Serotonin Neuronal Function and Selective Serotonin

Serotonin Neuronal Function and Selective Serotonin
Reuptake Inhibitor Treatment in Anorexia and Bulimia
Nervosa
Walter Kaye, Kelly Gendall, and Michael Strober
Anorexia nervosa (AN) and bulimia nervosa (BN) are
disorders characterized by aberrant patterns of feeding
behavior and weight regulation, and disturbances in
attitudes toward weight and shape and the perception of
body shape. Emerging data support the possibility that
substantial biologic and genetic vulnerabilities contribute
to the pathogenesis of AN and BN. Multiple neuroendocrine and neurotransmitter abnormalities have been documented in AN and BN, but for the most part, these
disturbances are state-related and tend to normalize after
symptom remission and weight restoration; however, elevated concentrations of 5-hydroxyindoleacetic acid in the
cerebrospinal fluid after recovery suggest that altered
serotonin activity in AN and BN is a trait-related characteristic. Elevated serotonin activity is consistent with
behaviors found after recovery from AN and BN, such as
obsessionality with symmetry and exactness, harm avoidance, perfectionism, and behavioral overcontrol. In BN,
serotonergic modulating antidepressant medications suppress symptoms independently of their antidepressant
effects. Selective serotonin reuptake inhibitors (SSRIs) are
not useful when AN subjects are malnourished and underweight; however, when given after weight restoration,
fluoxetine may significantly reduce the extremely high rate
of relapse normally seen in AN. Nonresponse to SSRI
medication in ill AN subjects could be a consequence of an
inadequate supply of nutrients, which are essential to
normal serotonin synthesis and function. These data raise
the possibility that a disturbance of serotonin activity may
create a vulnerability for the expression of a cluster of
symptoms that are common to both AN and BN and that
nutritional factors may affect SSRI response in depression,
obsessive– compulsive disorder, or other conditions characterized by disturbances in serotonergic pathways.
Biol Psychiatry 1998;44:825– 838 © 1998 Society of Biological Psychiatry
From the Department of Psychiatry, University of Pittsburgh, School of Medicine,
Western Psychiatric Institute and Clinic, Pittsburgh, Pennsylvania (WK, KG);
and Neuropsychiatric Institute and Hospital, School of Medicine, University of
California at Los Angeles, Los Angeles, California (MS).
Address reprint requests to Walter H. Kaye, MD, University of Pittsburgh Medical
Center, Western Psychiatric Institute and Clinic, 3811 O’Hara Street, Pittsburgh, PA 15213.
Received January 6, 1998; revised April 24, 1998; accepted April 30, 1998.
© 1998 Society of Biological Psychiatry
Key Words: Anorexia nervosa, bulimia nervosa, serotonin, selective serotonin reuptake inhibitors, treatment,
eating disorders
Introduction
A
norexia nervosa (AN) and bulimia nervosa (BN) are
disorders characterized by aberrant patterns of feeding behavior and weight regulation, and disturbances in
attitudes toward weight and shape and the perception of
body shape. In AN, there is an inexplicable fear of weight
gain and unrelenting obsession with fatness even in the
face of increasing cachexia. BN usually emerges after a
period of dieting, which may or may not have been
associated with weight loss. Binge eating is followed by
either self-induced vomiting, or by some other means of
compensation for the excess of food ingested. The majority of people with BN have irregular feeding patterns, and
satiety may be impaired. Although abnormally low body
weight is an exclusion for the diagnosis of BN, some
25–30% of patients with BN who present to treatment
centers have a prior history of AN; however, all individuals with BN have pathological concern with weight and
shape. Common to individuals with AN or BN are low
self-esteem, depression, and anxiety.
In certain respects, both diagnostic labels are misleading. Individuals affected with AN rarely have complete
suppression of appetite, but rather exhibit a volitional, and
more often than not, ego syntonic resistance to feeding
drives while eventually becoming preoccupied with food
and eating rituals to the point of obsession. Similarly, BN
may not be associated with a primary, pathological drive
to overeat; rather, like individuals with AN, those with BN
have a seemingly relentless drive to restrain their food
intake, an extreme fear of weight gain, and often have a
distorted view of their actual body shape. Loss of control
with overeating usually occurs intermittently and typically
only some time after the onset of dieting behavior.
Episodes of binge eating ultimately develop in a significant proportion of people with AN (Halmi et al 1991),
0006-3223/98/$19.00
PII S0006-3223(98)00195-4
826
W. Kaye et al
BIOL PSYCHIATRY
1998;44:825– 838
whereas some 5% of those with BN will eventually
develop AN (Hsu and Sobkiewicz 1989). Because restrained eating behavior and dysfunctional cognitions
relating weight and shape to self-concept are shared by
patients with either of these syndromes, and considering
that transitions between syndromes occur in many, it has
been argued that (Schweiger and Ficther 1997) AN and
BN share at least some risk and liability factors in
common.
The etiology of AN and BN is presumed to be complex
and multiply influenced by developmental, social, and
biological processes (Garner 1993; Treasure and Campbell
1994); however, the exact nature of these interactive
processes remains incompletely understood. Certainly,
cultural attitudes toward standards of physical attractiveness have relevance to the psychopathology of eating
disorders, but it is unlikely that cultural influences in
pathogenesis are very prominent. For one, dieting behavior and the drive toward thinness is quite commonplace in
industrialized countries throughout the world, yet AN and
BN affect only an estimated 0.3– 0.7 %, and 1.7–2.5%,
respectively, of the general female population. Moreover,
the fact that numerous clear descriptions of AN date from
the middle of the nineteenth century (Treasure and Campbell 1994) suggests that factors other than our current
culture play an etiologic role. Second, both syndromes,
AN in particular, have a relatively stereotypic clinical
presentation, sex distribution, and age of onset supporting
the possibility of some biologic vulnerability.
Illness Phenomenology
Variations in feeding behavior have been used to subdivide individuals with AN into two meaningful diagnostic
subgroups that have been shown to differ in other psychopathological characteristics (Garner et al 1985; Strober et
al 1982). In the restricting subtype of AN, subnormal body
weight and an ongoing malnourished state are maintained
by unremitting food avoidance; in the bulimic subtype of
AN, there is comparable weight loss and malnutrition, yet
the course of illness is marked by supervening episodes of
binge eating, usually followed by some type of compensatory action such as self-induced vomiting or laxative
abuse. Individuals with the bulimic subtype of AN are also
more likely to exhibit histories of behavioral dyscontrol,
substance abuse, and overt family conflict in comparison
to those with the restricting subtype. Particularly common
in individuals with AN are personality traits of marked
perfectionism, conformity, obsessionality, constriction of
affect and emotional expressiveness, and reduced social
spontaneity. These traits typically appear in advance of the
onset of illness and persist even after long-term weight
recovery, indicating they are not merely epiphenomena of
acute malnutrition and disordered eating behavior (Casper
1990; O’Dwyer et al 1996; Srinivasagam et al 1995;
Strober 1980).
Individuals with BN remain at normal body weight,
although many aspire to ideal weights far below the range
of normalcy for their age and height. The core features of
BN include repeated episodes of binge eating followed by
compensatory self-induced vomiting, laxative abuse, or
pathologically extreme exercise, as well as abnormal
concern with weight and shape. The DSM-IV has specified a distinction within this group between those individuals with BN who engage in self-induced vomiting or
laxative, diuretic, or enema abuse (purging type), and
those who exhibit other forms of compensatory action
such as fasting or exercise (nonpurging type). Beyond
these differences, it has been speculated (Vitousek and
Manke 1994) that there are two clinically divergent
subgroups of individuals with BN differing significantly in
psychopathological characteristics: a so-called multi-impulsive type, in whom bulimia occurs in conjunction with
more pervasive difficulties in behavioral self-regulation
and affective instability, and a second type whose distinguishing features include self-effacing behaviors, dependence on external rewards, and extreme compliance. BN
patients of the multi-impulsive type are far more likely to
have histories of substance abuse and display other impulse control problems such as shoplifting and selfinjurious behaviors. Considering these differences, it has
been postulated that individuals with multi-impulsive type
BN rely on binge eating and purging as a means of
regulating intolerable states of tension, anger, and fragmentation; in contrast, individuals with BN who are not of
this multi-impulsive type may have binge episodes precipitated through dietary restraint with compensatory behaviors maintained through reduction of guilty feelings associated with fears of weight gain.
Course of Illness
Most cases of AN emerge during the period of adolescence, although the condition can be observed in children.
There is no clear consensus as of yet on whether or not
prepubertal onset of the illness confers a more or less
ominous prognosis (Halmi et al 1979; Theander 1985).
Recovery from the illness tends to be protracted, but
studies of long-term outcome reveal the illness course to
be highly variable; roughly 50% of individuals will eventually have reasonably complete resolution of the illness,
whereas another 30% will have lingering residual features
that wax and wane in severity long into adulthood. Ten
percent of people with AN will pursue a chronic, unremitting course, and the remaining 10% of those affected will
Serotonin and SSRIs in Eating Disorders
eventually die from the disease (Crisp et al 1992; Isager et
al 1985; Strober et al 1997a; Sullivan 1995).
