Clinical Aspects of Eating Disorders

Anorexia Nervosa
The term anorexia nervosa literally means “lack of appetite
induced by nervousness.” This definition is something of a
misnomer, however, as a lack of appetite is neither the core
difficulty nor necessarily even true. At the heart of anorexia
nervosa is an intense fear of gaining weight or becoming fat,
combined with behaviors that result in a significantly low
body weight. The DSM-5 criteria for anorexia nervosa are
shown in the DSM criteria box. An important change from
DSM-IV to DSM-5 is that in DSM-5 amenorrhea (cessation
of menstruation) is no longer required for a person to be given
the diagnosis. Studies have suggested that women who con-
tinue to menstruate but meet all the other diagnostic criteria
for anorexia nervosa are very similar psychologically to women
who have amenorrhea and have ceased menstruating (Attia &
Roberto, 2009). Amenorrhea is also not
a criterion that can be used for males, nor
can it be assessed in prepubescent girls or in (A) (B)
women who use hormonal contraceptives. figure 9.1
Watch the Video Natasha: Anorexia on Gull’s patient with anorexia nervosa. (A) Before treatment.
MyPsychLab (B) After treatment.
Although we may think of anorexia nervosa as a Source: Gull, W. (1888). Anorexia nervosa. Lancet, pp. i, 516–17.
modern problem, it is centuries old. Descriptions of
extreme fasting that were probably signs of anorexia
nervosa can be found in early religious literature 18-year-old girl and a 16-year-old boy, who suffered from a
(Vandereycken, 2002). The first known medical account “nervous consumption” that resulted in wasting of body tis-
of anorexia nervosa, however, was published in 1689 by sue. The female patient eventually died because she refused
Richard Morton (see Silverman, 1997, for a good general treatment.
historical overview). Morton described two patients, an The disorder did not receive its current name until 1873,
when Charles Lasègue in Paris and Sir William Gull in London
independently described the clinical syndrome. In his last publi-
cation on the condition, Gull (1888) described a 14-year-old girl
criteria for who began “without apparent cause, to evince a repugnance to
Anorexia Nervosa DSM-5 food; and soon afterwards declined to take any whatever, except
half a cup of tea or coffee.” After being prescribed to eat light
A. Restriction of energy intake relative to requirements, lead- food every few hours, the patient made a good recovery. Gull’s
ing to a significantly low body weight in the context of illustrations of the patient before and after treatment appear in
age, sex, developmental trajectory, and physical health.
Significantly low weight is defined as a weight that is less
Figure 9.1.
than minimally normal or, for children and adolescents, less Even though they may look painfully thin or even emaci-
than that minimally expected. ated, many patients with anorexia nervosa deny having any
B. Intense fear of gaining weight or of becoming fat, or problem. Indeed, they may come to feel fulfilled by their
persistent behavior that interferes with weight gain, even weight loss. Despite this quiet satisfaction, however, they may
though at a significantly low weight. feel ambivalent about their weight. Efforts may be made to
C. Disturbance in the way in which one’s body weight or conceal their thinness by wearing baggy clothes or carrying
shape is experienced, undue influence of body weight or hidden bulky objects so that they will weigh more when mea-
shape on self-evaluation, or persistent lack of recognition
sured by others. Patients with anorexia nervosa may even resort
of the seriousness of the current low body weight.
to drinking large amounts of water to increase their weight
Source: Reprinted with permission from the Diagnostic and Statistical Manual of
Mental Disorders, Fifth Edition, (Copyright 2013). American Psychiatric Association.
temporarily.
There are two types of anorexia nervosa: the restricting
type and the binge-eating/purging type. The central difference

eating disorders and obesity CHAPTER 9  295

M09_BUTC4286_16_SE_C09.indd 295 10/06/13 4:42 PM

 Patients with anorexia nervosa may
be emaciated while still denying table 9.1
that they have any problems with
their weight. They will go to great
Distorted Thinking in Anorexia Nervosa
lengths to conceal their thinness by “I have a rule when I weigh myself. If I’ve gained then I starve the
wearing baggy clothes, or by drink- rest of the day. But if I’ve lost, then I starve too.”
ing massive amounts of water prior
“Bones define who we really are, let them show.”
to being weighed (in a hospital set-
ting, for instance). “An imperfect body reflects an imperfect person.”
“Anorexia is not a self-inflicted disease, it’s a self-controlled
lifestyle.”
“It’s not deprivation, it’s liberation.”

case reminds us that eating disorders can occur in young chil-

dren as well as in boys. It further highlights the high comorbidity
between eating disorders and obsessive-compulsive symptoms and
personality traits, which we will subsequently discuss. In fact, Tim
warrants an additional diagnosis of obsessive-compulsive disorder.

Tim: Obsessed With His Weight Tim is 8 years

old. He is extremely concerned about his weight and weighs him-
between these two subtypes concerns the way in which patients self daily. Tim complains that he is too fat, and if he does not lose
maintain their very low weight. In the restricting type, every weight, he cuts back on food. He has lost 10 pounds in the past
effort is made to limit the quantity of food consumed. Caloric year and still feels that he is too fat. However, it is clear that he is
intake is tightly controlled. Patients often try to avoid eating in underweight. In desperation, his parents have removed the
the presence of other people. When they are at the table, they scales from the house; as a result, Tim is keeping a record of the
may eat excessively slowly, cut their food into very small pieces, calories that he eats daily. He spends an excessive amount of
or dispose of food secretly (Beaumont, 2002). time on these calculations, checking and rechecking that he has
The relentless restriction of food intake is not possible for all done it just right.
patients with anorexia nervosa. Patients with the binge-eating/ In addition, Tim is described as being obsessed with clean-
purging type of anorexia nervosa differ from patients with liness and neatness. Currently, he has no friends because he
restricting anorexia nervosa because they either binge, purge, or refuses to visit them, feeling that their houses are “dirty.” He
binge and purge. A binge involves an out-of-control consump- becomes upset when another child touches him. He is always
tion of an amount of food that is far greater than what most checking whether he is doing things the way they “should” be
people would eat in the same amount of time and under the done. He often feels very agitated and anxious about this. Each
same circumstances. These binges may be followed by efforts to day, he has to wake up at least 2 hours before leaving for school
purge, or remove from their bodies, the food they have eaten. in order to allow himself sufficient time to get ready. Recently, he
Methods of purging commonly include self-induced vomiting woke up at 1:30 a.m. to prepare for school.
or misusing laxatives, diuretics, and enemas. Other compensa- Source: Adapted with permission from R. L. Spitzer (2000). DSM-IV-TR
Casebook: A Learning Companion to the Diagnostic and Statistical Manual of
tory behaviors that do not involve purging are excessive exercise Mental Disorders, Fourth Edition, Text Revision. Washington, DC: American
or fasting. Even purging strategies, however, do not prevent the Psychiatric Association.
absorption of all calories from food.
Indicative of the distorted values of eating-disordered
patients, (see Table 9.1 for personal comments that provide Because the artistic standards of their profession emphasize
examples of this) those with the restricting type of anorexia ner- a slender physique, ballet dancers are at an especially high risk
vosa are often greatly admired by others with eating disorders. for eating disorders (see the case study of Ms. R. on p. 297).
One patient reported that she had not been “successful” in her Gelsey Kirkland, who developed an eating disorder while
anorexia nervosa because of her failure to reach an extremely she was a premier ballerina with the New York City Ballet,
low weight. Her belief was that the hallmark of a truly successful described the existence of a “concentration camp aesthetic”
person with anorexia nervosa was death from starvation, and that within the company. This was no doubt fostered by the famous
patients who were able to accomplish this should somehow be choreographer George Balanchine, who, as described by Kirk-
revered (see Bulik & Kendler, 2000). land in her autobiography, tapped her on the ribs and sternum
In the example that follows, we describe the case of Tim, who after one event and exhorted that he “must see the bones”
is suffering from the restricting type of anorexia nervosa. Tim’s (Kirkland, 1986, pp. 55–56).

296  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 296 10/06/13 4:42 PM

 goes to great lengths to avoid such situations. This places great
limitations on her social life.
Source: Adapted with permission from DSM-IV Case Studies: A Clinical Guide
to Differential Diagnosis, Copyright 1996. American Psychiatric Association.

Bulimia Nervosa
Bulimia nervosa is characterized by uncontrollable binge
eating and efforts to prevent resulting weight gain by using
inappropriate behaviors such as self-induced vomiting and
excessive exercise. Bulimia nervosa was recognized as a
psychiatric syndrome relatively recently. The British psychia-
trist G. F. M. Russell (1997) proposed the term in 1979, and it
was adopted into the DSM in 1987. The word bulimia comes
from the Greek bous (which means “ox”), and limos (“hunger”).
It is meant to denote a hunger of such proportions that the
person “could eat an ox.”
Ballet dancers are at very high risk for developing eating disorders. The DSM-5 criteria for bulimia nervosa are shown in the
According to Gelsey Kirkland, once the premier ballerina at the New York DSM box below. Compared to DSM-IV, the diagnostic criteria
City Ballet, the value placed on being thin can create a “concentration
camp aesthetic” supported and encouraged by the dance company. for bulimia nervosa have been relaxed. Binge eating and purg-
ing now have to occur on average once a week (instead of twice
a week) over a 3-month period. This change was made after
Ms. R.: Going to Extremes Ms. R. is a very thin, research showed that people with sub-threshold bulimia nervosa
19-year-old, single ballet student who comes in at the insistence were remarkably similar to those who had the full syndrome
of her parents for a consultation concerning her eating behavior. (Eddy, Doyle, et al., 2008; Fairburn et al., 2007).
Ms. R. has had a lifelong interest in ballet. She began to attend
classes at age 5 and by age 8 was recognized by her teachers as
having impressive talent. Since the age of 14 she has been a
member of a national ballet company. criteria for
Ms. R. has had difficulties with eating since age 15 when, for Bulimia Nervosa DSM-5
reasons she is unable to explain, she began to induce vomiting
after what she felt was overeating. The vomiting was preceded A. Recurrent episodes of binge eating. An episode of binge
by many years of persistent dieting, which began with the eating is characterized by both of the following:
encouragement of her ballet teacher. Over the past 3 years, 1. Eating, in a discrete period of time (e.g., within any
Ms. R.’s binges have occurred once a day in the evening and have 2-hour period), an amount of food that is definitely
been routinely followed by self-induced vomiting. The binges larger than what most individuals would eat in a
similar period of time under similar circumstances.
consist of dozens of rice cakes or, more rarely, half a gallon of ice
2. A sense of lack of control over eating during the epi-
cream. Ms. R. consumes this food late at night after her parents
sode (e.g., a feeling that one cannot stop eating or
have gone to bed. For some time, Ms. R.’s parents have been control what or how much one is eating).
concerned that their daughter has a problem with her eating, but
B. Recurrent inappropriate compensatory behaviors in order
she consistently denied difficulties until about a month before this to prevent weight gain, such as self-induced vomiting;
consultation. misuse of laxatives, diuretics, or other medications; fasting;
Ms. R. reached her full height of 5 feet 8 inches at age 15. Her or excessive exercise.
greatest weight was 120 pounds at age 16, which she describes C. The binge eating and inappropriate compensatory behav-
as being fat. For the past 3 years, her weight has been reasonably iors both occur, on average, at least once a week for
stable in the range of 100 to 104 pounds. She exercises regularly 3 months.
as part of her profession, and she denies using laxatives, diuret- D. Self-evaluation is unduly influenced by body shape and
weight.
ics, or diet pills as a means of weight control. Except when she is
binge eating, she avoids the consumption of foods high in fat or E. The disturbance does not occur exclusively during
episodes of anorexia nervosa.
sugar.
Since age 15, she has been a strict vegetarian and consumes Source: Reprinted with permission from the Diagnostic and Statistical Manual of
Mental Disorders, Fifth Edition, (Copyright 2013). American Psychiatric Association.
no meat or eggs and little cheese. For the past 3 or 4 years, Ms.
R. has been uncomfortable eating in front of other people and

