, et al. A placebo controlled
17. American Psychiatric Association. Diagnostic and Statistical study of trazodone in bulimia nervosa. J Clin Psychopharmacol.
Manual of Mental Disorders. 3rd ed. Washington, DC: American 1989;9:254-259.
Psychiatric Association; 1980. 40. Fichter MM, Leibl K, Rief W, et al. Fluoxetine versus placebo: a
18. Vandereycken W. Neuroleptics in the short-term treatment of ano- double-blind study with bulimic inpatients undergoing intensive
rexia nervosa. A double-blind placebo-controlled study with sulpir- psychotherapy. Pharmacopsychiatry. 1991;24:1-7.
ide. Br J Psychiatry. 1984;144:288-292. 41. Walsh BT, Hadigan CM, Devlin MJ, et al. Long-term outcome of
19. Biederman J, Herzog DB, Rivinus TM, et al. Amitriptyline in the antidepressant treatment for bulimia nervosa. Am J Psychiatry.
treatment of anorexia nervosa: a double-blind, placebo-controlled 1991;148:1206-1212.
study. J Clin Psychopharmacol. 1985;5:10-16. 42. Agras WS, Rossiter EM, Arnow B, et al. Pharmacologic and
20. Halmi KA, Eckert E, LaDu TJ, Cohen J. Anorexia nervosa. cognitive-behavioral treatment for bulimia nervosa: a controlled
Treatment efficacy of cyproheptadine and amitriptyline. Arch Gen comparison. Am J Psychiatry. 1992;149:82-87.
Psychiatry. 1986;43:177-181. 43. Fluoxetine Bulimia Nervosa Collaborative Study Group. Fluoxetine in
21. Birmingham CL, Goldner EM, Bakan R. Controlled trial of zinc sup- the treatment of bulimia nervosa. A multicenter, placebo-controlled,
plementation in anorexia nervosa. Int J Eat Disord. 1994;15:251-255. double-blind trial. Arch Gen Psychiatry. 1992;49:139-147.
22. Brambilla F, Draisci A, Peirone A, et al. Combined cognitive- 44. Kennedy SH, Goldbloom DS, Ralevski E, et al. Is there a role for
behavioral, psychopharmacological and nutritional therapy in eat- selective monoamine oxidase inhibitor therapy in bulimia nervosa?
ing disorders II. Anorexia nervosa-binge-eating/purging type. A placebo-controlled trial of brofaromine. J Clin Psychopharmacol.
Neuropsychopharmacology. 1995;32:64-67. 1993;13:415-422.
23. Klibanski A, Biller BM, Schoenfeld DA, et al. The effects of estro- 45. Agras WS, Rossiter EM, Arnow B, et al. One-year follow-up of
gen administration on trabecular bone loss in young women with psychosocial and pharmacologic treatments for bulimia nervosa. J
anorexia nervosa. J Clin Endocrinol Metab. 1995;80:898-904. Clin Psychiatry. 1994;55:179-183.
24. Szmukler GI, Yung GP, Miller G, et al. A controlled trial of cis- 46. Goldstein DJ, Wilson MG, Thompson VL, et al. Long-term fluox-
apride in anorexia nervosa. Int J Eat Disord. 1995;17:347-357. etine treatment of bulimia nervosa. Fluoxetine Bulimia Nervosa
25. Attia E, Haiman C, Walsh BT, et al. Does fluoxetine augment the Research Group. Br J Psychiatry. 1995;166:660-666.
inpatient treatment of anorexia nervosa? Am J Psychiatry. 47. Kanerva R, Rissanen A, Sarna S. Fluoxetine in the treatment of
1998;155:548-551. anxiety, depressive symptoms, and eating-related symptoms in
26. Ricca V, Mannucci E, Paionni A, et al. Venlafaxine versus fluoxet- bulimia nervosa. Nord J Psychiatry. 1994;49:237-242.