BN is usually precipitated by dieting and weight loss,
yet it can occur in the absence of apparent dietary restraint.
The frequency of binge episodes, their duration, and
amount of food consumed during any one episode all vary
considerably among patients. Age of onset is somewhat
more variable in BN as compared to AN, with most cases
developing during the period from mid to late adolescence
through the mid twenties (Fairburn et al 1997). Follow-up
studies of clinic samples 5–10 years after presentation
showed a 50% rate of recovery, while nearly 20% continued to meet full criteria for BN (Keel and Mitchell 1997).
Following onset, disturbed eating behavior will wax and
wane over the course of several years in a high percentage
of clinic cases. Approximately 30% of women who had
been in remission experienced relapse into bulimic symptoms.
Evidence of Genetic Influences
As noted previously, there is no convincing evidence that
cultural factors are the primary or most formidable determinants of these disorders; however, emerging evidence
suggests that both AN and BN are familial and that
clustering of the disorder in families may arise partly from
genetic transmission of risk (Lilenfeld et al in press;
Strober 1991). In one large family study of AN (Strober et
al 1990), risk of the disorder in mothers and sisters of
probands was estimated at 4%, or roughly eight times the
lifetime expectancy in the general female population.
Moreover, the present authors have reported recently that
a more broadly defined eating disorder phenotype that
differs from AN or BN in degree of severity occurs far
more often in relatives of both AN and BN probands
compared to relatives of normal controls. Likewise, analysis of data from a large, epidemiological sample of twins
obtained via the Virginia Twin Registry (Kendler et al
1991; Walters and Kendler 1995) adds to evidence of a
strong association between AN and BN. Specifically, it
was found that the cotwin of a twin affected with AN was
2.6 times more likely to have a lifetime diagnosis of BN
compared to cotwins of unaffected twins. In short, evidence suggests at least some sharing of familial risk and
liability factors between AN and BN. This finding is of
considerable importance to future family– genetic research
if it should be confirmed that the broader, subclinical
phenotype is more common among relatives than the
narrow, more severe phenotype.
Paralleling these accounts are several reports of greater
pairwise concordance rates of eating disorders in monozygotic compared to dizygotic twin pairs, with heritability
estimates in the range of 50 –90% for AN (Holland et al
BIOL PSYCHIATRY
1998;44:825– 838
827
1988; Treasure and Holland 1989) and 35–50% for BN
(Kendler et al 1991); however, with the exception of
reports from Kendler’s group, based on diagnostic study of
the Virginia Twin Registry sample, existing accounts of
differential concordance of eating disorders in twin pairs
are hampered by unsystematic, and potentially biased,
sampling.
As noted earlier, a range of general psychiatric symptoms are found commonly in patients with AN or BN. In
many cases, they develop secondary to malnutrition and
other disabling psychological effects of aberrant eating
psychological disability; yet, in some, they clearly antedate disordered eating, or arise following recovery from
low body weight or binge eating. Whether or not particular
psychiatric disorders increase liability to eating disorders
or are expressions of a shared underlying diathesis is a
question of heuristic and clinical importance.
Several family and twin studies have examined the
covariation between eating disorders and various other
psychiatric conditions that co-occur with AN and BN.
These studies have been reviewed in detail (Lilenfeld et al
1997; Strober 1991). With regard to major affective
illness, studies of AN probands have yielded familial risk
estimates in the range of 7–25%, with relative risk
estimates in studies employing normal controls in the
range of 2.1–3.4. Likewise, studies of BN probands have
shown, with rare exceptions, that their first-degree relatives are several times more likely to develop affective
disorders than relatives of control subjects. At the same
time, most studies considering the effects of proband
comorbidity on familial risk have shown that affective
illness is more likely to be transmitted by probands with
this same diagnostic comorbidity. In short, although AN
and BN often co-occur with major mood disorders, unipolar depression in particular, the two conditions do not seem
to express a single, shared transmitted liability.
Family studies investigating rates of substance use
disorders suggest relatively low rates among relatives of
restricting AN probands (Holderness et al 1994; Lilenfeld
et al in press). In contrast, rates are elevated in relatives of
probands with BN; however, results from two studies
(Kaye et al 1996; Schuckit et al 1996) indicate that there
is no evidence of a cross-transmission of BN and substance use disorder in families, whereas twin data (Kendler
et al 1991) have shown that the genetic variation influencing susceptibility to alcoholism was independent of those
genetic factors underlying risk for BN. Tentative evidence
of independent familial transmission of obsessive– compulsive disorders and AN and BN has also been reported
by the present authors (Lilenfeld et al in press). On the
other hand, we have reported preliminary data suggesting
a common familial transmission of AN and obsessive–
compulsive personality disorder, thus suggesting the exis-
828
W. Kaye et al
BIOL PSYCHIATRY
1998;44:825– 838
tence of a broad, genetically influenced phenotype with
core features of rigid perfectionism and propensity for
extreme behavioral constraint. Hence, in spite of the
formidable challenges encountered in the biological study
of nutritionally compromised individuals, efforts to better
understand the pathophysiology of AN and BN continue to
have clinical and heuristic value.
Persistent Psychological Disturbances after
Recovery from AN and BN
People who have an eating disorder (ED) often have a
variety symptoms aside from pathological eating behaviors. Physiological symptoms include an abundance of
neuroendocrine, autonomic, and metabolic disturbances.
Psychological symptoms include depression, anxiety, substance abuse, and personality disorders. Determining
whether such symptoms are a consequence or a potential
cause of pathological feeding behavior or malnutrition is a
major methodological issue in the study of EDs. It is
impractical to study EDs prospectively due to the young
age of onset and difficulty in premorbid identification of
people who will develop an ED; however, subjects can be
studied after long-term recovery from an ED. The assumed
absence of confounding nutritional influences in recovered
ED women raises a possibility that persistent psychobiological abnormalities might be trait-related and potentially
contribute to the pathogenesis of this disorder. A limited
number of studies have investigated people who have
recovered from AN and BN. Although the definition of
recovery from an ED has not been formalized, investigators tend to include people formerly ill with AN after they
have been at a stable and healthy body weight for months
or years of time and have not been malnourished or
engaged in pathological eating behavior during that period
of recovery. For BN, investigators tend to include subjects
who have been abstinent from bingeing and purging for
months or years of time. Some investigators include a
criteria of normal menstrual cycles and a minimal duration
of recovery, such as 1 year of time.
Investigators (Casper 1990; O’Dwyer et al 1996; Srinivasagam et al 1995; Strober 1980) have found that women
who were long-term recovered from AN had a persistence
of obsessional behaviors as well as inflexible thinking,
restraint in emotional expression, and a high degree of
self- and impulse control. In addition, they have social
introversion, overly compliant behavior, and limited social
spontaneity as well as greater risk avoidance and harm
avoidance. Moreover, individuals who are long-term recovered from AN had continued core eating disorder
symptoms, such as ineffectiveness, a drive for thinness,
and significant psychopathology related to eating habits.
Similarly, people who have recovered from BN continue
to be overconcerned with body shape and weight, abnormal eating behaviors, and dysphoric mood (Collings and
King 1994; Fallon et al 1991; Johnson-Sabine et al 1992;
Kaye et al in press; Norring and Sohlberg 1993). Recovered AN and BN women had increased perfectionism, and
their most common obsessional target symptoms were the
need for symmetry and ordering/arranging. Considered
together, these residual behaviors can be characterized as
over concerns with body image and thinness, obsessionality with symmetry, exactness, and perfectionism, and
dysphoric/negative affect. In general, pathological eating
behavior and malnutrition appear to exaggerate the magnitude of these concerns. Thus, the intensity of these
symptoms is less after recovery, but the content of these
concerns remains unchanged. The persistence of these
symptoms after recovery raises a question of whether such
behaviors are premorbid traits that contribute to the
pathogenesis of AN and BN.
Studies of Neurotransmitters
The role of biology in the etiology of AN has been
proposed for the past 60 years (Russell 1970). Earlier
theories raised the question of whether people with AN
had a pituitary or hypothalamic disturbance. More recently, a growing understanding of neurotransmitter modulation of appetitive behaviors has raised the question of
whether some disturbance of neurotransmitter function
causes AN and/or BN (Fava et al 1989; Leibowitz 1986;
Morley and Blundell 1988). It is possible that disturbances
of brain neuropeptides and/or monoamines could contribute to other symptoms and behaviors, such as neuroendocrine or autonomic abnormalities, or alterations of mood
and behavior, in people with AN or BN. It is important to
emphasize that monoamine or neuropeptide disturbances
could be a consequence of dietary abnormalities, or
premorbid traits that contribute to a vulnerability to
develop AN or BN. One way to tease apart cause and
effect is to study people with AN or BN at various stages
in their illness; that is, while symptomatic and after
recovery.