eating disorders and obesity CHAPTER 9  297

M09_BUTC4286_16_SE_C09.indd 297 10/06/13 4:42 PM

 The clinical picture of the binge-eating/purging type of first time, she started binge eating. Distressed by the increase in
anorexia nervosa has much in common with bulimia nervosa. her weight and by increasingly frequent binge eating, Catherine
Indeed, some researchers have argued that the bulimic type of began vomiting in a desperate attempt to lose weight.
anorexia nervosa should really be considered another form of Over the last six months Catherine has skipped breakfast and
bulimia nervosa. The difference between a person with buli- lunch, eaten a normal evening meal with her family, but gone on
mia nervosa and a person with the binge-eating/purging type to binge eat in the late evening usually two or three times a week.
of anorexia nervosa is weight. By definition, the person with She binges in response to feeling upset and worried. In a recent
anorexia nervosa is severely underweight. This is not true of the typical binge she ate four slices of toast with butter and jam, six
person with bulimia nervosa. packets of potato chips, three large bars of chocolate, half a box
Consequently, if the person who binges or purges also meets of cereal and a large bowl of ice cream. She had been feeling
criteria for anorexia nervosa, the diagnosis is anorexia nervosa rejected by a friend. Catherine drinks large quantities of water to
(binge-eating/purging type) and not bulimia nervosa. In other help her induce vomiting after bingeing. More recently she has
words, the anorexia nervosa diagnosis “trumps” the bulimia started to take 30 to 40 laxatives, as well as vomiting. She has a
nervosa diagnosis. This is because there is a far greater mortality strenuous exercise regime, including 500 daily sit-ups and an aero-
rate associated with anorexia nervosa than with bulimia nervosa. bic workout. Food is divided rigidly into good and bad categories;
Recognizing this, the DSM requires that the more severe form of food in the bad category (such as cookies, chocolate, and cheese)
eating pathology take precedence diagnostically. is not allowed.
People with anorexia nervosa and bulimia nervosa share a Catherine has become increasingly self-conscious. She
common fear of being or becoming fat. However, unlike patients refuses to weigh herself, has given up swimming (she used to
with anorexia nervosa, those with bulimia nervosa are typically of swim on her school team) and if she goes shopping she will not
normal weight or sometimes even slightly overweight. The fear try on clothes unless she has privacy. Catherine dislikes her body.
of becoming fat helps explain the development of bulimia ner- She is preoccupied with her shape and reports that a little voice
vosa. Bulimia typically begins with restricted eating motivated in her head constantly says, “You’re fat and ugly, I can’t stand the
by the desire to be slender. During these early stages, the person way you look.” Catherine worries that her friends will think less of
diets and eats low-calorie foods. Over time, however, the early her if she does not lose weight (when out with friends she avoids
resolve to restrict gradually erodes, and the person starts to eat eating and sucks her tummy in to the point of pain in order to
“forbidden foods” such as potato chips, pizza, cake, ice cream, appear thinner). She feels very bad about herself if she thinks she
and chocolate. Of course, some patients binge on whatever food has gained weight.
is available, including such things as raw cookie dough. During Source: Adapted from Cooper, Todd, & Wells, (2000). Bulimia Nervosa:
A Cognitive Therapy Programme. London: Jessica Kingsley Publishers.
an average binge, someone with bulimia nervosa may consume
as many as 4,800 calories (Johnson et al., 1982). After the binge,
in an effort to manage the breakdown of self-control, the person
begins to vomit, fast, exercise excessively, or abuse laxatives. This
pattern then persists because, even though those with bulimia Binge Eating Disorder
nervosa are disgusted by their behavior, the purging alleviates the A new addition to DSM-5 is the diagnosis of binge eating
fear of gaining weight that comes from eating. disorder (BED). Previously, in DSM-IV, binge eating was
Whereas people with anorexia nervosa often deny the seri- given provisional status. This encouraged researchers to study
ousness of their disorder and are surprised by the shock and con- it. Research has now supported the idea that BED is a distinct
cern with which others view their emaciated conditions, those clinical syndrome (Wonderlich et al., 2009). It has therefore
with bulimia nervosa are often preoccupied with shame, guilt, entered the DSM as a new formal diagnosis. Watch the
and self-deprecation. They make efforts to conceal their behavior Video Stacy: Binge Eating Disorder on MyPsychLab
as they struggle (often unsuccessfully) to master their urges to Although BED has some clinical features in common with
binge. The case described below depicts a typical pattern. bulimia nervosa, there is an important difference.
After a binge (which may be at a level comparable to that of
Catherine: Distressed by Her Weight Catherine a patient with bulimia nervosa), the person with BED does not
is 20 years old. Catherine has been concerned about her weight engage in any form of inappropriate “compensatory” behavior.
and shape for several years. As a teenager she felt plump and This might include purging, using laxatives, or even exercising
was frequently on a diet, even though she was within the normal to limit weight gain (Table 9.2 summarizes the similarities and
weight range for her age and height. As she became older, these differences between different types of eating disorders). There is
concerns became more serious. She severely restricted her food also much less dietary restraint in BED than is typical of either
intake and started punishing exercise regimens. At one stage she bulimia nervosa or anorexia nervosa (Wilfley et al., 2000). Not
became significantly underweight. Medical help led to weight surprisingly, binge-eating disorder is associated with being over-
gain, which was quickly followed by more dieting and, for the weight or even obese (Hudson et al., 2007; Pike et al., 2001),

298  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 298 10/06/13 4:42 PM

 9.2
table
Comparing Anorexia Nervosa, Bulimia Nervosa, and Binge Eating Disorder
Binge Eating
Symptom Anorexia Nervosa Bulimia Nervosa Disorder
Restricting Binge/Purge Purging Nonpurging
Body weight Markedly low Markedly low Normal weight or Normal weight or Typically overweight
slightly overweight slightly overweight or obese
Fear of weight gain, yes yes yes yes no
becoming fat
Body image Distorted perception Distorted perception Overconcerned with Overconcerned with May be unhappy
weight weight with body and
weight
Binge eating no yes yes yes yes
Purging no yes yes no no
Use of nonpurging yes yes yes yes no
methods to avoid
weight gain
Feeling of lack of no During binges yes yes yes
control over eating

although weight is not a factor involved in making the diagnosis.

criteria for
Interestingly, individuals with binge-eating disorder are more
Binge-Eating Disorder DSM-5 likely to have overvalued ideas about the importance of weight
and shape than overweight or obese patients who do not have
A. Recurrent episodes of binge eating. An episode of binge binge-eating disorder. In this respect, they also resemble people
eating is characterized by both of the following: with bulimia nervosa (Allison et al., 2005).
1. Eating, in a discrete period of time (e.g., within any
2-hour period), an amount of food that is definitely Ms. A.: Feeling Out of Control Ms. A. was a
larger than what most people would eat in a similar 38-year-old African American woman who was single, lived alone,
period of time under similar circumstances.
and was employed as a personnel manager at a hotel in
2. A sense of lack of control over eating during the epi- New York. Her height was 6 feet, and she weighed 292 pounds
sode (e.g., a feeling that one cannot stop eating or
(BMI = 39.6; see Table 9.4 on p. 310) when she was initially seen at
control what or how much one is eating).
the eating disorders clinic. Her chief reason for going to the clinic
B. The binge-eating episodes are associated with three
was that she felt her eating was out of control and, as a result, she
(or more) of the following:
had gained approximately 80 pounds over the previous year.
1. Eating much more rapidly than normal.
Ms. A. reported a lifetime history of obesity and a history of
2. Eating until feeling uncomfortably full.
binge eating beginning at approximately age 11. At her intake
3. Eating large amounts of food when not feeling session, she described her eating. She felt out of control and ate
physically hungry.
large amounts of food nearly every day, typically in the evenings
4. Eating alone because of feeling embarrassed by how when she was on her way home from work or alone at home.
much one is eating.
She tended to feel out of control throughout the day, which con-
5. Feeling disgusted with oneself, depressed, or very tributed to her daily snacking on three or four regular candy bars
guilty afterward.
or three or four medium cookies and one ice cream bar. Ms. A.
C. Marked distress regarding binge eating is present. would consequently feel that a binge episode was inevitable.
D. The binge eating occurs, on average, at least once a week A typical binge episode consisted of two pieces of
for 3 months. chicken, one small bowl of salad, two servings of mashed
E. The binge eating is not associated with the recurrent use potatoes, one hamburger, one large serving of french fries,
of inappropriate compensatory behavior as in bulimia ner- one fast-food serving of apple pie, one large chocolate shake,
vosa and does not occur exclusively during the course of
one large bag of potato chips, and 15 to 20 small cookies—all
bulimia nervosa or anorexia nervosa.
within a 2-hour period. During her binge episodes, Ms. A. ate
Source: Reprinted with permission from the Diagnostic and Statistical Manual of
Mental Disorders, Fifth Edition, (Copyright 2013). American Psychiatric Association.
much more rapidly than usual until she felt uncomfortably full,
ate large amounts of food when she didn’t feel physically hun-
gry, ate alone because she was embarrassed by the ­quantity

eating disorders and obesity CHAPTER 9  299

M09_BUTC4286_16_SE_C09.indd 299 10/06/13 4:42 PM

 of food she ate, and felt disgusted with herself and very guilty
after eating. She was also extremely distressed about her
weight. She sadly acknowledged that her weight and shape
were the most important factors in determining her feelings
about herself.
Source: Adapted with permission from the American Journal of Psychiatry,
Copyright 2000. American Psychiatric Association.