ine in the treatment of atypical anorectic outpatients: a preliminary 48. Fichter MM, Kruger R, Rief W, et al. Fluvoxamine in prevention of
study. Eat Weight Disord. 1999;4:10-14. relapse in bulimia nervosa: effects on eating-specific psychopathol-
27. Hill K, Bucuvalas J, McClain C, et al. Pilot study of growth hor- ogy. J Clin Psychopharmacol. 1996;16:9-18.
mone administration during the refeeding of malnourished anorexia 49. Beaumont PJ, Russell JD, Touyz SW, et al. Intensive nutritional
nervosa patients. J Child Adolesc Psychopharmacol. 2000;10:3-8. counseling in bulimia nervosa: a role for supplementation with
28. Kaye WH, Nagata T, Weltzin TE, et al. Double-blind placebo- fluoxetine? Aust N Z J Psychiatry. 1997;31:514-524.
controlled administration of fluoxetine in restricting- and restrict- 50. Fichter MM, Leibl C, Kruger R, et al. Effects of fluvoxamine on
ing-purging-type anorexia nervosa. Biol Psychiatry. 2001;49:644-652. depression, anxiety, and other areas of general psychopathology in
29. Ruggiero GM, Laini V, Mauri MC, et al. A single blind comparison bulimia nervosa. Pharmacopsychiatry. 1997;30:85-92.
of amisulipride, fluoxetine and clomipramine in the treatment of 51. Goldbloom DS, Olmsted M, Davis R, et al. A randomized control-
restricting anorectics. Prog Neuropsychopharmacol Biol Psychiatry. led trial of fluoxetine and cognitive behavioral therapy for bulimia
2001;25:1049-1059. nervosa: short-term outcome. Behav Res Ther. 1997;35:803-811.
30. Fassino S, Leombruni P, Daga G, et al. Efficacy of citalopram in 52. Walsh BT, Wilson GT, Loeb KL, et al. Medication and psycho-
anorexia nervosa: a pilot study. Eur Neuropsychopharmacol. therapy in the treatment of bulimia nervosa. Am J Psychiatry.
2002;12:453-459. 1997;154:523-531.
31. Barbarich NC, McConaha CW, Halmi KA, et al. Use of nutritional 53. Goldstein DJ, Wilson MG, Ascroft RC, et al. Effectiveness of fluox-
supplements to increase the efficacy of fluoxetine in he treatment etine therapy in bulimia nervosa regardless of comorbid depres-
of anorexia nervosa. Int J Eat Disord. 2004;35:10-15. sion. Int J Eat Disord. 1999;25:19-27.
32. Miller KK, Grieco KA, Klibanski A. Testosterone administration in 54. Wilson GT, Loeb KL, Walsh BT, et al. Psychological versus phar-
women with anorexia nervosa. J Clin Endocrinol Metab. macological treatments of bulimia nervosa: predictors and proc-
2005;90:1428-1433. esses of change. J Consult Clin Psychol. 1999;67:451-459.
33. Mondraty N, Birmingham CL, Touyz S, et al. Randomized control- 55. Faris PL, Kim SW, Meller WH, et al. Effect of decreasing afferent
led trial of olanzapine in the treatment of cognitions in anorexia vagal activity with ondansetron on symptoms of bulimia nervosa: a
nervosa. Australas Psychiatry. 2005;13:72-75. randomized, double blind trial. Lancet. 2000;355:792-797.
34. Walsh BT, Kaplan AS, Attia E, et al. Fluoxetine after weight resto- 56. Carruba MO, Cuzzolaro M, Riva L, et al. Efficacy and tolerability
ration in anorexia nervosa: a randomized controlled trial. JAMA. of moclobemide in bulimia nervosa: a placebo-controlled trial. Int
2006;295:2605-2612. J Clin Psychopharmacol. 2001;16:27-32.
35. Brambilla F, Garcia CS, Fassino S, et al. Olanzapine therapy in 57. Mitchell JE, Fletcher L, Hanson K, et al. The relative efficacy of
anorexia nervosa: psychobiological effects. Int Clin fluoxetine and manual-based self-help in the treatment of outpatients
Psychopharmacol. 2007;22:197-204. with bulimia nervosa. J Clin Psychopharmacol. 2001;21:298-304.
36. Brambilla F, Monteleone M, Maj M. Olanzapine-induced weight 58. Mitchell JE, Halmi K, Wilson GT, et al. A randomized secondary
gain in anorexia nervosa; involvement of leptin and ghrelin secre- treatment study of women with bulimia nervosa who fail to
tion. Psychoneuroendocrinology. 2007;32:402-406. respond to CBT. Int J Eat Disord. 2002;32:271-281.