Neuropeptides
Multiple neuroendocrine abnormalities have been documented in AN and BN (Sharp and Freeman 1997),
including alteration of the hypothalamic–pituitary– gonadal axis, the hypothalamic–pituitary–adrenal axis, thyroid system, growth hormone secretion, and fluid conservation as well as autonomic instability and reduced
metabolic function. For the most part, these neuroendocrine disturbances are state-related and tend to normalize
after clinical recovery.
Serotonin and SSRIs in Eating Disorders
Starvation-induced alterations of neuropeptide activity
probably contribute to neuroendocrine dysfunctions in the
eating disorders. For example, corticotropin-releasing hormone (CRH) alterations contribute to hypercortisolemia
(Gold et al 1986; Kaye et al 1987) and neuropeptide Y
(NPY) alterations may contribute to amenorrhea (Kaye et
al 1990). Hypersecretion of vasopressin and reduced
cerebrospinal fluid (CSF) oxytocin in underweight AN
subjects may maintain or exacerbate the persistent preoccupation with the adverse consequences of food intake
(Demitrack et al 1990). Hence, alterations in a number of
different peptides could contribute to and perpetuate the
characteristic psychophysiological disturbances, such as
reduced feeding, in acutely ill AN and BN patients. Many
people with AN or BN cannot easily “reverse” their
illness. In AN, malnutrition may contribute to entering a
downward spiraling cycle with malnutrition sustaining and
perpetuating the desire for more weight loss and dieting.
Symptoms, such as increased satiety, as well as obsessions
and dysphoric mood, may be exaggerated by these neuropeptide alterations and thus contribute to this downward
spiral. Even after improved nutrition and weight gain,
many people with AN have much difficulty normalizing
their behavior. For example, abnormal eating attitudes,
affective disturbances, and menstrual dysfunction may
persist despite weight restoration (Copeland et al 1995;
Falk and Halmi 1982). While neuroendocrine and neuropeptide disturbances do not appear to be a permanent
feature or cause of AN, these disturbances are strongly
entrenched and are not easily corrected by improved
nutrition or short-term weight normalization. The fact that
neuropeptide disturbances are not found after long-term
recovery suggests that therapy must be sustained for
months after weight normalization.
Serotonin Neuronal Activity
There has been considerable interest in the role that
serotonin (5-HT) may play in AN and BN (Blundell 1992;
Brewerton 1995; Jimerson et al 1990; Kaye and Weltzin
1991; Treasure and Campbell 1994). A substantial number
of studies have shown alterations in 5-HT activity in the ill
state. Although less well studied, 5-HT disturbances appear to persist after recovery. In addition, people with AN
and BN (see below) respond to antidepressants in placebocontrolled trials.
5-HT pathways play an important role in postprandial
satiety. Treatments that increase intrasynaptic 5-HT, or
directly activate 5-HT receptors, tend to reduce food
consumption, whereas interventions that dampen serotonergic neurotransmission or block receptor activation reportedly increase food consumption and promote weight
gain (Blundell 1984; Leibowitz 1986). Moreover, central
BIOL PSYCHIATRY
1998;44:825– 838
829
Figure 1. CSF 5-HIAA concentrations when ill and after recovery in people with anorexia and bulimia nervosa. Values are
compared to those of healthy control women, which are set at
100%.
nervous system (CNS) 5-HT pathways have been implicated in the modulation of mood, impulse regulation and
behavioral constraint, and obsessionality. They also affect
a variety of neuroendocrine systems.
When underweight, patients with AN have a significant
reduction in basal concentrations of the 5-HT metabolite
5-hydroxyindoleacetic acid (5-HIAA) in the CSF compared to healthy controls (Figure 1) (Demitrack et al 1995;
Kaye et al 1988b), as well as blunted plasma prolactin
response to drugs with 5-HT activity (Hadigan et al 1995)
and reduced 3H-imipramine binding (Weizman et al
1986). Together, these findings suggest reduced serotonergic activity, which could be secondary to a diet-induced
reduction of availability of the amino acid, tryptophan, the
precursor of 5-HT. In contrast, CSF concentrations of
5-HIAA are reported to be elevated (Kaye et al 1991a) and
neuroendocrine responses to 5-HT-stimulating drugs are
normalized (O’Dwyer et al 1996) in women who are
long-term weight recovered from AN. These contrasting
findings of reduced and heightened serotonergic activity in
acutely ill and long-term recovered AN individuals, respectively, may seem counterintuitive; however, since
dieting lowers plasma tryptophan levels in otherwise
healthy women (Anderson et al 1990), resumption of
normal eating in AN may unmask intrinsic abnormalities
in serotonergic systems that mediate certain core behavioral or temperamental underpinnings of risk and vulnerability.
Considerable evidence also exists of dysregulation of
serotonergic processes in people who are ill with BN. This
includes blunted prolactin response to the 5-HT receptor
agonists m-chlorophenylpiperazine (m-CPP) (Brewerton
et al 1992; Levitan et al 1997), 5-hydroxytrytophan
(Goldbloom et al 1990a), and dl-fenfluramine (Jimerson et
al 1997; McBride et al 1991), increased platelet 5-HT
uptake (Goldbloom et al 1990b), reduced platelet imipra-
830
BIOL PSYCHIATRY
1998;44:825– 838
mine binding capacity (Marazziti et al 1988), and enhanced migraine-like headache response to m-CPP challenge (Brewerton et al 1988). Moreover, acute
perturbation of serotonergic tone by dietary depletion of
tryptophan has also been linked to increased food intake
and mood irritability in women with BN compared to
healthy controls (Weltzin et al 1994). Whereas ill BN have
normal CSF 5-HIAA levels (Figure 1), women who are
long-term recovered from BN have elevated concentrations of 5-HIAA in the CSF (Kaye et al in press).
Together, these data show that both recovered AN and
BN women have elevated CSF 5-HIAA concentrations. It
has been found that low levels of CSF 5-HIAA are
associated with impulsive and nonpremeditated aggressive
behaviors, which cut across traditional diagnostic boundaries. Behaviors found after recovery from AN and BN,
such as obsessions with symmetry, exactness, and perfectionism, and negative affect, tend to be opposite in
character to behaviors displayed by people with low
5-HIAA levels. Together, these studies contribute to a
growing literature that suggests that CSF 5-HIAA concentrations may correlate with a spectrum of behavior. These
data support the hypothesis (Cloninger et al 1993) that
increased CSF 5-HIAA concentrations may be associated
with exaggerated anticipatory overconcern with negative
consequences, while the lack of such concerns may
explain impulsive, aggressive acts that are associated with
low CSF 5-HIAA.
These data raise the possibility that a disturbance of
5-HT activity may create a vulnerability for the expression
of a cluster of symptoms that are common to both AN and
BN. The possibility of a common vulnerability for AN and
BN may seem puzzling given well-recognized differences
in behavior in these disorders; however, recent studies
suggest that AN and BN have a shared etiologic vulnerability (Kendler et al 1991). Other factors, that are independent of a vulnerability for the development an eating
disorder, may contribute to the development of eating
disorder subgroups. For example, people with restrictingtype AN have extraordinary self-restraint and self-control.
The risk for obsessive– compulsive personality disorder is
elevated only in this subgroup and in their families and
shows a shared transmission with restricting-type AN
(Lilenfeld et al in press). In other words, an additional
vulnerability for behavioral over-control and rigid and
inflexible mood states, combined with a vulnerability for
an eating disorder, may result in restricting-type AN.
The contribution of 5-HT to specific human behaviors
remains uncertain. That is, 5-HT has been postulated to
contribute to temperament or personalty traits, such as
harm avoidance (Cloninger 1987) or behavioral inhibition
(Soubrie 1986), or to categorical dimensions such as
obsessive– compulsive disorder (OCD) (Barr et al 1992),
W. Kaye et al
Figure 2. Hypothetical model of serotonin (5-HT) neuronal
activity in control women (CW) and women with BN. (a)
Normal 5-HT neuronal activity in CW. (b) Response of a CW to
diet-induced reduction of tryptophan (TRP) includes a downregulation of presynaptic and an up-regulation of postsynaptic
activity to maintain consistent 5-HT output. (c) Increased 5-HT
output in recovered BN results in obsessive, negative affect. We
hypothesize that BN subjects may have an intrinsic disturbance
of 5-HT neuronal function. For example, a defect in a presynaptic autoreceptor might make the system unresponsive to
increased extracellular 5-HT. Such a defect might explain increased 5-HT signal transmission in the premorbid and recovered
state, which might contribute to a vulnerability for restricted
eating. (d) People may discover that extreme dieting, which may
reduce plasma tryptophan availability, is a means by which they
can self-modulate their brain 5-HT functional activity and briefly
reverse a dysphoric mood state.
anxiety and fear (Charney et al 1990), or depression
(Graheme-Smith 1992) as well as satiety for food consumption. Importantly, these symptoms persist in AN and
BN after recovery.
BN, the most common ED, may be the prototypic
expression of a disturbance of 5-HT activity that contributes to the pathogenesis of eating disorders. Clinically,
people with BN have extremes of eating and behavior.