Age of Onset and Gender Differences

Eating disorders are often considered to be “modern” disorders,
yet pathological patterns of eating date back several centuries
(Silverman, 1997). St. Catherine of Sienna began to starve her-
self when she was around 16 years of age. She died in 1380 (at Competitive eating contests, such as the hot dog eating contest that is held
every year in Coney Island, New York, attract many spectators. During a
the age of 32 or 33) because she refused to consume either food 10-minute time span, some competitors are able to eat 60 or more hot dogs.
or water (Keel & Klump, 2003). Moreover, as far back as the sec- Do you think this is a healthy sport? What messages does it send about food?
ond century, the Greek physician Galen referred to a syndrome In what ways is competitive eating different from binge eating disorder?
characterized by overeating, vomiting, and fainting, which he
termed bulimos (see Ziolko, 1996). It was not until the 1970s underdiagnosed in the past because of the stereotype that they are
and 1980s, however, that eating disorders began to attract signif- female disorders. Another reason for the underdiagnosis of eating
icant attention. Clinicians began seeing more and more patients disorders in men is the gender bias in the DSM criteria. These
with pathological eating patterns. It soon became apparent that emphasize the type of weight and shape concerns (e.g., desire to
this was an important clinical problem. be thin) and methods of weight control (dieting) that are more
Anorexia nervosa and bulimia nervosa do not occur in appre- typical of women. For men, body dissatisfaction often involves a
ciable numbers before adolescence. Children as young as 7, though, wish to be more muscular. Over-exercising as a means of weight
have been known to develop eating disorders, especially anorexia control is also more common in men. As a result, men are less
nervosa (Bryant-Waugh & Lask, 2002). Anorexia nervosa is most likely to recognize that they have an eating disorder, are more
likely to develop in 15- to 19-year-olds. For bulimia nervosa, the likely to be misdiagnosed when they do, and are less likely to
age group at highest risk is young women falling in the age range receive specialist treatment (Jones & Morgan, 2010).
of 20 to 24 (Hoek & van Hoecken, 2003). Most patients with One established risk factor for eating disorders in men is
binge-eating disorder are older than those with anorexia nervosa or homosexuality. Gay and bisexual men have higher rates of eating
bulimia nervosa, generally between 30 and 50 years of age. disorders than heterosexual men do (Feldman & Meyer, 2007).
Eating disorders have long been regarded as occurring pri- Gay men (like heterosexual men) value attractiveness and youth in
marily in women. Although in the past it was thought that the their romantic partners. Because gay men (like women) are seeking
gender ratio was as high as 10:1, more recent estimates suggest to be sexually attractive to men, body dissatisfaction may therefore
that there are three females for every male with an eating disorder be more of an issue for gay men than it is for heterosexual men. In
(Jones & Morgan, 2010). This downward revision of the gender support of this idea, Smith and colleagues (2011) found that gay
ratio reflects the fact that eating disorders in men may have been men were more dissatisfied with their bodies and had higher levels

thinking critically about DSM-5

Other Forms of Eating Disorders DSM-5

I
n addition to anorexia nervosa, bulimia nervosa, and binge DSM-5, the diagnosis of EDNOS was given to around 60 percent
eating disorder, the DSM-5 also includes a new category that of adolescents and adults who sought treatment in outpatient
is used to diagnose other clinically significant problems involv- settings (Eddy, Doyle, et al., 2008; Fairburn & Bohn, 2005).
ing feeding and eating. Purging disorder (which involves purging Changes to the diagnostic criteria for anorexia nervosa and buli-
in normal weight people who have not eaten large amounts of mia nervosa in DSM-5 as well as the inclusion of BED as a new
food) is one example of this. Gone from the DSM is the diagno- diagnosis are expected to make EDNOS a less meaningful and
sis of eating disorder not otherwise specified (EDNOS). Prior to necessary diagnosis. As a result it has been removed.

300  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 300 10/06/13 4:42 PM

 Men are now experiencing socio- disorder and better detection by clinicians. There was also a rise in
cultural pressures to have toned
the number of new cases of bulimia nervosa from 1970 to 1993
and muscular bodies. For many
men, body dissatisfaction takes (Keel & Klump, 2003). However, much of this increase may have
the form of wanting to have a occurred in the time period up to 1982. A more recent analysis has
more muscular upper torso. indicated that the prevalence of bulimia nervosa decreased from
1982 to 1992 and remained stable from 1992 to 2002 (Keel et al.,
2006). Stable rates of bulimia nervosa from 1990 to 2004 have
also been reported (Crowther et al., 2008).
Despite the encouraging news of a decrease in the prevalence
of bulimia nervosa over the last few decades, many young peo-
ple, particularly girls and young women in their adolescence and
early adulthood, show some evidence of disturbed eating pat-
terns or have distorted self-perceptions about their bodies. For
example, in a sample of 4,746 middle and high school students,
41.5 percent of girls and 24.9 percent of boys reported problems
with body image, and more than a third of the girls and almost
a quarter of the boys said they placed a lot of importance on
of disordered eating than heterosexual men did. Moreover, gay weight and shape with regard to their self-esteem (Ackard et al.,
men tended to believe that a potential mate would want them to be 2007). Questionnaire studies further suggest that up to 19 per-
leaner than they themselves wanted to be. Other specific subgroups cent of students report some bulimic symptoms (Hoek, 2002).
of men who are at higher risk of eating disorders are wrestlers and Also of concern are findings from a survey of adults in Australia
jockeys, who need to “make weight” in order to compete or work showing a twofold increase in binge eating, purging, and strict
(Carlat et al., 1997). dieting or fasting in the period from 1995 to 2005 (Hay et al.,
2008). Disordered eating behaviors that do not meet criteria
for an eating disorder diagnosis are of concern because, in some
Prevalence of Eating Disorders cases, they may worsen over time, eventually leading to clinically
The most common form of eating disorder is binge-eating significant problems. Table 9.3 shows some sample items from
disorder. Worldwide, and based on the most recent data, the the Eating Disorders Inventory, a questionnaire measure that is
lifetime prevalence of binge eating disorder is around 2 percent often used in research.
(Kessler et al., 2013). In the United States, community-based
estimates from the National Comorbidity Survey indicate a life- Medical Complications of Eating Disorders
time prevalence of around 3.5 percent in women and 2 percent in
The tragic death of Ana Carolina Reston from kidney failure
men (Hudson et al., 2007). It is also worth noting that the preva-
serves as a sad reminder of just how lethal a disorder anorexia
lence of binge-eating disorder is higher in obese people and in the
nervosa is. In fact, anorexia nervosa has the highest mortality rate
range of 6.5 to 8 percent (Grilo, 2002; Sansone et al., 2008).
of any psychiatric disorder. The mortality rate for females with
Worldwide, the prevalence of bulimia nervosa is estimated
anorexia nervosa is more than 12 times higher than the mortality
at 1 percent (Kessler et al., 2013). Data from the National
rate for females aged 15 to 24 in the general U.S. population
Comorbidity Survey further show that the lifetime prevalence of
(Sullivan, 1995, 2002). When patients with this disorder die, it
bulimia nervosa in the United States is around 1.5 percent for
women and 0.5 percent for men (Hudson et al., 2007). Some-
what less frequent is anorexia nervosa. Estimates from the United
States suggest that this disorder has a lifetime prevalence of 0.9 table9.3
percent in women and 0.3 percent in men (Hudson et al., 2007). Sample Items from the Eating Disorders Inventory
These figures are comparable to prevalence estimates from Swe-
Sample Statement
den, where the rate of anorexia nervosa is 1.2 percent in women
I am terrified of gaining weight.
and 0.29 percent in men (Bulik et al., 2006). Although anorexia
I eat or drink in secrecy.
nervosa is sometimes viewed as a very rare disorder, in its severe
form it is about as common as schizophrenia. I think my hips are too big.

The risk of developing anorexia nervosa seemed to increase Source: Reproduced by special permission of the Publisher, Psychological
during the twentieth century. In one study, lifetime rates of this Assessment Resources, Inc., 16204 North Florida Avenue, Lutz, Florida
33549, from the Eating Disorder Inventory-3 by David M. Garner, PhD,
disorder were higher in people born after 1945 than before this Copyright 1984, 1991, 2004, by Psychological Assessment Resources, Inc.
time (Klump et al., 2007). This is true for both males and females. (PAR). Further reproduction is prohibited without permission of PAR.
This increase is not fully explained by increased awareness of the