37. Alexander W. Prozac wins new indication for bulimia nervosa. 59. Romano SJ, Halmi KA, Sarkar NP, et al. A placebo-controlled
Drug Topics. 1997;141:6. study of fluoxetine in continued treatment of bulimia nervosa after

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010


Pharmacotherapy of Eating Disorders / Jackson et al 159

successful acute fluoxetine treatment. Am J Psychiatry. 2002;159: 68. McElroy SL, Casuto LS, Nelson EB, et al. Placebo controlled trial
96-102. of sertraline in the treatment of binge eating disorder. Am J
60. Hoopes SP, Reimherr FW, Hedges DW, et al. Treatment of bulimia Psychiatry. 2000;157:1004-1006.
nervosa with topiramate in a randomized, double-blind, placebo- 69. Ricca V, Mannucci E, Mezzani B, et al. Fluoxetine and fluvoxam-
controlled trial, part 1: improvement in binge and purge measures. ine combined with individual cognitive behavior therapy in binge
J Clin Psychiatry. 2003;64:1335-1341. eating disorder: a one-year follow-up study. Psychother Psychosom.
61. Hedges DW, Reimherr FW, Hoopes SP, et al. Treatment of bulimia 2001;70:298-306.
nervosa with topiramate in a randomized, double-blind, placebo- 70. Arnold LM, McElroy SL, Hudson JI, et al. A placebo controlled,
controlled trial, part 2: improvement in psychiatric measures. J randomized trial of fluoxetine in the treatment of binge-eating
Clin Psychiatry. 2003;64:1449-1454. disorder. J Clin Psychiatry. 2002;63:1028-1033.
62. Walsh BT, Fairburn CG, Mickley D, et al. Treatment of bulimia 71. Appolinario JC, Bacaltchuk J, Sichieri R, et al. A randomized, dou-
nervosa in a primary care setting. Am J Psychiatry. 2004;161: ble-blind, placebo-controlled study of sibutramine in the treatment
556-561. of binge-eating disorder. Arch Gen Psychiatry. 2003;60:1109-1116.
63. Sundblad C, Landen M, Eriksson T, et al. Effects of the androgen 72. McElroy SL, Hudson JI, Malhotra S, et al. Citalopram in the treat-
antagonist flutamide and the serotonin reuptake inhibitor citalo- ment of binge-eating disorder: a placebo controlled trial. J Clin
pram in bulimia nervosa: a placebo-controlled pilot study. J Clin Psychiatry. 2003;64:807-813.
Psychopharmacol. 2005;25:85-88. 73. McElroy SL, Arnold LM, Shapira NA, et al. Topiramate in the treat-
64. Agras WS, Telch CF, Arnow B, et al. Weight loss, cognitive- ment of binge eating disorder associated with obesity: a randomized,
behavioral, and desipramine treatments in binge eating disorder: placebo controlled trial. Am J Psychiatry. 2003;160:255-261.
an additive design. Behav Ther. 1994;25:225-238. 74. Pearlstein T, Spurell E, Hohlstein LA, et al. A double blind,
65. Stunkard A, Berkowitz R, Tanrikut C, et al. D-fenfluramine placebo-controlled trial of fluvoxamine in binge eating disorder: a
treatment of binge eating disorder. Am J Psychiatry. 1996;153: high placebo response. Arch Women Ment Health. 2003;6:147-151.
1455-1459. 75. Grilo CM, Masheb RM, Wilson GT. Efficacy of cognitive behavio-
66. Hudson JI, McElroy SL, Raymond NC, et al. Fluvoxamine in the ral therapy and fluoxetine for the treatment of binge eating disor-
treatment of binge-eating disorder: a multicenter placebo-controlled, der: a randomized double-blind placebo-controlled comparison.
double-blind trial. Am J Psychiatry. 1998;155:1756-1762. Biol Psychiatry. 2005;57:301-309.
67. Laederach-Hofmann K, Graf C, Horber F, et al. Imipramine and 76. Grilo CM, Masheb RM, Salant SL. Cognitive behavioral therapy
diet counseling with psychological support in the treatment of guided self-help and orlistat for the treatment of binge eating dis-
obese binge eaters: a randomized, placebo-controlled double-blind order: a randomized, double-blind, placebo-controlled trial. Biol
study. Int J Eat Disord. 1999;26:231-244. Psychiatry. 2005;57:1193-1201.

Downloaded from http://ncp.sagepub.com at AUBURN UNIV on May 4, 2010