They tend to eat few normal meals. They tend to either diet
or overeat. Similarly, they tend to fluctuate between
minimization and inhibition of mood states and extremes
of mood and catastrophic over-concerns. These clinical
observations, coupled with data from studies in ill and
recovered BN women, lead to the speculation that the
5-HT system in people with BN is inherently unstable and
poorly modulated (Figure 2). Certain traits, such as restricted eating and obsessive perfectionism, and exactness,
harm avoidance, and negative affect might be consistent
with increased 5-HT transmission in a nondieting state. In
contrast, a diet-induced reduction in synaptic 5-HT release
could result in a reduction of this dysphoric state, but
Serotonin and SSRIs in Eating Disorders
might lead in turn to extremes of unstable mood and binge
eating. It is possible that people with such an inherent
modulatory defect in 5-HT function may be prone to
develop an ED. Because of their modulatory 5-HT defect,
they cannot respond appropriately and precisely to stress
or stimuli, or modulate their affective states. They may
learn that extremes of dietary intake, by effects on plasma
tryptophan, are a means by which they can crudely
modulate their brain 5-HT functional activity. Several
investigators (Jimerson et al 1992; Kaye et al 1988a) have
proposed a model in which individuals with BN may
restrict eating or overeat as a means of self-modulating
5-HT activity. That is, dieting or binge episodes could
alter the tryptophan to large neutral amino acid ratio in
plasma, which in turn alters tryptophan availability to the
brain, which results in changes in 5-HT synthesis and
release (Fernstrom and Faller 1978). In fact, recent studies
(Weltzin et al 1994) show ill patients with BN, after
tryptophan depletion, have an increase in labile and
dysphoric mood and overeat compared to control women,
supporting the possibility that women with BN have a
fragile and dysregulated 5-HT system that is vulnerable to
dietary manipulations.
Treatment of Bulimia Nervosa
It is fair to say that progress to date in establishing the
efficacy of specific psychological and pharmacologic therapies for EDs has been more dramatic for BN than for AN
(Arnow 1997). With regard to psychotherapy, although the
number of controlled clinical trials are still few in number,
most indicate that cognitive behavior therapy (CBT) is an
effective treatment for upwards of 60 –70% of individuals
with BN, with remission of binge eating and purging
achieved in some 30 –50% of cases (Connors et al 1984;
Fairburn et al 1993; Kirkley et al 1985; Wolf and
Crowther 1992). Evidence for the efficacy of antidepressant pharmacotherapy in BN is impressive; however, the
benefits may diminish over time in a significant proportion
of individuals with BN who respond initially, and only a
minority have complete suppression of their symptoms
with antidepressant monotherapy (Agras et al 1992; Crow
and Mitchell 1996; Fluoxetine Bulimia Nervosa Collaborative Study Group 1992; Hoffman and Halmi 1993;
Jimerson et al 1996; Kaye et al in submission; Mitchell
and de Zwaan 1993; Mitchell et al 1990, 1993; Russell
1988; Walsh 1991a).
The results of most double-blind, placebo-controlled
randomized trials reported to date indicate that antidepressants show at least some superiority over placebo in
reducing the frequency of binge eating episodes (Walsh
1991b). In addition, some studies show a reduction in
intensity of some other symptoms commonly seen in BN,
BIOL PSYCHIATRY
1998;44:825– 838
831
such as preoccupation with food and depression (Goldbloom and Olmsted 1993). These findings have been
demonstrated with a variety of antidepressive medications,
including tricyclic antidepressants (TCAs) (imipramine,
desipramine, clomipramine, amitriptyline) (Agras et al
1987; Barlow et al 1988), monoamine oxidase inhibitors
(MAOIs) (phenelzine, isocarboxazid) (Walsh et al 1984),
and selective serotonin reuptake inhibitors (SSRIs) (fluoxetine, fluvoxamine) (Goldstein et al 1995). Patients participating in these trials typically reported from 8 to 10
episodes of binge eating per week at baseline. The
“average” decrease in binge frequency for patients receiving the antidepressant medication was about 55%, with
wide variation across studies. Placebo responses were
similarly variable, but were generally less than half the
size of the response for the active treatment; however, only
a minority of the patients actually achieved full abstinence
from bingeing and purging behaviors (Mitchell et al 1997).
Most trials have shown no correlation between improvement in mood and reduction in BN symptoms. In addition,
antidepressants suppress bulimic symptoms also in nondepressed bulimic patients, suggesting a mode of action
other than through their antidepressant effects (Blouin et al
1989). In some studies, the patients receiving the antidepressive medication demonstrated a significant increase in
dietary restraint and a reduction in the tendency for
stressors to trigger binge eating.
There is little evidence for superiority of response to a
single class of medication. Thus, differences in side effects
may be a significant factor in the clinical choice of the
antidepressant. The dosage level needed to achieve an
effect appears to be similar to that required in major
depression, with a tendency for higher dosages (e.g.,
200 –300 mg/day of the relevant TCA) to be more effective (Walsh 1991b). The time course of action of the
medications also appears to be similar to that found in
depression, with improvement occurring over the course
of several weeks. Improvement in frequency of selfinduced vomiting generally parallels the decrease in
bingeing. Treatment failure should be considered only if
adequate doses have been administered for 8 –12 weeks.
Fluoxetine 60 mg/day has proved significantly superior to
20 mg/day and to placebo in the reduction of binge eating
and BN-related features (Fluoxetine Bulimia Nervosa
Collaborative Study Group 1992). This dosage has been
shown to be safe with minimal adverse effects even over
an extended treatment period. Other antidepressants have
not been studied systematically in the treatment of BN.
Bupropion is contraindicated in BN patients, as it has been
associated with increased risk of seizures (Horne et al
1988).
The question of how long to continue antidepressant
therapy in BN has not been addressed systematically, as
832
BIOL PSYCHIATRY
1998;44:825– 838
most trials have incorporated only a relatively brief duration of treatment. A few follow-up studies extending from
4 to 24 months have found a high relapse rate upon the
discontinuation of the antidepressant medication (Agras
1997). As BN itself often runs a chronic course with
significant rate of relapse, antidepressants apparently have
to be continued for prolonged periods of time in patients
responding well in the initial stages of treatment. It should
be noted, however, that attenuation of the antibinge effect
of the antidepressant has been noted over time.
Five studies (Agras et al 1992; Fichter et al 1991;
Leitenberg et al 1994; Mitchell et al 1990; Walsh et al
1997) have been published or presented to date that
assessed the relative efficacy of combining psychotherapy
(in most trials of the CBT type) and antidepressants for the
management of BN, compared with the isolated treatments
themselves. Although differing in many respects, these
studies suggest that the improvement in bulimic symptoms
with CBT alone was greater than with the medication
alone. Adding medication to the psychotherapy generally
did not improve significantly the outcome over psychotherapy alone in terms of eating behaviors, nor did it
increase the speed of the therapeutic response; however,
one prolonged follow-up evaluation found that combined
treatment was more effective on a number of eating
variables than CBT alone (Agras et al 1992). Another
study showed the superiority of combined therapy in
reducing the rates of anxiety and depression (Mitchell et al
1990).
Treatment of Anorexia Nervosa
People with AN have responded less effectively to treatment (Herzog et al 1992). Extended hospitalizations can
be lifesaving because such treatment can restore weight to
emaciated people, which, in turn, reverses medical complications; however, hospitalizations can be lengthy and
expensive. In fact, the hospital utilization rate for people
with AN is higher than for any other psychiatric disorder,
aside from schizophrenia and organic disorders (McKenzie and Joyce 1992). Moreover, relapse after hospitalization has been high (Russell et al 1987).
The evaluation of the efficacy of medications in augmenting weight gain in AN is limited, because most trials
have been conducted in outpatients or inpatients participating in behavioral and nutritional ED programs, which
are themselves efficient in the short run. Nevertheless, in
these settings controlled trials have not provided consistent evidence for the efficacy of antidepressant medications in the treatment of AN (Biederman et al 1985; Gross
et al 1981; Lacey and Crisp 1980). Cyproheptadine hydrochloride, a 5-HT antagonist, may increase the rate of
weight gain in restricting-type AN patients (Halmi et al
W. Kaye et al
1986) but only in high doses (Goldberg et al 1979;
Vigersky and Loriaux 1977). Limited efficacy of pharmacologic and psychological treatment in AN may be due, in
part, to the fact that past treatments have mainly focused
on attempts to increase the rate of weight gain of emaciated patients in a hospital setting. Inpatient treatment,
consisting of nursing care, behavior modification, and
supportive psychotherapy, succeeds in restoring the
weight of most emaciated AN patients. Thus, it is difficult
to prove that an active medication is effective in such a
setting.