eating disorders and obesity CHAPTER 9  301

M09_BUTC4286_16_SE_C09.indd 301 10/06/13 4:42 PM

 is most often because of medical complications. Overall, approxi- in key electrolytes such as potassium (Mitchell & Crow, 2010).
mately 3 percent of people with anorexia nervosa die from the con- Chronically low levels of potassium (hypokalemia) can also result in
sequences of their self-imposed starvation (Signorini et al., 2007). kidney damage and renal failure severe enough to require dialysis.
Malnutrition also takes its toll in other ways (see Mitchell & Abuse of laxatives, which occurs in 10 to 60 percent of
Crow, 2010). Many patients with anorexia nervosa disorder look patients with eating disorders (Roerrig et al., 2010) makes all of
extremely unwell. Their hair on the scalp thins and becomes these problems much worse. Laxatives are used to induce diar-
brittle, as do their nails. Their skin becomes very dry, and downy rhea so that the person feels thinner or to remove unwanted
hair (called “lanugo”) starts to grow on the face, neck, arms, calories from the body. Laxative abuse can lead to dehydration,
back, and legs. Many patients also develop a yellowish tinge to electrolyte imbalances, and kidney disease as well as damage to
their skin, especially on the palms of their hands and bottoms of the bowels and gastrointestinal tract.
their feet. Some of these problems are illustrated in Figure 9.2. Bulimia nervosa is much less lethal than anorexia nervosa,
Because they are so undernourished, people with this disor- although it is still associated with a mortality rate that is approxi-
der have a difficult time coping with cold temperatures. Their mately twice that found in people of comparable age in the gen-
hands and feet are often cold to the touch and have a purplish- eral population (Arcelus et al., 2011). Bulimia nervosa also creates
blue tinge due to problems with temperature regulation and lack a number of medical concerns (Mitchell & Crow, 2010). Purging
of oxygen to the extremities. As a consequence of chronically low can cause electrolyte imbalances and low potassium, which, as we
blood pressure, patients often feel tired, weak, dizzy, and faint. have already mentioned, puts the patient at risk for heart abnor-
Thiamin (vitamin B1) deficiency may also be present; this could malities. Another complication is damage to the heart muscle,
account for some of the depression and cognitive changes docu- which may be due to using ipecac syrup (a poison that causes
mented in low-weight anorexia patients (Winston et al., 2000). vomiting). More typically, however, patients develop calluses on
Although many of these problems resolve when patients gain their hands from sticking their fingers down their throats to make
weight, anorexia nervosa may result in increased risk for osteopo- themselves sick. In extreme cases, where objects such as a tooth-
rosis in later life. This is because peak bone density is normally brush are used to induce vomiting, tears to the throat can occur.
attained during the years of early adulthood. The failure to eat Because the contents of the stomach are acidic, patients dam-
healthily during this time may result in more brittle and fragile age their teeth when they throw up repeatedly. Brushing teeth
bones forever (Attia & Walsh, 2007). immediately after vomiting damages them even more. Mouth
People with anorexia nervosa can die from heart arrhythmias ulcers and dental cavities are a common consequence of repeated
(irregular heartbeats). Sometimes this is caused by major imbalances purging, as are small red dots around the eyes that are caused by the
pressure of throwing up. Finally, patients with buli-
Anorexia affects your whole body mia very often have swollen parotid (salivary) glands
caused by repeatedly vomiting. These are known as
Brain and Nerves Hair “puffy cheeks” or “chipmunk cheeks” by many buli-
can’t think right, fear of hair thins and gets brittle
gaining weight, sad, moody, mia sufferers. Although such swellings are not pain-
irritable, bad memory, fainting, Blood
anemia and other ful, they are often quite noticeable to others.
changes in brain chemistry
blood problems
Heart Muscles and Joints
low blood pressure, weak muscles, swollen joints, Course and Outcome
slow heart rate, fluttering fractures, osteoporosis
of the heart (palpitations), We begin this section with a sobering statistic. After
heart failure Kidneys
kidney stones, kidney failure
medical complications, the second most common
Body Fluids cause of death in those who suffer from anorexia
Hormones
periods stop, bone loss,
low potassium, magnesium, nervosa is suicide. Recent estimates suggest that
and sodium
problems growing, trouble one out of five deaths in individuals with anorexia
getting pregnant. If Intestines
pregnant, higher risk for constipation, bloating
nervosa is the result of suicide (Arcelus et al., 2011).
miscarriage, having Somewhere between 3 and 23 percent of patients
a C-section, baby with Skin
bruise easily, dry skin, growth with anorexia will make a suicide attempt, and rates
low birthweight, and
post partum depression. of fine hair all over body, of completed suicide are 50 times greater than they
get cold easily, yellow skin,
nails get brittle
are in the general population (Franko & Keel, 2006;
Dash line indicates that organ is behind other main organs.
Keel et al., 2003). It has been suggested that patients
who have lost their ability to maintain an “emotion-
figure 9.2 ally protective” low body weight are at particularly
Anorexia nervosa takes its toll on the body and causes many medical problems. When high risk of suicide (Crisp et al., 2006). Patients who
starved, the body turns on itself in an effort to provide energy. Fat is burned first, then
muscle. Eventually, organs are destroyed in the body’s struggle to find fuel. Watch are older when they first receive clinical attention for
the Video Eating Disorders on MyPsychLab their disorder are also more likely to have a premature

302  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 302 10/06/13 4:42 PM

 death (Arcelus et al., 2011). Bulimia nervosa is not associated with with one form of eating disorder to be later diagnosed with
increased risk of completed suicide, although suicide attempts are another eating disorder. Over a 7-year period, Eddy and col-
made in 25 to 30 percent of cases (Franko & Keel, 2006). leagues (2008) report that the majority of women in their
Although the clinical outcome for some patients is tragic, study experienced diagnostic crossover. Bidirectional transitions
over the very long term recovery is possible. Löwe and colleagues between the two subtypes of anorexia nervosa (restricting and
(2001) examined what happened to patients with anorexia ner- binge-purging) were especially common.
vosa 21 years after they had first sought treatment. Reflecting Shifts from anorexia nervosa to bulimia nervosa also
the high morbidity associated with anorexia nervosa, 16 percent occurred in about a third of patients. Interestingly, however,
of the patients (all of whom were women) were no longer alive, there were no cases of direct transition from the restricting type
having died primarily from complications of starvation or from of anorexia nervosa directly into bulimia nervosa. Instead, the
suicide. Another 10 percent were still suffering from the disorder, transition to bulimia nervosa seems to occur after an earlier
and a further 21 percent had partially recovered. On the positive transition to the binge-purging subtype of anorexia nervosa. You
side, however, 51 percent of the individuals were fully recovered may also recall that the main difference between patients with
at the time of the follow-up. In another 6-year follow-up study, the binge-purging subtype of anorexia nervosa and bulimia ner-
52.1 percent of patients who had received outpatient cognitive vosa is weight (and associated amenorrhea). So if someone with
behavioral therapy for their anorexia nervosa eventually recov- anorexia nervosa (binge-purge subtype) gains weight, the diagno-
ered (Castellini et al., 2011). These findings provide grounds for sis will change to bulimia nervosa to reflect this fact, even though
optimism for those who suffer from this disorder. They indicate there may not be a big clinical change in the illness itself. More-
that even after a series of treatment failures it is still possible for over, even after they have crossed over into bulimia nervosa,
women with anorexia nervosa to become well again. these women remain vulnerable to relapsing back into anorexia
With regard to bulimia nervosa, in the long term, progno- nervosa. This suggests that clinicians should pay attention to a
sis tends to be quite good. Two outcome studies have shown past history of anorexia nervosa even when patients no longer
that around 70 percent of women initially diagnosed with this meet the low-weight criterion necessary for its diagnosis.
disorder will be in remission and will no longer meet diagnostic Only a minority of patients with bulimia nervosa transition
criteria for any eating disorder by the end of an 11- to 12-year into anorexia nervosa. In one study the figure was 14 percent (Eddy,
follow-up (Keel et al., 1999; Fichter & Quadflieg, 2007). The Dorer, et al., 2008). In another it was 9.2 percent (Castellini et al.,
remaining 30 percent, however, will continue to experience sig- 2011). Crossovers from the restricting subtype of anorexia nervosa
nificant difficulties with their eating. Substance-abuse problems, into binge-eating disorder do not seem to occur at all. Diagnostic
more frequent binges, more shape concerns, as well as a longer crossover from bulimia nervosa into binge-eating disorder occurs in
duration of illness predict worse outcomes over time (Castellini about 10.9 percent of cases (Castellini et al., 2011).
et al., 2011; Keel et al., 1999). Finally, we note that binge-eating disorder and anorexia
Finally, like patients with bulimia nervosa, patients with nervosa appear to be quite distinct disorders. Over the course of
binge-eating disorder also have high rates of clinical remission. a 12-year follow-up, no patient with binge-eating disorder devel-
Following a period of intensive treatment, two-thirds of a sample oped anorexia nervosa and no patient with anorexia nervosa devel-
of 60 patients no longer had any form of eating disorder oped binge-eating disorder (Fichter & Quadflieg, 2007). In other
(Fichter & Quadflieg, 2007). In a larger study involving 137 Ital- words, there was no diagnostic crossover between these diagnoses.
ian patients who received individual cognitive-behavior therapy, However, around 10 percent of patients who previously had
60 percent were found to be recovered when they were assessed binge-eating disorder transitioned into bulimia nervosa during
6 years after the end of treatment (Castellini et al., 2011). this time. Figure 9.3 on p. 304 illustrates some of these trends.
It is worth noting that, even when well, many individuals
who recover from anorexia nervosa and bulimia nervosa still
harbor residual food issues. They may be excessively concerned Association of Eating Disorders With Other
about shape and weight, restrict their dietary intakes, and overeat Forms of Psychopathology
and purge in response to negative mood states (Sullivan, 2002). Eating disorder diagnoses are commonly associated with other
In other words, the idea of recovery is relative. Someone who no diagnosable psychiatric conditions. In fact, comorbidity is the rule
longer meets all of the diagnostic criteria for an eating disorder rather than the exception (Hudson et al., 2007). For instance,
may still have issues with food and body image. approximately 68 percent of patients with anorexia nervosa,
63 percent of patients with bulimia nervosa, and almost 50 per-
cent of people with binge-eating disorder are also diagnosed with
Diagnostic Crossover depression (Brewerton et al., 1995; Halmi, et al., 1991; Hudson
One way in which eating disorders differ from other types of et al., 2007; O’Brien & Vincent, 2003). Obsessive-compulsive
disorders is that there is a lot of diagnostic crossover. What this disorder is often found in patients with anorexia nervosa and
means is that it is quite common for someone who is diagnosed bulimia nervosa (Kaye et al., 2004; Milos et al., 2002;

eating disorders and obesity CHAPTER 9  303

M09_BUTC4286_16_SE_C09.indd 303 10/06/13 4:42 PM

 and obsessionality (Keys et al., 1950). We must therefore be cau-
tious in our conclusions.
Even though the physiological consequences of eating dis-
AN–R AN–BP orders may exacerbate personality disturbances, they may only
be enhancing traits that were present prior to the development
of the illness. Research suggests that some personality traits in
eating-disordered patients might both predate the onset of the
disorder and remain even when the eating disorder remits and
the patient has recovered (Kaye et al., 2004; Klump et al., 2004).
BED BN Consistent with this, around two-thirds of a sample of patients
with anorexia nervosa reported that they were rigid and perfec-
tionistic, even as children (Anderluh et al., 2003).