Some, but not all (Strober et al 1997b), recent studies
that have focused on preventing relapse show more promise. Several psychotherapies specifically developed to treat
AN appear to show reduced relapse at a 1–2-year follow-up (Russell et al 1987; Treasure et al 1996). Our
group (Kaye et al 1991b; in submission) has found in
separate open and double-blind placebo-controlled studies
that fluoxetine improved outcome and reduced relapse
after weight restoration. That is, fluoxetine was associated
with a significant reduction in core ED symptoms, depression, anxiety, and obsessions and compulsions. In a recent
double-blind, placebo-controlled study, our group found
that fluoxetine, when given after weight restoration, significantly reduces the extremely high rate of relapse
normally seen in AN. Subjects were started on fluoxetine
(n 5 16) or placebo (n 5 19) after inpatient weight
restoration, discharged from the hospital, and followed for
1 year as outpatients. Ten of 16 (63%) subjects on
fluoxetine remained well over 1 year of outpatient followup, whereas only 3 of 19 (16%) remained well on placebo
(p 5 .006). (Figure 3). Fluoxetine administration was
associated with a significant weight gain and a significant
reduction in obsessions and compulsions. Thus fluoxetine
improves outcome in AN by reducing symptoms and helps
maintain a healthy body weight in outpatient treatment.
Our group (Ferguson et al in press) and others (Attia et
al 1998) have found that SSRIs are not useful when people
with AN are malnourished and underweight. As noted by
Tollefson (1995), SSRIs are dependent on neuronal release
of 5-HT for their action. If the release of 5-HT from
presynaptic neuronal storage sites was substantially compromised, and net synaptic 5-HT concentration was negligible, a clinically meaningful response to an SSRI might
not occur. In fact, malnourished individuals with AN have
reduced CSF 5-HIAA, the major 5-HT metabolite in the
brain (Kaye et al 1984), suggesting reduced synaptic
5-HT. This could be due to reduced availability of tryptophan, the essential amino acid precursor to 5-HT
(Schweiger et al 1986).
This link between dietary intake and SSRI efficacy is
supported by data that have repeatedly shown that dieting
in healthy normal-weight and obese women reduces tryp-
Serotonin and SSRIs in Eating Disorders
Figure 3. Survival curve showing proportion of anorexia nervosa
patients remaining well on fluoxetine or placebo after weight
restoration.
tophan availability and thereby limits potential serotonergic production (Anderson et al 1990; Gatti et al 1993;
Goodall 1990; Walsh et al 1995; Wolfe et al 1997).
Moreover, studies in animals show that food restriction
decreases 5-HT and its synthesis rate in the brain (Haleem
and Haider 1996) and down-regulates the density of 5-HT
transporters (Huether et al 1997). Finally, depletion of
tryptophan, the precursor of 5-HT, reverses the effects of
SSRI antidepressants in depressed patients (Barr et al
1994; Bremner et al 1997; Delgado et al 1990). In
addition, in the patients studied by Delgado et al (1990),
plasma tryptophan was inversely related to depression
scores. In AN, weight restoration normalizes nutrition, and
CSF 5-HIAA concentrations become elevated (Kaye et al
1991a). These changes in nutrients and 5-HT activity may
account for why individuals with AN may become responsive to fluoxetine after weight restoration.
Other Potential Effects of Nutrition on
Serotonin Activity and SSRI Efficacy
Impaired 5-HT functional activity could also be a consequence of nutrients other than tryptophan, which are
essential to normal 5-HT synthesis and function, and are
reduced in underweight AN patients (McClain et al 1993;
Mira et al 1989; Nunez et al 1995; Rock and CurranCelentano 1994; Rock and Vasantharajan 1995). These
include insulin activity, essential fatty acids, zinc, and
pyridoxine (vitamin B6). Insulin mediates the increased
brain uptake of tryptophan and 5-HT synthesis following
carbohydrate ingestion, and the low basal and postingestive insulin concentrations observed in underweight AN
BIOL PSYCHIATRY
1998;44:825– 838
833
patients (Alderdice et al 1985) may hamper this process.
Very low-fat diets, as typically consumed by women with
AN, have been found to diminish neuronal 5-HT activity
in animal studies (Muldoon et al 1992). In addition, the
type of dietary fat consumed determines the biophysical
properties of cell membranes and hence influences neurotransmitter receptor function. Specifically, it has been
hypothesized that an inadequate n-3 fatty acid intake
adversely affects serotonergic function (Hibbeln and Salem 1995).
Zinc also influences membrane stability to the extent
that zinc deficiency increases membrane fluidity, causing
the loss of receptor function (McClain et al 1993). Poor
zinc status has been observed in individuals with AN
(Rock and Curran-Celentano 1994), which may contribute
to the ineffectiveness of SSRIs during the acute stage of
the illness. The B vitamins are also involved in metabolic
pathways influencing 5-HT availability. Vitamin B12 and
folate are required for the formation of tetrahydrobiopterin, which is a hydroxylase cofactor in the rate-limiting
step of 5-HT synthesis (Coppen et al 1989). Vitamin B6,
in the form of pyridoxal-5-phosphate, is a cofactor required in the reaction that converts 5-hydroxytryptophan
to 5-HT (Dakshinamurti et al 1990; Sharma and Dakshinamurti 1994). Dietary deficiency and lowered biochemical activity of these B vitamins in some patients with AN
(Mira et al 1989; Nunez et al 1995; Rock and CurranCelentano 1994; Rock and Vasantharajan 1995) may
therefore compromise the ability of SSRIs to facilitate
serotonergic function.
An alternative possibility is that reduced serotonergic
activity in dieting women and women underweight with
AN is mediated through low levels of gonadal steroids that
are observed in malnourished states (Pirke et al 1989;
Wakeling et al 1979). Estrogen has been found to stimulate significant increases in the density of central 5-HT2A
binding sites (Fink et al 1996) and 5-HT transporter
binding sites (McQueen et al 1996), to inhibit monoamine
oxidase activity (Chakravorty and Halbreich 1997), and to
be positively correlated to platelet 5-HT content
(Guicheney et al 1988). A deficiency of estrogen in
malnutrition may, therefore, thwart the functional actvity
of the serotonergic system and the potential responsiveness to SSRI medication.
In summary, AN is a condition in which inadequate
nutriture appears to impair SSRI efficacy. It has also been
shown that people with BN show signs of starvation when
bingeing and purging (Pirke et al 1985, 1989). Although
not as well studied, it is possible that malnutrition could
also compromise SSRI efficacy in individuals with BN.
These data raise the possibility that nutritional factors may
affect SSRI response in depression, OCD, or other conditions characterized by disturbances in serotonergic path-
834
BIOL PSYCHIATRY
1998;44:825– 838
ways. When patients do not respond to one SSRI medication, physicians often switch to another SSRI medication.
Another strategy would be to assess nutritional status.
Better nutrition might serve to improve SSRI efficacy. The
possibility that suboptimal nutrition status is responsible
for nonresponse to SSRI treatment in depression or OCD
remains to be determined.
Summary
Emerging data support the possibility that substantial
biologic vulnerabilities contribute to the pathogenesis of
AN and BN. The development of an ED is often attributed
to the effects of our cultural environment, such as mass
media, on body image; however, although all women in
our society are exposed to cultural mores that value
slimness, only a small percentage of women exposed to
these messages develop an ED. Thus, it is possible that
there may be an underlying biologic diathesis that places
someone “at risk” for developing BN. Similar shifts in the
understanding of the pathophysiology of obesity have
been driven by the discovery of potentially new mechanisms of weight regulation.
This work was presented at the Neuroscience Discussion Forum “A
Decade of Serotonin Research” held at Amelia Island, Florida in
November 1997. The conference was sponsored by the Society of
Biological Psychiatry through an unrestricted educational grant provided
by Eli Lilly and Company.
References
Agras WS (1997): Pharmacotherapy of bulimia nervosa and
binge eating disorder: Longer-term outcomes. Psychopharmacol Bull 33:433– 436.
Agras WS, Dorian B, Kirkley BG, Arnow B, Bachman J (1987):
Imipramine in the treatment of bulimia: A double-blind
controlled study. Int J Eat Disord 6:29 –38.
Agras WS, Rossiter EM, Arnow B, et al (1992): Pharmacologic
and cognitive-behavioral treatment for bulimia nervosa: A
controlled comparison. Am J Psychiatry 149:82– 87.
Alderdice JT, Dinsmore WW, Buchanan KD, Adams C (1985):
Gastrointestinal hormones in anorexia nervosa. J Psychiatr
Res 19:207–213.
Anderson IM, Parry-Billings M, Newsholme EA, Fairburn CG,
Cowen PJ (1990): Dieting reduces plasma tryptophan and
alters brain 5-HT function in women. Psychol Med 20:785–
791.
Arnow B (1997): Psychotherapy of anorexia and bulimia Ballieres Clin Psychiatry 3:235–257.
Attia E, Haiman C, Walsh BT, Falter SR (1998): Does fluoxetine
augment the inpatient treatment of anorexia nervosa? Am J
Psychiatry 155:548 –551.
Barlow J, Blouin J, Blouin A, Perez E (1988): Treatment of
W. Kaye et al
bulimia with desipramine: A double-blind crossover study.
Can J Psychiatry 34:24 –29.
Barr LC, Goodman WK, Price LH, McDougle CJ, Charney DS
(1992): The serotonin hypothesis of obsessive compulsive
disorder: Implications of pharmacologic studies. J Clin Psychiatry 53:17–28.