Eating Disorders Across Cultures

Although the majority of research on eating disorders is conducted
AN–R = Anorexia nervosa–restricting subtype in the United States and Europe, eating disorders are not confined
AN–BP = Anorexia nervosa–binge-purge subtype to these areas. Widespread eating disorder difficulties have been
BN = Bulimia nervosa reported among both Caucasian and black South African college
BED = Binge eating disorder
students (le Grange et al., 1998). Anorexia nervosa and bulimia
figure 9.3 nervosa are also clinical problems in Japan, Hong Kong, Taiwan,
Diagnostic Crossover in Eating Disorders. Diagnostic crossover is Singapore, and Korea (Lee & Katzman, 2002).
­common in eating disorders with transitions between the two subtypes Cases of eating disorders have also been documented in
of anorexia nervosa being very common. Transitions from the binge-
purge subtype of anorexia nervosa to bulimia nervosa also often occur.
India and Africa. The prevalence of eating disorders in Iran is
In the figure, the width of the arrow denotes the relative likelihood of comparable to that in the United States (Nobakht & Dezhkam,
any ­transition. (Adapted from Fichter & Quadflieg, 2007, and Eddy, 2000). And a few years ago, the first published report of five
Dorer, et al., 2008.)
men in central China who were diagnosed with eating disor-
ders appeared (Tong et al., 2005). Far from being confined to
O’Brien & Vincent, 2003). In addition, there is frequent co- industrialized Western countries, eating disorders are becoming a
occurrence of substance-abuse disorders in the binge-purging problem worldwide.
subtype of anorexia nervosa as well as in bulimia nervosa. The Being Caucasian, however, does appear to be associated with
restrictive type of anorexia nervosa, however, tends not to be asso- subclinical problems that may place individuals at higher risk for
ciated with higher rates of substance abuse (Halmi, 2010). developing eating disorders. Examples of such problems include
Comorbid personality disorders are frequently diagnosed in body dissatisfaction, dietary restraint, and a drive for thinness.
people with eating disorders (Rø et al., 2005; see Chapter 10). A meta-analysis involving a total of 17,781 participants has
Those with the restrictive type of anorexia nervosa are inclined shown that such attitudes and behaviors are significantly more
toward personality disorders in the anxious-avoidant cluster prevalent in whites than in nonwhites (Wildes et al., 2001).
(Cluster C; Skodol et al., 1993). In contrast, eating disorders Although Asian women exhibit levels of pathological eating
that involve binge/purge syndromes (both anorexia nervosa and similar to those of white women (Wildes et al., 2001), it has long
bulimia nervosa) are more likely to be associated with dramatic,
emotional, or erratic (Cluster B) problems, especially borderline
research Close-Up
personality disorder (Halmi, 2010).
Consistent with this, more than a third of patients with eating
disorders have engaged in the kinds of self-harming behaviors (cut-
Meta-analysis
ting or burning themselves, for example) that are symptomatic of
borderline personality disorder (Paul et al., 2002). Personality disor- A meta-analysis is a statistical method used to combine the
ders are similarly reported in patients with BED, although no clear results of a number of similar research studies. The data from
pattern has emerged (Wilfley et al., 2000). People with BED also each separate study are transformed into a common metric
have high rates of anxiety disorders (65%), mood disorders (46%), called the effect size. Doing this allows data from the various
and substance use disorders (23%; see Hudson et al., 2007). studies to be combined and then analyzed. You can think of
a meta-analysis as being just like the research with which you
One problem with simple examinations of personality disor-
are already familiar, except that the “participants” are indi-
ders in patients with eating disorders is that some of the distur-
vidual research studies, not individual people!
bances found in these patients could reflect the consequences of
malnourishment. Starvation is known to increase both irritability

304  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 304 10/06/13 4:42 PM

 been held that African Americans are less susceptible to subclini- Britain but who had South Asian (Indian, Pakistani,
cal types of eating problems and body image concerns than Bangladeshi) ethnic origins also were less likely than patients
Caucasians are. For example, in one sample of 1,061 black with English ethnic origins to show evidence of fat phobia
women, no case of anorexia nervosa was found. In contrast, out (Tareen et al., 2005).
of a sample of 985 white women, 15 (1.5%) met clinical criteria In yet another study, young women in Ghana who had
for this disorder. Fewer black than white women also had buli- anorexia nervosa were also not especially concerned about their
mia nervosa (0.4% versus 2.3%; see Striegel-Moore et al., 2003). weight or shape. Rather, they emphasized religious ideas of
However, as minorities become more and more integrated self-control and denial of hunger as the motivation for their
and internalize white, middle-class societal values about the self-starvation (Bennett et al., 2004). In a final example, Japa-
desirability of thinness, we should expect to see increases in the nese women with eating disorders reported significantly lower
rates of eating disorders. As an example of this, Alegria and col- levels of perfectionism and less of a drive for thinness than did
leagues (2007) have demonstrated that rates of eating disorders American women with eating disorders (Pike & Mizushima,
were higher in Latinos who were born in the United States com- 2005). Findings such as these highlight the considerable role
pared with those who were not. In The World Around Us box played by culture in the clinical presentation of eating disorders.
below we discuss the protective role of ethnic identity in African Cases of anorexia nervosa have been reported throughout
American women. history. They have furthermore been shown to occur all over
A select number of the clinical features of diagnosed the world. In light of this reality, Keel and Klump (2003)
forms of eating disorders may also vary according to culture. have concluded that anorexia nervosa is not a culture-bound
For instance, about 58 percent of anorexia nervosa patients in syndrome. Of course, as we have just noted, culture may influ-
Hong Kong are not excessively concerned about fatness. The ence the disorder’s clinical manifestation. The more important
reason they give for refusing food is fear of stomach bloating point, however, is that anorexia nervosa is not a disorder that
(Lee et al., 1993). Anorexia nervosa patients who were living in occurs simply because of exposure to Western ideals and the

the WORLD around us

Ethnic Identity and Disordered Eating

I
n contrast to young white, Asian American, and Hispanic This highlights the importance of healthy role models for young
girls, black adolescent girls seem less inclined to use weight black girls. Without these, black women will no longer remain
and appearance to fuel their sense of identity and self-worth unique among American ethnic groups in their lower levels of
(Grabe & Hyde, 2006; Polivy et al., 2005). This may provide body dissatisfaction.
them with some protection from the development of eating
disorders. However, any protection afforded to black women
seems to be linked to how strongly they identify with their
ethnic group and how much they receive culturally consistent
messages that value what their bodies naturally look like and
support who they are.

In a study of 322 African American female college students,

Rogers and colleagues (2010) found that women who had stron-
ger ethnic identities were less likely to have internalized U.S.
societal beauty ideals about the importance of being thin and
attractive. This is an important finding because, in the same
study, internalization of these ideals was found to be linked to
disordered eating. An important factor in the ethnic distribution
and occurrence of eating disorders is therefore the extent to
which minority women are both exposed to and willing to inter-
nalize the same kind of white, middle-class values that have been
African American women who have a strong identification with their
linked to increased risk for eating disorders. There is already some ethnicity and cultural heritage have less risk of developing eating
evidence that the body dissatisfaction difference between blacks disorders than African American women who have internalized white
and whites is getting smaller over time (Grabe & Hyde, 2006). middle-class values about the desirability of thinness.

eating disorders and obesity CHAPTER 9  305

M09_BUTC4286_16_SE_C09.indd 305 10/06/13 4:42 PM

 modern emphasis on thinness. In contrast, bulimia nervosa anorexia nervosa and bulimia nervosa themselves. In one large
does seem to be a culture-bound syndrome. More specifi- family study of eating disorders, the risk of anorexia nervosa for
cally, it seems to occur in people who have had some exposure the relatives of individuals with anorexia nervosa was 11.4 times
to Western ideals about thinness, who have access to large greater than for the relatives of the healthy controls (Strober
amounts of food, and who, because of modern plumbing, can et al., 2000).
purge in private (Keel, 2010). For the relatives of people with bulimia nervosa, the risk of
bulimia nervosa was 3.7 times higher than it was for the relatives
in review of the healthy controls (Strober et al., 2000). It is of additional
• What are the major clinical differences between patients with interest that the relatives of patients with eating disorders are
anorexia nervosa and patients with bulimia nervosa? What clini- more likely to suffer from a variety of other disorders as well
cal features do these two forms of eating disorders have in as eating disorders. For example, high rates of major depres-
common? sive disorder are found in the relatives of patients with anorexia
• How do the prevalence rates for eating disorders vary according nervosa, bulimia nervosa, and binge-eating disorder (Lilenfeld
to gender, sexual orientation, and ethnicity? et al., 1998; Lilenfeld et al., 2006; Mangweth et al., 2003). The
• What kinds of medical problems do patients with eating disor- relatives of people suffering from bulimia nervosa also have an
ders suffer from? increased likelihood of having problems with alcohol and drug
• What is the long-term outcome for patients with anorexia dependence, while the relatives of those diagnosed with anorexia
nervosa? In what ways is the clinical outcome of patients with nervosa are at increased risk of obsessive-compulsive disorder and
anorexia nervosa different from the clinical outcome of patients obsessive-compulsive personality disorder (Lilenfeld et al., 1998).
with bulimia nervosa? As you know, family studies do not allow for the untangling
• Why is bulimia nervosa viewed as a culture-bound syndrome of the different contributions of genetic and environmental
when anorexia nervosa is not? influences. These kinds of questions are best resolved by twin
and adoption studies. Presently, we have none of the latter. For-
tunately, a number of twin studies do exist. Considered together,
Risk and Causal Factors these studies suggest that both anorexia nervosa and bulimia ner-
in Eating Disorders vosa are heritable disorders (Fairburn & Harrison, 2003; Wade,
2010). Indeed, it has been suggested that the contribution of
There is no single cause of eating disorders. In all probability, genetic factors to the development of eating disorders may be
they reflect the complex interaction between genetic and envi- about as strong as the contribution of genetic factors to bipolar
ronmental factors. Biological, sociocultural, family, and individ- disorder and schizophrenia (Kaye, 2008).
ual variables likely all play a role. However, it is important not to There is provocative evidence for a gene (or genes) on chro-
regard these areas as distinct and in competition with each other. mosome 1 that might be linked to susceptibility to the restrictive
The question of whether eating disorders are caused by biologi- type of anorexia nervosa (Grice et al., 2002). However, a recent
cal factors or by cultural pressures is not an appropriate one. study failed to find any links between 182 different candidate
Biological and cultural explanations are interlinked. For example, genes and various aspects of eating disorders such as lowest body
being exposed to cultural attitudes that emphasize thinness is mass index, drive for thinness, or body dissatisfaction (Root et al.,
significant in the development of eating disorders. However, 2011). These disappointing findings highlight the challenges asso-
only a small number of people who are exposed to such attitudes ciated with trying to identify the genetic basis of anorexia nervosa.
actually develop eating disorders. Genetic factors may give rise Evidence has also suggested that susceptibility to bulimia nervosa,
to individual differences that help explain why certain people particularly self-induced vomiting, may be linked to chromosome
are more sensitive than others to cultural attitudes and other 10 (Bulik, Devlin, et al., 2003). Of course, until these findings are
environmental risk factors. In other words, eating disorders are widely replicated, they must be viewed as preliminary.
best conceptualized in terms of the diathesis-stress model, where Eating disorders have additionally been linked to genes that
genes render some people more susceptible to environmental are involved in the regulation of the neurotransmitter serotonin
pressures and hence to the development of problematic eating (Wade, 2010). This makes sense, given the role that serotonin is
attitudes and behaviors. known to play in the regulation of eating behavior. Furthermore,
serotonin is known to be involved in mood. With this in mind,
Biological Factors it is interesting to note that mood disorders and eating disorders
Genetics Much research attention is now being devoted to often cluster together in families (Halmi, 2010).
the study of genetic factors in eating disorders. This is because At present, researchers are still some distance away from
the tendency to develop an eating disorder has been shown to understanding the precise role genes play in the development
run in families (Wade, 2010). The biological relatives of people of eating disorders. Given the high degree of overlap between
with anorexia nervosa or bulimia nervosa have elevated rates of anorexia nervosa and bulimia nervosa (many women with