Barr LC, Goodman WK, McDougle CJ, et al (1994): Tryptophan
depletion in patients with obsessive-compulsive disorder who
respond to serotonin reuptake inhibitors. Arch Gen Psychiatry
51:309 –317.
Biederman J, Herzog DB, Rivinus TM, et al (1985): Amitriptyline in the treatment of anorexia nervosa: A double-blind,
placebo-controlled study. J Clin Psychopharmacol 5:10 –16.
Blouin J, Blouin A, Perez E (1989): Bulimia: Independence of
antibulimic and antidepressant properties of desipramine.
Can J Psychiatry 34:24 –29.
Blundell JE (1984): Serotonin and appetite. Neuropharmacology
23:1537–1551.
Blundell JE (1992): Serotonin and the biology of feeding. Am J
Clin Nutr 55:155S–159S.
Bremner JD, Innis RB, Saloman EM, et al (1997): Positron
emission tomography measurement of cerebral metabolic
correlates of tryptophan depletion-induced depressive relapse.
Arch Gen Psychiatry 54:364 –374.
Brewerton (1995): Toward a unified theory of serotonin dysregulation in eating and related disorders. Psychoneuroendocrinology 20:561–590.
Brewerton TD, Murphy DL, Mueller EA, Jimerson DC (1988):
The induction of migraine-like headaches by the serotonin
agonist, m-chlorophenylpiperazine. Clin Pharmacol Ther 43:
605– 609.
Brewerton TD, Mueller EA, Lesem MD, et al (1992): Neuroendocrine responses to m-chlorophenylpiperazine and L-tryptophan in bulimia. Arch Gen Psychiatry 49:852– 861.
Casper RC (1990): Personality features of women with good
outcome from restricting anorexia nervosa. Psychosom Med
52:156 –170.
Chakravorty SG, Halbreich U (1997): The influence of estrogen
on monoamine oxidase activity. Psychopharmacol Bull 33:
229 –233.
Charney DS, Wood SW, Krystal JH, Heninger GR (1990):
Serotonin function and human anxiety disorders. Ann NY
Acad Sci 600:558 –573.
Cloninger CR (1987): A systematic method for clinical description and classification of personality variants: a proposal.
Arch Gen Psychiatry 44:573–588.
Cloninger CR, Svrakic DM, Przybeck TR (1993): A psychobiological model of temperament and character. Arch Gen
Psychiatry 50:975–990.
Collings S, King M (1994): Ten-year follow-up of 50 patients
with bulimia nervosa. Br J Psychiatry 164:80 – 87.
Connors ME, Johnson CL, Stuckey MK (1984): Treatment of
bulimia with brief psychoeducational group therapy. Am J
Psychiatry 141:1512–1516.
Copeland PM, Sacks NR, Herzog DB (1995): Longitudinal
follow-up of amenorrhea in eating disorders. Psychosom Med
57:121–126.
Coppen A, Swade C, S.A. J, Armstrong RA, Blair JA, Leeming
Serotonin and SSRIs in Eating Disorders
RJ (1989): Depression and tetrahydrobiopterin: The folate
connection. J Affect Disord 16:103–107.
Crisp AH, Callender JS, Halek C, Hsu LKG (1992): Long-term
mortality in anorexia nervosa: A 20-year follow-up of the St
George’s and Aberdeen cohorts. Br J Psychiatry 161:104 –
107.
Crow SJ, Mitchell JE (1996): Integrating cognitive therapy and
medications in treating bulimia nervosa. Psychiatr Clin North
Am 19:755–760.
Dakshinamurti K, Paulose CS, Viswanathan M, Siow YL,
Sharma SK (1990): Neurobiology of pyridoxine. Ann NY
Acad Sci 585:128 –144.
Delgado PL, Price LH, Aghajanian GK, Landis H, Heninger GR
(1990): Serotonin function and the mechanism of antidepressant action. Reversal of antidepressant-induced remission by
rapid depletion of plasma tryptophan. Arch Gen Psychiatry
47:411– 418.
Demitrack MA, Lesem MD, Listwak SJ, Brandt HA, Jimerson
DC, Gold PW (1990):CSF oxytocin in anorexia nervosa and
bulimia nervosa: Clinical and pathophysiologic considerations. Am J Psychiatry 147:882– 886.
Demitrack MA, Heyes MP, Altemus M, Pigott TA, Gold PW
(1995): Cerebrospinal fluid levels of kynurenine pathway
metabolites in patients with eating disorders. Biol Psychiatry
37:512–520.
Fairburn CG, Marcus MD, Wilson GT (1993): Cognitive-behavioral therapy for binge eating and bulimia nervosa: A comprehensive treatment manual. In: Fairburn CG, Wilson GO,
editors. Binge Eating: Nature, Assessment and Treatment.
New York: Guildford Press, pp 361– 404.
Fairburn CG, Welch SL, Doll HA, Davies BA, O’Connor ME
(1997): Risk factors for bulimia nervosa. A community-based
case-control study. Arch Gen Psychiatry 54:509 –517.
Falk JR, Halmi KA (1982): Amenorrhea in anorexia nervosa:
Examination of the critical body weight hypothesis. Biol
Psychiatry 17:799 – 806.
Fallon BA, Walsh BT, Sadik C, Saoud JB, Lukasik V (1991):
Outcome and clinical course in inpatient bulimic women: A
2- to 9-year follow-up study. J Clin Psychiatry 52:272–278.
Fava M, Copeland PM, Schweiger U, Herzog DB (1989):
Neurochemical abnormalities in anorexia nervosa and bulimia nervosa. Am J Psychiatry 146:963–971.
Ferguson CP, La Via MC, Crossan PJ, Kaye WH (in press): Are
SSRIs effective in underweight anorexia nervosa? Int J Eat
Disord.
Fernstrom JD, Faller DV (1978): Neutral amino acids in the
brain: Changes in response to food ingestion. J Neurochem
30:1531–1538.
Fichter MM, Leibl K, Rief W, Brunner E, Schmidt-Auberger S,
Engel RR (1991): Fluoxetine versus placebo: A double-blind
study with bulimic inpatients undergoing intensive psychotherapy. Pharmacopsychiatry 24:1–7.
Fink G, Sumner BEH, Rosie R, Grace O, Quinn JP (1996):
Estrogen control of central neurotransmission: Effect on
mood, mental state and memory. Cell Mol Neurobiol 16:325–
344.
Fluoxetine Bulimia Nervosa Collaborative Study Group (1992):
Fluoxetine in the treatment of bulimia nervosa. Arch Gen
Psychiatry 49:139 –147.
BIOL PSYCHIATRY
1998;44:825– 838
835
Garner DM (1993): Pathogenesis of anorexia nervosa. Lancet
341:1631–1635.
Garner DM, Garfinkel PE, O’Shaughnessy M (1985): The
validity of the distinction between bulimia with and without
anorexia nervosa. Am J Psychiatry 142:581–587.
Gatti E, Porroni M, Noe D, Crovetti R, Testolin G (1993):
Plasma amino acids changes in obese patients on very
low-calorie diets. Int J Vitam Nutr Res 63:81– 85.
Gold PW, Gwirtsman H, Avgerinos PC, et al (1979): Cyproheptadine in anorexia nervosa. Br J Psychiatry 134:67–70.
Goldbloom DS, Olmsted MP (1993): Pharmacotherapy of bulimia nervosa and its associated attitudinal disturbances: The
clinical significance of change. Am J Psychiatry 150:770 –
774.
Goldbloom DS, Garfinkel PE, Katz R, Brown GM (1996): The
hormonal response to L-5-hydroxytryptophan in bulimia
nervosa. J Psychosom Res 40:289 –297.
Goldbloom DS, Hicks LK, Garfinkel PE (1990b): Platelet
serotonin uptake in bulimia nervosa. Biol Psychiatry 28:644 –
647.
Goldstein DJ, Wilson MG, Thompson VL, Potvin JH, Rampey
AH, and the Fluoxetine Bulimia Nervosa Research Group
(1995): Long-term fluoxetine treatment of bulimia nervosa.
Br J Psychiatry 166:660 – 666.
Goodall E (1990): Dieting, tryptophan and mood. BNF Nutr Bull
15:137–141.
Graheme-Smith DG (1992): Serotonin in affective disorders. Int
J Clin Psychopharmacol 6:S5–S13.
Gross HA, Ebert MH, Faden VB, Goldberg SC, Nee LE, Kaye
WH (1981): A double-blind controlled trial of lithium carbonate in primary anorexia nervosa. J Clin Psychopharmacol
1:376 –381.
Guicheney P, Leger D, Barrat J, et al (1988): Platelet serotonin
content and plasma tryptophan in peri- and postmenopausal
women: Variations with plasma oestrogen levels and depressive symptoms. Eur J Clin Invest 18:297–304.
Hadigan CM, Walsh TB, Buttinger C, Hollander E (1995):
Behavioral and neuroendocrine responses to meta-CPP in
anorexia nervosa. Biol Psychiatry 37:504 –511.