306  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 306 10/06/13 4:42 PM

 anorexia nervosa later develop bulimia nervosa and many women One important kind of physiologic opposition designed
with bulimia nervosa report histories of anorexia nervosa), it to prevent us from moving far from our set point is hunger. As
is likely that these two eating disorders may have some genetic we lose more and more weight, hunger may rise to extreme lev-
factors in common (Helder & Collier, 2010). There is also rea- els, encouraging eating, weight gain, and a return to a state of
son to believe that genetic factors may play more of a role after equilibrium. Far from having little or no appetite, patients with
puberty than before it (Klump et al., 2007). In other words, the anorexia nervosa may think about food constantly and make
magnitude of genetic influences may increase over the course of intense efforts to suppress their increasing hunger. Accordingly,
development. This could be because genetic factors may deter- chronic dieting may well enhance the likelihood that a person
mine how different individuals respond to the psychosocial risk will encounter periods of seemingly irresistible impulses to gorge
factors (pressure to be thin, family conflict) that may develop or on large amounts of high-calorie food. For patients with bulimia
become more important during the adolescent years. nervosa, these hunger-driven impulses may escalate into uncon-
trollable binge eating.
Brain Abnormalities An area of the brain that plays an im-
portant role in eating is the hypothalamus. Animal studies have Serotonin Serotonin is a neurotransmitter that has been
demonstrated that when the lateral hypothalamus (the side areas implicated in obsessionality, mood disorders, and impulsivity.
of the hypothalamus) is stimulated electrically the animal will It also modulates appetite and feeding behavior. Because many
start to eat, even if it has eaten very recently. During stimulation patients with eating disorders respond well to treatment with
of the lateral hypothalamus, animals will also tolerate shock and antidepressants (which target serotonin), some researchers have
other unpleasant stimuli in order to gain access to food. concluded that eating disorders involve a disruption in the sero-
There is no good evidence that abnormalities in the hypo- tonergic system (Bailer & Kaye, 2011).
thalamus play a central role in eating disorders, however. Uher Serotonin is made from an essential amino acid called tryp-
and Treasure (2005) review a series of case reports of patients tophan. This can only be obtained from food. After tryptophan
with tumors in the hypothalamus. Although these were some- is consumed, it is converted to serotonin via a series of chemi-
times associated with an increase or loss of appetite, there was no cal reactions. People with anorexia nervosa have low levels of
evidence that they resulted in specific eating disorders. In con- 5-HIAA, which is a major metabolite of serotonin. This may
trast, damage to the frontal and the temporal cortex did seem to be because they are eating so little food. In contrast, levels of
be linked to the development of anorexia nervosa in some cases 5-HIAA are normal in people with bulimia nervosa. What is
and bulimia nervosa in others. This is interesting because the interesting is that, after recovery, both of these patient groups
temporal cortex is known to be involved in body image percep- have higher levels of 5-HIAA than control women do; they also
tion. Parts of the frontal cortex (particularly an area called the have higher levels of 5-HIAA than they had when they were in
orbitofrontal cortex) also play a role in monitoring the pleasant- the ill state (Kaye, 2008). Although the finding of higher levels
ness of stimuli such a smell and taste (van Kuyck et al., 2009). of 5-HIAA in recovered patients compared to controls seems
Although very speculative at this time, it is possible that the counterintuitive, it has been suggested that resuming normal eat-
lateral hypothalamus acts as a site that integrates information ing makes it possible to detect abnormalities in the serotonin sys-
relevant for regulating food intake. The lateral hypothalamus tem (such as higher levels of serotonin in several different brain
receives information from many parts of the brain, including areas) that might be involved in risk for eating disorders. Kaye
the frontal cortex and the amygdala (which is a part of the brain and colleagues have further suggested that people with serotonin
involved in emotion and fear learning). Animal research suggests overactivity may use dieting as a way to regulate this by decreas-
that a network involving these (and other) brain areas may be ing the amount of tryptophan that is available to make serotonin
important not only for overeating in response to environmental (Bailer & Kaye, 2010).
cues but for suppressing eating in response to fear (Petrovich, Of course, it is important to remember that neurotransmit-
2011). As research progresses we will learn more about how the ters like serotonin do not work in isolation. A change in the sero-
pieces of the puzzle fit together to result in different types of eat- tonin system will have implications for other neurotransmitter
ing disorders. systems too (e.g., dopamine, norepinephrine). So the situation
is undoubtedly complex. Nonetheless, investigating the role of
Set Points There is a well-established tendency for our bod- serotonin in eating disorders is still an active area of research.
ies to resist marked variation from some sort of biologically
determined set point or weight that our individual bodies try to Sociocultural Factors
“defend” (Garner, 1997). Anyone intent on achieving and main- What is the ideal body shape for women in Western culture?
taining a significant decrease in body mass below his or her in- Next time you glance at a glossy fashion magazine, take a
dividual set point may be trying to do this in the face of internal moment to consider the messages contained in its pages. The
physiologic opposition, which is aimed at trying to get the body overall body size of the models that appear on the covers of such
back close to its original set-point weight. magazines as Vogue and Cosmopolitan has become increasingly

eating disorders and obesity CHAPTER 9  307

M09_BUTC4286_16_SE_C09.indd 307 10/06/13 4:42 PM

 thinner over the years (Sypeck et al., 2004). Young women are British supermodel Kate Moss
has said that one of her mottos
avid consumers of such magazines and are bombarded with
is “Nothing tastes as good as
images of unrealistically thin models. These magazines are also skinny feels.”
widely available all over the world. British Vogue is published in
40 or more countries and can be found in India, Argentina, and
Kenya among other places (see Gordon, 2000). Moreover, social
pressures toward thinness may be particularly powerful in higher
socioeconomic backgrounds, from which a majority of girls and
women with anorexia nervosa appear to come (McClelland &
Crisp, 2001).
It is likely that thinness became deeply rooted as a cultural
ideal in the 1960s, although prior to this time women had
certainly been concerned with their weight and appearance.
However, the type of body that was regarded as glamorous
and attractive (e.g., Marilyn Monroe) was more curvaceous.
One landmark event was the arrival of Twiggy on the fashion
scene. Twiggy was the first superthin supermodel. Although
her appearance was initially regarded as shocking, it did not with qualities that were highly valued such as being strong, able
take long for the fashion industry to embrace the look she to work, and kind and generous. Being thin, in contrast, was
exemplified. regarded negatively because it was thought to reflect being sickly,
The emphasis on thinness in the fashion industry continues. incompetent, or having somehow received poor treatment.
Although, as we noted earlier, some efforts are being made to Culturally, fatness was preferred over thinness, and dieting was
exclude excessively thin models from the runways, these changes viewed as offensive. What was also striking was the total absence
are still being resisted by many designers. In an example of of any condition that could be considered an eating disorder.
the glamorization of anorexia, British supermodel Kate Moss After television came to Fiji, however, the cultural climate
has even made up a new word to describe herself. The word is changed. Not only were Fijians able to see programs such as
“rexy,” a hybrid term that combines “anorexic” and “sexy.” Beverly Hills 90210 and Melrose Place that were popular at that
A provocative illustration of the importance of the media time, but many young women also began to express concerns
in creating pressures to be thin comes from a now classic study about their weight and dislike of their bodies. For the first
that was done by Anne Becker and her colleagues (2002). time, women in Fiji started to diet in earnest. The young Fijian
When Becker first began conducting research in Fiji in the women studied by Becker also made comments that suggested
1990s, she was struck by the considerable percentage of Fijians that their body dissatisfaction and wish to lose weight were
(especially women) who were overweight with respect to their motivated by a desire to emulate the actors they had seen on
Western counterparts. From a cultural perspective, however, television.
this made sense. Within Fijian culture, being fat was associated Although Becker did not collect information about eating
disorders themselves (she measured attitudes toward eating), this
British model Twiggy was “natural experiment” provides us with some anecdotal information
the first superthin model as to the way in which Western values about thinness may begin
to achieve international
fame.
to infest themselves into foreign cultural environments. In another
reflection of this, after 51 years of Nigeria being unsuccessful in
the Miss World beauty pageant, a 19-year-old Nigerian contestant
finally won the coveted crown in 2001. Her success was later attrib-
uted to the fact that, for the first time, Nigerians had selected a
contestant to represent them who was not considered to be beauti-
ful by local standards on account of being too thin (Onishi, 2002).

Family Influences
Clinicians have long been aware that certain problems seem to
regularly characterize the families of patients with anorexia ner-
vosa, prompting many clinicians to advocate a family therapy
approach to treatment intervention (Lock et al., 2001). Echoing
this sentiment, more than one-third of patients with anorexia

308  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 308 10/06/13 4:42 PM

 In her home country of individual with an eating disorder in the family is likely to affect
Nigeria, Agbani Darego
family functioning in a negative way. That is, the causal con-
was considered too thin
to be beautiful. Nonethe- nection, if any, might be in the other direction. In fact, recent
less, she was selected to longitudinal data from a twin study suggest that disordered eat-
represent the country in ing attitudes may predate parent–child conflict (Spanos et al.,
the 2001 Miss World con-
test and won the coveted 2010). If this finding is replicated in future studies, we may need
crown. to reconsider the role of family conflict as a risk factor for the
development of eating problems.

Individual Risk Factors

Not everyone who lives in a society that places excessive empha-
sis on being thin goes on to develop an eating disorder. If that
were the case, eating disorders would be much more prevalent
than they are. There must be other factors that increase a given
person’s susceptibility to the development of disordered eating.
As we noted earlier, some of these factors may be biological,
while others may be more psychological in nature. In a further
reflection of gene–environment interaction, genetic factors may
actually influence some of the traits (e.g., perfectionism, obses-
siveness, anxiety) that may make some people more likely to
respond to cultural pressures with disturbed eating patterns.

Gender As you have already learned, eating disorders are

much more frequently found in women than in men. Being
female is a strong risk factor for developing eating disorders,
particularly anorexia nervosa and bulimia nervosa (Jacobi et al.,
2004). Moreover, the greatest period of risk for these disorders
nervosa reported that family dysfunction was a factor that con- occurs in adolescence. Binge-eating disorder does not follow this
tributed to the development of their eating disorder (Tozzi et al., pattern, however. The onset of binge-eating disorder is typically
2003). Patients with anorexia nervosa perceive their families as well after adolescence. Binge-eating disorder is also much more
more rigid, less cohesive, and as having poorer communication likely to be found in males as well as in females.
than healthy control participants do (Vidovic et al., 2005). For men (but not for women) sexual orientation is a risk
In addition, many of the parents of patients with eating factor for disordered eating. As we noted earlier, this may be
disorders have long-standing preoccupations regarding the because gay and bisexual men are trying to be attractive to
desirability of thinness, dieting, and good physical appearance men, who (regardless of sexual orientation) typically place great
(Garner & Garfinkel, 1997). Like their children, they have per- emphasis on physical appearance. However, being in a relation-
fectionistic tendencies (Woodside et al., 2002). ship may moderate the risk for disordered eating. Brown and
Family factors have also been studied in connection with Keel (2012a) have recently shown that gay and bisexual men
bulimia nervosa. Both white and ethnic minority (black, Hispanic, in their 30s and 40s who were in relationships reported less of
and biracial) adolescents with bulimia nervosa perceive their fami- a drive for thinness and dieted less frequently than men of the
lies to be less cohesive than their parents do (Hoste et al., 2007). same age who were single and without a partner.
Fairburn and colleagues (1997) have also noted that women with
bulimia nervosa could be differentiated from both a general psy- Internalizing the Thin Ideal The Duchess of Windsor
chiatric control group and a healthy control group on such risk once said that you could never be too rich or too thin. Clearly
factors as high parental expectations, other family members’ diet- she had internalized the thin ideal, buying into the notion that
ing, and degree of critical comments from other family members being thin is highly desirable. Think for a moment about the ex-
about shape, weight, or eating. In a large sample of college-age tent to which you think this way. Do you regard thin people as
women, the strongest predictor of bulimic symptoms was the unhealthy and weak? Or do you associate being thin with feeling
extent to which family members made disparaging comments attractive, being popular, and being happy? The extent to which
about the woman’s appearance and focused on her need to diet people internalize the thin ideal is associated with a range of
(Crowther et al., 2002). problems that are thought to be risk factors for eating disorders.
Still, when attempting to depict family characteristics associ- These include body dissatisfaction, dieting, and negative affect
ated with eating disorders, we must remember that having an (Stice, 2002). In fact, there is empirical evidence to suggest that

eating disorders and obesity CHAPTER 9  309

M09_BUTC4286_16_SE_C09.indd 309 10/06/13 4:42 PM

 internalizing the thin ideal may be an early component of the Body dissatisfaction is a risk
factor for disordered eating.
causal chain that culminates in disordered eating (McKnight In-
vestigators, 2003; Stice, 2001).

Perfectionism Perfectionism (defined as the pursuit of

unattainably high standards combined with an intolerance of
mistakes) has long been regarded as an important risk factor for
eating disorders (Bruch, 1973). This is because people who are
perfectionistic may be much more likely to subscribe to the thin
ideal and relentlessly pursue the “perfect body.” It has also been
suggested that perfectionism helps maintain bulimic pathol-
ogy through the rigid adherence to dieting that then drives the
binge/purge cycle (Fairburn et al., 1997).
Research supports the association of perfectionism and eat-
ing disorders (Bardone-Cone et al., 2007). This is especially
true for anorexia nervosa. In a study of 322 women, Halmi and
colleagues (2000) found that women with anorexia nervosa
scored higher on a measure of perfectionism than did a sample of world) do not display such body image distortions (Platte et al.,
controls without an eating disorder. The women with anorexia 2000). This supports the notion that sociocultural influences are
nervosa scored higher on perfectionism regardless of whether implicated in the discrepancy between the way many young girls
they had the restricting subtype of anorexia nervosa or subtypes and women perceive their own bodies and the “ideal” female
that involved either purging or binge eating and purging. A form as represented in the media. Such perceptual biases lead
large proportion of bulimia nervosa patients also demonstrate a girls and women to believe that men prefer more slender shapes
long-standing pattern of excessive perfectionism (Anderluh et al., than they in fact do. Many women also feel evaluated by other
2003; Garner & Garfinkel, 1997). women, believing that their female peers have even more strin-
Of course, any personality characteristics found in eating- gent standards of weight and shape than they do themselves.
disordered patients could be the result of the eating disorder It would be one thing if women had a reasonable chance of
itself rather than contributory in a causal sense. But if this were attaining their “ideal” bodies simply by not exceeding an average
the case, recovery from anorexia nervosa would be followed by caloric intake or by maintaining a healthy weight. Quite simply,
a reduction in perfectionism. Yet this does not seem to happen this is not possible for most people. In fact, as pointed out by
(Bardone-Cone et al. 2007). Perfectionism also seems to predate Garner (1997), the average body weight of young American women
disordered eating in those with anorexia nervosa (Halmi et al., has been increasing over the past four decades at least, probably as
2012) again suggesting that perfectionism is not a simple corre- a consequence of general improvements in nutrition, pediatric
late of eating problems. Overall, the research findings support the health care, and other factors as well (e.g., the widespread availabil-
idea that perfectionism is an enduring personality trait that places ity of high-calorie foods). Yet, as women’s average weight has been
people at higher risk for the development of eating disorders increasing since the late 1950s, the weight of such cultural icons of
(Lilenfeld et al., 2006; Stice, 2002). Perfectionism may also have attractiveness as Playboy centerfolds and Miss America contestants
a genetic basis. In a twin study, high levels of perfectionism were has decreased at a roughly comparable rate. It has been calculated
found in the twins with anorexia nervosa, as expected. However, that 70 percent of Playboy centerfolds have a body mass index (see
high levels of perfectionism were also found in the co-twins, who Table 9.4) below 18.5 (Katzmarzyk & Davis, 2001).
did not suffer from eating disorders (Wade et al., 2008).
Interestingly, men with eating disorders are less perfectionistic
than are women with eating disorders (Woodside et al., 2004). If
table 9.4
men are generally less perfectionistic than women, this might help
Calculating Body Mass Index
them avoid having some of the weight and shape concerns that
seem to be a stepping stone to the development of eating disorders. weight (lb.)
* 703 = BMI
height (in.)2
Negative Body Image One consequence of sociocultural BMI
pressure to be thin is that some young girls and women develop Healthy 18.5–24.9
highly intrusive and pervasive perceptual biases regarding how Overweight 25–29.9
“fat” they are (e.g., Fallon & Rozin, 1985; Rodin, 1993; Obese 30–39.9
Wiseman et al., 1992; Zellner et al., 1989). In sharp contrast, Morbidly obese 40
young Amish people (who live radically separated from the modern

310  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 310 10/06/13 4:42 PM

 Even children’s toys promote unrealistically slender ideals. As Stice (2002) notes in his comprehensive review of this
Consider, for instance, the size and shape of the Barbie doll. For topic, when our efforts to diet fall short, it is almost inevitable
an average woman to achieve Barbie’s proportions she would that we will feel bad about ourselves (see also Ackard et al.,
have to be 7 feet 2 inches tall, lose 10 inches from her waist, and 2002). Dieting itself may therefore not be the real problem
add 12 inches to her bust (Moser, 1989). (which is good news for people who are obese and who need
The research literature strongly implicates body dissat- to lose weight). Rather, it may be the case that the people who
isfaction as an important risk factor for pathological eating often report that they are trying to lose weight are the people
(McKnight Investigators, 2003). Indeed, in a recent prospective who are most unhappy with their bodies and who are most
longitudinal study, body dissatisfaction emerged as the most inclined to feel negatively about themselves when they fail to
powerful predictor of the onset of eating disorders in a sample of stick to their diet plans. Other factors, such as perfectionism,
almost 500 adolescent girls (Stice et al., 2011). Body dissatisfac- may also play a role. Those who have the highest expectations
tion is also associated with dieting and with negative affect. Point- of themselves may be the people who are most likely to feel bad
edly, if we don’t like how we look, we are likely to feel bad about when they fail to meet their own self-imposed high standards.
ourselves. We may also try to lose weight in order to look better.
Negative Emotionality Negative affect (feeling bad) is
Dieting When people wish to be thinner, they typically go on a a causal risk factor for body dissatisfaction (Stice, 2002). When
diet. Nearly all instances of eating disorders begin with the “normal” we feel bad, we tend to become very self-critical. We may focus
dieting that is routine in our culture. Have you ever dieted? Most on our limitations and shortcomings while magnifying our flaws
people have, at some point in their lives (Jeffrey et al., 1991). Esti- and defects. This seems to be especially true of individuals with
mates are that, at any one time, approximately 39 percent of women eating disorders, who, like people with depression, tend to show
and 21 percent of men are trying to lose weight (Hill, 2002). distorted ways of thinking and of processing information
Dieting is a risk factor for the development or worsening received from the environment (e.g., Butow et al., 1993; Garner
of eating disorders (Jacobi et al., 2004; Striegel-Moore & Bulik, et al., 1997). In many cases, there is widespread negative self-
2007). In a large sample of adolescent girls, body dissatisfaction evaluation (e.g., Fairburn et al., 1997). These cognitive distor-
and dieting predicted symptoms of bulimia nervosa at a 1-year tions (“I’m fat; I’m a failure; I’m useless”) have the potential to
follow-up (Johnson & Wardle, 2005). In another large-scale lon- make people feel even worse about themselves.
gitudinal study, it was found that the majority of adolescent girls Longitudinal studies have confirmed that depression and
who went on to develop anorexia nervosa had been dieters (Patton general negative affect are predictive of a high risk for later develop-
et al., 1990). ing an eating disorder (Johnson, Cohen, et al., 2002; Leon et al.,
As we all know, however, not everyone who diets develops 1997). Moreover, evidence suggests that negative affect may work
an eating disorder. And in some cases diets can be helpful. For to maintain binge eating (see Stice, 2002). Patients often report
example, when overweight women were randomly assigned to that they engage in binges when they feel stressed, down, or bad
either a low-calorie diet or a waiting list control group (that about themselves. They further indicate that in the very short term,
did not involve a diet), those who received the diet lost weight eating provides much needed comfort. These reports are highly
and showed a decrease in bulimic symptoms (Presnell & Stice, consistent with affect-regulation models that view binge eating as
2003). So why has dieting been linked to eating disorders? a distraction from negative feelings (see Vögele & Gibson, 2010).
Of course, a major problem is that after binges patients feel disap-
Dieting is a risk factor for the pointed or even disgusted with themselves. In short, a difficult situ-
development or worsening of ation leads to behavior that makes circumstances even worse.
eating disorders

Childhood Sexual Abuse Childhood sexual abuse has been

implicated in the development of eating disorders (Jacobi et al.,
2004). However, there is some debate about whether sexual abuse
is truly a risk factor for eating disorders (see Stice, 2002). In one
prospective study, Vogeltanz-Holm and colleagues (2000) failed
to find that early sexual abuse predicted the later onset of binge
­eating. However, another prospective study found that children
who had been sexually abused or physically neglected had higher
rates of eating disorders and eating problems in adolescence and
adulthood (Johnson, Cohen, et al., 2002). A meta-analysis of
53 studies also revealed a weak—but positive—association
­between childhood sexual abuse and eating pathology (Smolak &
Murnen, 2002).

eating disorders and obesity CHAPTER 9  311

M09_BUTC4286_16_SE_C09.indd 311 10/06/13 4:43 PM

 This suggests that the two variables are linked in some way, anorexia nervosa are especially likely to terminate inpatient treat-
although the precise nature of the link is not yet clear. One ment prematurely (Steinhausen, 2002; Woodside et al., 2004).
possibility is that being sexually abused increases the risk of Making the situation even worse, there have been surprisingly
developing other known risk factors for eating disorders, such as few controlled studies on which to base an informed judgment
having a negative body image or high levels of negative affect. In about which treatment modality will work best (le Grange &
other words, the causal pathway from early abuse to later eating Lock, 2005; Wilson, Grilo, & Vitousek, 2007). In part, this is
disorder may be an indirect one (rather than a direct one) that due to the fact that the disorder is rare. However, patients who
involves an array of other intervening variables. suffer from anorexia nervosa are also often extremely reluctant to
seek treatment. And when they do, taking part in research stud-
in review ies is unlikely to be a priority for them. These factors combine to
make treatment research extremely difficult.
• What evidence suggests that genetic factors may play a role in
the development of eating disorders? The most immediate concern with patients who have
anorexia nervosa is to restore their weight to a level that is no
• What brain areas are implicated in eating and eating disorders?
longer life threatening. In severe cases, this requires hospitaliza-
• What neurotransmitter has been implicated (and most well stud- tion and extreme measures such as intravenous feeding. This is
ied) in eating disorders?
followed by rigorous control of the patient’s caloric intake so as to
• What individual characteristics are associated with increased risk progress toward a targeted range of weight gain (Andersen et al.,
for eating disorders?
1997). Normally, this short-term effort is successful. However,
• How might a diathesis-stress model be used to explain the without treatment designed to address the psychological issues
development of eating disorders?
that fuel the anorexic behavior, any weight gain will be temporary
and the patient will soon need medical attention again. In some
Treatment of Eating Disorders cases aggressive treatment efforts can backfire (Strober, 2004).