Haleem DJ, Haider S (1996): Food restriction decreases serotonin and its synthesis rate in the hypothalamus. NeuroReport
7:1153–1156.
Halmi KA, Casper RC, Eckert ED, Goldberg SC, Davis JM
(1979): Unique features associated with age of onset of
anorexia nervosa. Psychiatry Res 1:209 –215.
Halmi KA, Eckert E, LaDu TJ, Cohen J (1986): Anorexia
nervosa. Treatment efficacy of cyproheptadine and amitriptyline. Arch Gen Psychiatry 43:177–181.
Halmi KA, Eckert E, Marchi P, Sampugnaro V, Apple R, Chen
J (1991): Comorbidity of psychiatric diagnoses in anorexia
nervosa. Arch Gen Psychiatry 48:712–718.
Herzog DB, Keller MB, Strober M, Yeh CJ, Pal SY (1992): The
current status of treatment for anorexia nervosa and bulimia
nervosa. Int J Eat Disord 12:215–220.
Hibbeln JR, Salem NJ (1995): Dietary polyunsaturated fatty
acids and depression: When cholesterol does not satisfy. Am J
Clin Nutr 62:1–9.
Hoffman L, Halmi KA (1993): Psychopharmacology in the
836
BIOL PSYCHIATRY
1998;44:825– 838
treatment of anorexia and bulimia nervosa. Psychiatr Clin
North Am 16:767–778.
Holderness CC, Brooks-Gunn J, Warren WP (1994): Co-morbidity of eating disorders and substance abuse: Review of the
literature. Int J Eat Disord 16:1–34.
Holland AJ, Sicotte N, Treasure J (1988): Anorexia nervosa:
Evidence for a genetic basis. J Psychosom Res 32:561–571.
Horne RL, Ferguson JM, Pope HG, et al (1988): Treatment of
bulimia with bupropion: A multicenter controlled trial. J Clin
Psychiatry 49:262–266.
Hsu LKG, Sobkiewicz TA (1989): Bulimia nervosa: A four- to
six-year follow-up study. Psychol Med 19:1035–1038.
Huether G, Zhou D, Schmidt S, Wiltfang J, Ruther E (1997):
Long-term food restriction down-regulates the density of
serotonin transporters in the rat frontal cortex. Biol Psychiatry
41:1174 –1180.
Isager T, Brinch M, Kreiner S, Tolstrop K (1985): Death and
relapse in anorexia nervosa: Survival analysis of 151 cases.
J Psychiatr Res 19:515–521.
Jimerson D, Ebert M, Rittmaster R, Loriaux L, Chrousos GP
(1986): Abnormal hypothalamic-pituitary-adrenal function in
anorexia nervosa. N Engl J Med 314:1335–1342.
Jimerson DC, Lesem MD, Hegg AP, Brewerton TD (1990):
Serotonin in human eating disorders. Ann NY Acad Sci
600:532–544.
Jimerson DC, Lesem MD, Kaye WH, Brewerton TD (1992):
Low serotonin and dopamine metabolite concentrations in
cerebrospinal fluid from bulimic patients with frequent binge
episodes. Arch Gen Psychiatry 49:132–138.
Jimerson DC, Wolfe BE, Brotman AW, Metzger ED (1996):
Medications in the treatment of eating disorders. Psychiatr
Clin North Am 19:739 –754.
Jimerson DC, Wolfe BE, Metzger ED, Finkelstein DM, Cooper
TB (1997): Decreased serotonin function in bulimia nervosa.
Arch Gen Psychiatry 54:529 –534.
Johnson-Sabine E, Reiss D, Dayson D (1992): Bulimia nervosa:
A 5-year follow-up study. Psychol Med 22:951–959.
Kaye WH, Weltzin TE (1991): Neurochemistry of bulimia
nervosa. J Clin Psychiatry 52:21–28.
Kaye WH, Ebert MH, Raleigh M, Lake R (1984): Abnormalities
in CNS monoamine metabolism in anorexia nervosa. Arch
Gen Psychiatry 41:350 –355.
Kaye WH, Gwirtsman HE, George DT, et al (1987): Elevated
cerebrospinal fluid levels of immunoreactive corticotropinreleasing hormone in anorexia nervosa: Relation to state of
nutrition, adrenal function and intensity of depression. J Clin
Endocrinol Metab 64:203–208.
Kaye WH, Gwirtsman HE, Brewerton TD, George DT, Wurtman
RJ (1988a): Bingeing behavior and plasma amino acids: A
possible involvement of brain serotonin in bulimia nervosa.
Psychiatry Res 3:31– 43.
Kaye WH, Gwirtsman HE, George DT, Ebert H (1988b):
CSF-5HIAA concentrations in anorexia nervosa: Reduced
values in underweight subjects normalize after weight gain.
Biol Psychiatry 23:102–105.
Kaye WH, Berrettini W, Gwirtsman H, George DT (1990):
Altered cerebrospinal fluid neuropeptide Y and peptide YY
immunoreactivity in anorexia and bulimia nervosa. Arch Gen
Psychiatry 47:548 –556.
W. Kaye et al
Kaye WH, Gwirtsman HE, George DT, Ebert H (1991a): Altered
serotonin activity in anorexia nervosa after long-term weight
recovery. Arch Gen Psychiatry 48:556 –562.
Kaye WH, Weltzin T, Hsu LKG, Bulik CM (1991b): An open
trial of fluoxetine in patients with anorexia nervosa. J Clin
Psychiatry 52:464 – 471.
Kaye WH, Lilenfeld LR, Plotnikov K, et al (1996): Bulimia
nervosa and substance dependence: Association and family
transmission. Alcohol Clin Exp Res 20:878 – 881.
Kaye WH, Greeno CG, Moss H, et al (in press): Alterations in
serotonin activity and psychiatric symptomatology after recovery from bulimia nervosa. Arch Gen Psychiatry.
Kaye WH, Nagata T, Weltzin TE, et al (in submission): Successful outcome of restricting-type anorexia nervosa after the
double-blind placebo-controlled administration of fluoxetine.
Keel PK, Mitchell JE (1997): Outcome in bulimia nervosa. Am J
Psychiatry 154:313–321.
Kendler KS, MacLean C, Neals M, Kessler R, Heath A, Eaves L
(1991): The genetic epidemiology of bulimia nervosa. Am J
Psychiatry 148:1627–1637.
Kirkley BG, Schneider JA, Agras WS, Bachman JA (1985):
Comparison of two group treatments for bulimia. J Consult
Clin Psychol 53:43– 48.
Lacey JH, Crisp AH (1980): Hunger, food intake and weight:
The impact of clomipramine on a refeeding anorexia nervosa
population. Postgrad Med J 56(suppl 1):S79 –S85.
Leibowitz SF (1986): Brain monoamines and peptides: Role in
the control of eating behavior. Fed Proc 45:1396 –1403.
Leitenberg J, Rosen JC, Wolf J, Vara LS, Detzer MJ, Srebnik D
(1994): Comparison of cognitive-behavior therapy and desipramine in the treatment of bulimia nervosa. Behav Res Ther
32:37– 45.
Levitan RD, Kaplan AS, Joffe RT, Levitt AJ, Brown GM (1997):
Hormonal and subjective responses to intravenous metachlorophenylpiperazine in bulimia nervosa. Arch Gen Psychiatry 54:521–527.
Lilenfeld LR, Kaye WH, Greeno CG, et al (1997): Psychiatric
disorders in women with bulimia nervosa and their firstdegree relatives: Effects of comorbid substance dependence.
Int J Eat Disord 22:253–264.
Lilenfeld LR, Kaye WH, Greeno CG, et al (1998): A controlled
family study of anorexia and bulimia nervosa: Psychiatric and
effects of proband comorbidity. Arch Gen Psychiatry 55:603–
610.
Marazziti D, Macchi E, Rotondo A, Franco Placidi G, Cassano
GB (1988): Involvement of serotonin system in bulimia. Life
Sci 43:2131–2126.
McBride PA, Anderson GM, Khait VD, Sunday SR, Halmi KA
(1991): Serotonergic responsivity in eating disorders. Psychopharmacol Bull 27:365–372.
McClain C, Stuart M, Kasarskis E, Humphries L (1993): Zinc,
appetite regulation and eating disorders. Prog Clin Biol Res
380:47– 64.
McKenzie JM, Joyce PR (1992): Hospitalization for anorexia
nervosa. Int J Eat Disord 11:235–241.
McQueen JK, Wilson H, Dow RC, Fink GJ (1996): Oestradiol-17 beta increases serotonin transporter binding sites
(SERT) and SERT mRNA expression in discrete regions of
the female rat brain. J Physiol (Lond) 495:114.
Serotonin and SSRIs in Eating Disorders
Mira M, Stewart PM, Abraham SF (1989): Vitamin and trace
element status of women with disordered eating. Am J Clin
Nutr 50:940 –944.
Mitchell JE, de Zwaan M (1993): Pharmacological treatments of
binge eating. In: Fairburn CG, Wilson GO, editors. Binge
Eating: Nature, Assessment and Treatment. New York:
Guildford Press, pp 250 –269.
Mitchell JE, Pyle RL, Eckert ED, Hatsukami D, Pomeroy C,
Zimmerman R (1990): A comparison study of antidepressants
and structured intensive group psychotherapy in the treatment
of bulimia nervosa. Arch Gen Psychiatry 47:149 –157.
Mitchell JE, Raymond N, Specker S (1993): A review of
controlled trials of pharmacotherapy and psychotherapy in the
treatment of bulimia nervosa. Int J Eat Disord 14:229 –247.
Mitchell JE, de Zwaan M, Crow S (1997): Psychopharmacology
of eating disorders. Ballieres Clin Psychiatry 3:217–234.
Morley JE, Blundell JE (1988): The neurobiological basis of
eating disorders: Some formulations. Biol Psychiatry 23:53–
78.
Muldoon MF, Kaplan JR, Manuck SB (1992): Effects of a
low-fat diet on brain serotonergic responsivity in cynomolgus
monkeys. Biol Psychiatry 331:739 –742.
Norring CE, Sohlberg SS (1993): Outcome, recovery, relapse
and mortality across six years in patients with clinical eating
disorders. Acta Psychiatr Scand 87:437– 444.
Nunez C, Moreiras O, Carbajal A (1995): Guidelines for the
dietetic treatment of patients with anorexia nervosa based on
assessment of their nutritional status. Rev Clin Esp 195:226 –
232.
O’Dwyer AM, Lucey JV, Russell GFM (1996): Serotonin
activity in anorexia nervosa after long term weight restoration: Response to D-fenfluramine challenge. Psychol Med
26:353–359.
Pirke KM, Pahl J, Schweiger WM (1985): Metabolic and
endocrine indices of starvation in bulimia: A comparison with
anorexia nervosa. Psychiatry Res 15:33–39.
Pirke KM, Schweiger U, Strowitzki T, Tusch RJ (1989): Dieting
causes menstrual irregularities in normal weight young
women through impairment of episodic luteinizing hormone
secretion. Fertil Steril 5:263–268.
Rock CL, Curran-Celentano J (1994): Nutritional disorders of
anorexia nervosa: A review. Int J Eat Disord 15:187–203.
Rock CL, Vasantharajan S (1995): Vitamin status of eating
disorder patients: Relationship to clinical indices and effect of
treatment. Int J Eat Disord 18:257–262.
Russell GFM (1970): Anorexia nervosa: Its identity as an illness.
In: Price JE, editor. Modern Trends in Psychological Medicine. London: Butterworth, pp 131–164.
Russell GFM (1988): The role of drug treatment in anorexia and
bulimia nervosa. Psychopharmacology (Berl) 96:124.
Russell GFM, Szmukler GI, Dare C, Eisler I (1987): An
evaluation of family therapy in anorexia nervosa and bulimia
nervosa. Arch Gen Psychiatry 44:1047–1056.
Schuckit MA, Tipp JE, Anthenall RM, Bucholz KK, Hesselbrock
VM, Nurnberger JI (1996): Anorexia nervosa and bulimia
nervosa in alcohol-dependent men and women and their
relatives. Am J Psychiatry 153:74 – 82.
Schweiger U, Ficther M (1997): Eating disorders: Clinical
BIOL PSYCHIATRY
1998;44:825– 838
837
presentation, classification and etiologic models. Ballieres
Clin Psychiatry 3:199 –216.
Schweiger U, Warnhoff M, Pahl J, Pirke KM (1986): Effects of
carbohydrate and protein meals on plasma large neutral
amino acids, glucose, and insulin plasma levels of anorectic
patients. Metabolism 35:938 –943.
Sharma SK, Dakshinamurti K (1994): Effects of serotonergic
agents on plasma prolactin levels in pyridoxine-deficient
adult male rats. Neurochem Res 19:687– 692.
Sharp CW, Freeman CP (1997): Medical complications and
management. Ballieres Clin Psychiatry 3:303–318.
Soubrie P (1986): Reconciling the role of central serotonin
neurosis in human and animal behavior. Behav Brain Sci
9:319 –363.
Srinivasagam NM, Kaye WH, Plotnikov KH, Greeno C, Weltzin
TE, Rao R (1995): Persistent perfectionism, symmetry and
exactness after recovery from anorexia nervosa. Am J Psychiatry 152:1630 –1634.
Strober M (1980): Personality and symptomatological features in
young, nonchronic anorexia nervosa patients. J Psychosom
Res 24:353–359.
Strober M (1991): Family genetic studies of eating disorders.
J Clin Psychiatry 52:9 –12.
Strober M, Salkin B, Burroughs J, Morrell W (1982): Validity of
the bulimia-restrictor distinction in anorexia nervosa. J Nerv
Ment Dis 170:345–351.
Strober M, Lampert C, Morrell W, Burroughs J (1990): A
controlled family study of anorexia nervosa: Evidence of
familial aggregation and of shared transmission with affective
disorders. Int J Eat Disord 9:239 –253.
Strober M, Freeman R, Morrell W (1997a): The long term course
of severe anorexia nervosa in adolescents: Survival analysis
of recovery, relapse and outcome predictors over 10-15 years
in a prospective study. Int J Eat Disord 22:339 –360.
Strober M, Freeman R, DeAntonio M, Lampert C, Diamond J
(1997b): Does adjunctive fluoxetine influence the posthospital course of restricting-type anorexia nervosa? A
24-month prospective, longitudinal follow up and comparison with historical controls. Psychopharmacol Bull 33:
425– 431.
Sullivan PF (1995): Mortality in anorexia nervosa. Arch Gen
Psychiatry 152:1073–1074.
Theander S (1985): Outcome and prognosis in anorexia nervosa
and bulimia. J Psychiatr Res 19:493–508.
Tollefson GD (1995): Selective serotonin reuptake inhibitors. In:
Schatzberg AF, Nemeroff CB, editors. Textbook of Psychopharmacology. Washington, DC: American Psychiatric Press,
pp 161–182.
Treasure J, Campbell I (1994): The case for biology in the
aetiology of anorexia nervosa. Psychol Med 24:3– 8.
Treasure J, Holland AJ (1989): Genetic vulnerability to eating
disorders: Evidence from twin and family studies. In: Remschmidt H, Schmidt MH, editors. Child and Youth Psychiatry:
European Perspectives. New York: Hogrefe & Huber, pp
59–68.
Treasure J, Schmidt U, Troop N, Todd G (1996): Sequential
treatment for bulimia nervosa incorporating a self-care manual. Br J Psychiatry 168:94 –98.
838
BIOL PSYCHIATRY
1998;44:825– 838
Vigersky RA, Loriaux DL (1977): The effect of cyproheptadine
in anorexia nervosa: A double-blind trial. In: Vigersky RA,
editor. Anorexia Nervosa. New York: Raven Press, pp 349 –
356.
Vitousek K, Manke F (1994): Personality variables and disorders
in anorexia nervosa and bulimia nervosa. J Abnorm Psychol
103:137–147.
Wakeling A, DeSouza FA, Gore MBR, Sabur M, Kingstone D,
Boss AMB (1979): Amenorrhoea, body weight and serum
hormone concentrations, with particular reference to prolactin
and thyroid hormones in anorexia nervosa. Psychol Med
9:265–272.
Walsh BT (1991a): Fluoxetine treatment of bulimia nervosa.
J Psychosom Res 35:33– 40.
Walsh BT (1991b): Psychopharmacological treatment of bulimia
nervosa. J Clin Psychiatry 52(suppl 10):34 –38.
Walsh BT, Stewart JW, Roose SP, Gladis M, Glassman AH
(1984): Treatment of bulimia with phenelzine. Arch Gen
Psychiatry 41:1105–1109.
Walsh AES, Oldman AD, Franklin M, Fairburn CG, Cowen PJ
(1995): Dieting decreases plasma tryptophan and increases
W. Kaye et al
the prolactin response to d-fenfluramine in women but not
men. J Affect Disord 33:89 –97.
Walsh BT, Wilson GO, Leob KL, et al (1997): Medication and
psychotherapy in the treatment of bulimia nervosa. Am J
Psychiatry 154:523–531.
Walters E, Kendler K (1995): Anorexia nervosa and anorexiclike syndromes in a population-based female twin sample.
Am J Psychiatry 152:64 –71.
Weizman R, Carmi M, Tyano S, Apter A, Rehavi M (1986):
High affinity [3H]imipramine binding and serotonin uptake to
platelets of adolescents suffering from anorexia nervosa. Life
Sci 38:1235–1242.
Weltzin TE, Fernstrom JD, McConaha C, Kaye WH (1994):
Acute tryptophan depletion in bulimia: Effects on large
neutral amino acids. Biol Psychiatry 35:388 –397.
Wolf EM, Crowther JH (1992): An evaluation of behavioral and
cognitive-behavioral group intervention for the treatment of
bulimia nervosa in women. Int J Eat Disord 11:3–16.
Wolfe BE, Metzger ED, Stollar C (1997): The effects of dieting
on plasma tryptophan concentration and food intake in
healthy women. Physiol Behav 61:537–541.