Patients with eating disorders are often very conflicted about Medications Antidepressants are sometimes used in the treat-
getting well. Approximately 17 percent of patients with severe ment of anorexia nervosa, although there is no evidence that
eating disorders have to be committed to a hospital for treatment they are especially effective (Brown & Keel, 2012b). In contrast,
against their will (Watson, Bowers, & Andersen, 2000). Suicide research suggests that treatment with an antipsychotic medication
attempts are often made, and clinicians need to be mindful of called olanzapine may be beneficial. Antipsychotic medications
this risk, even when patients have received a great deal of treat- (which help with disturbed thinking) are routinely used in the
ment (Franko et al., 2004). treatment of schizophrenia (see Chapter 13). These medications
This ambivalence toward recovery is apparent in the behavior also provide benefits in the treatment of anorexia nervosa, which is
of individuals admitted to inpatient units. Hospitalization also characterized by distorted beliefs about body shape and size. More
often means that those with anorexia nervosa will be exposed importantly, one side effect of olanzapine is weight gain. Although
to other patients who are thinner and more experienced than this is a problem for patients with schizophrenia, in the treatment
they are. This can lead to competitive pressure to be the “best of anorexia nervosa weight gain is obviously much more desirable.
anorexic” patient on the unit (Wilson, Grilo, & Vitousek, 2007).
Some “pro-ana” (short for “pro-anorexia”) or “pro-mia” Family Therapy For adolescents with anorexia nervosa,
(“pro-bulima”) websites actively support anorexic or bulimic family therapy is considered to be the treatment of choice.
behavior, creating much concern in the treatment community. The best-studied approach, which (very importantly) blames
These sites are often visited by young people (Custers & Van neither the parents nor the child for the anorexia nervosa, is
den Bulck, 2009). An experimental study has also shown that known as the Maudsley model (le Grange & Lock, 2005). A
exposure to the material on these websites has significant nega- typical treatment program involves 10 to 20 sessions spaced over
tive consequences for young women. After viewing material on a 6 to 12 months. The treatment has three phases. In the refeeding
pro-anorexia website, undergraduate participants reported more phase, the therapist works with the parents and supports their
negative affect and felt less positive about themselves and their efforts to help their child (typically a daughter) to eat health-
bodies than did participants assigned to view other websites that ily once more. Family meals are observed by the therapist, and
concerned fashion or home décor (Bardone-Cone & Cass, 2007). efforts are made to guide the parents as a functioning support
team for their daughter’s recovery. After the patient starts to gain
Treatment of Anorexia Nervosa weight, the negotiations for a new pattern of relationships phase
Individuals with anorexia nervosa view the disorder as a chronic begins, and family issues and problems begin to be addressed.
condition and are generally pessimistic about their potential for Later, in the termination phase of treatment, the focus is on the
recovery (Holliday et al., 2005). They have a high dropout rate development of more healthy relationships between the patient
from therapy, and patients with the binge-purging subtype of and her parents (see Lock et al., 2001).

312  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 312 10/06/13 4:43 PM

 Randomized controlled trials have shown that patients research Close-Up
treated with family therapy for 1 year do better than patients
who are assigned to a control treatment (where they receive sup- Randomized Controlled Trials
portive counseling on an individual basis). Five years after treat-
ment, 75 to 90 percent of patients show full recovery (le ­
Grange & Lock, 2005). In a more recent study, 121 adolescents A randomized controlled trial involves a specific treatment
group (which is the group the researchers are most inter-
were randomly assigned to receive either family-based treat-
ested in) as well as a control treatment group (against which
ment or individual therapy. Both treatments involved a total of
the treatment group will be compared). Participants have an
24 hours of treatment spaced over the course of a year. At the equal chance of being placed in either group because which
end of treatment, 42 percent of patients who received family group they go into is randomly determined.
treatment were in full remission. For those who received individ-
ual therapy, the corresponding figure was 23 percent. When fol-
lowed up another year later, 49 percent of the patients who had
received family-based therapy were well, compared with 23 per- to do better than patients who have been ill for longer or who
cent of those who had been treated with individual therapy have bulimia nervosa (Dare & Eisler, 2002). This highlights the
(Lock et al., 2010). What is encouraging about these findings is importance of early treatment, which may save some patients
that that overwhelming majority of adolescents completed the from a lifetime of suffering. For patients who are older or who
treatments and did not drop out. This suggests that, although have a long history of anorexia nervosa, the Maudsley approach
individual therapy (which encouraged weight gain, the develop- unfortunately provides little clinical benefit (Wilson, Grilo, &
ment of autonomy, and accepting responsibility for food-related Vitousek, 2007).
issues) was slightly less efficacious overall than family treatment,
it was still very acceptable to the adolescents who received it and Cognitive-Behavioral Therapy Cognitive-Behavioral
it still provided some benefit. This is good news, especially for Therapy (CBT), which involves changing behavior and mal-
adolescents whose family members are unable or unwilling to adaptive styles of thinking, has proved to be very effective in
participate in family treatment. treating bulimia nervosa. Because anorexia nervosa shares many
Not surprisingly, family treatment is more helpful for some features with bulimia nervosa, CBT is often used with anorexia
patients than it is for others. Patients who develop anorexia ner- nervosa patients as well (Vitousek, 2002). The recommended
vosa before age 19 and have been ill for fewer than 3 years seem length of treatment is 1 to 2 years. A major focus of the treat-
ment involves modifying distorted beliefs concerning weight
and food, as well as distorted beliefs about the self that may
have contributed to the disorder (e.g., “People will reject me
unless I am thin”).
Pike and colleagues (2003) treated a sample of 33 women
who had anorexia nervosa, after they had been discharged from
the hospital. Over the course of 1 year, the women received
either 50 sessions of CBT or nutritional counseling. Despite this,
only 17 percent of patients who received CBT showed full recov-
ery. None of the women who received nutritional counseling was
fully well (i.e., normal weight, no binge eating or purging, and
with eating attitudes and concerns about weight within normal
limits) at the end of treatment. The limited success of CBT for
patients with anorexia nervosa may be due to the extreme cog-
nitive rigidity that is characteristic of those with this disorder
(Brown & Keel, 2012a). There is clearly a need for new treat-
ment developments, particularly for older patients with more
long-standing problems.

Treatment of Bulimia Nervosa

Medications It is quite common for patients with bulimia
nervosa to be treated with antidepressant medications. Research-
ers became interested in using these medications after it became
Family therapy is now regarded as the treatment of choice for adoles-
cents with anorexia nervosa. The therapy is most effective for those in clear that many patients with bulimia nervosa also suffer from
the earlier stages of the illness. mood disorders. Generally speaking (and in contrast to patients

eating disorders and obesity CHAPTER 9  313

M09_BUTC4286_16_SE_C09.indd 313 10/06/13 4:43 PM

 with anorexia nervosa), patients taking antidepressants do better changing the cognitions and behaviors that initiate or perpetuate
than patients who are given inert, placebo medications. A posi- a binge cycle. This is accomplished by challenging the dysfunc-
tive response is usually apparent within the first 3 weeks. People tional thought patterns typically present in bulimia nervosa,
who do not show early improvement are unlikely to benefit such as the “all-or-nothing” or dichotomous thinking described
from further treatment with the same medication (Sysko et al., earlier. For instance, CBT challenges the tendency to divide all
2010). Perhaps surprisingly, antidepressants seem to decrease the foods into “good” and “bad” categories. This is done by provid-
frequency of binges as well as improve patients’ mood and preoc- ing factual information, as well as by arranging for the patients
cupation with shape and weight (McElroy et al., 2010). to demonstrate to themselves that ingesting “bad” food does not
inevitably lead to a total loss of control over eating. Figure 9.4
Cognitive-Behavioral Therapy The leading treatment shows a cognitive worksheet that was completed by a patient. It
for bulimia nervosa is CBT. Most of the current treatment ap- provides a good example of the kind of “hot thought” that can
proaches are based on the work of Fairburn and colleagues in facilitate a binge.
Oxford, England. Multiple controlled studies that include post- Treatment with CBT clearly helps to reduce the severity
treatment and long-term follow-up outcomes attest to the clini- of symptoms in patients with bulimia nervosa. Still, patients
cal benefits of CBT for bulimia (Fairburn, Jones et al., 1993; with the disorder are rarely entirely well at the end of treatment
Wilson, 2010). Such studies have included comparisons with (Lundgren et al., 2004). Binging and purging is eliminated in
medication therapy (chiefly antidepressants) and with interper- around 30 to 50 percent of cases (Wilson, 2010). Even after
sonal psychotherapy (IPT; see Agras et al., 2000). They generally treatment, weight and shape concerns may remain. In an effort
reveal CBT to be superior. In fact, combining CBT and medica- to improve treatment efficacy, new approaches such as dialectical
tions produces only a modest increment in effectiveness over that behavior therapy (which is a treatment for borderline personal-
attainable with CBT alone. ity disorder; see Chapter 10) are now being explored, with some
The “behavioral” component of CBT for bulimia nervosa success (Safer et al., 2001; Chen & Safer, 2010). Another prom-
focuses on normalizing eating patterns. This includes meal plan- ising development involves using more individualized cognitive-
ning, nutritional education, and ending binging and purging behavior therapy approaches that are specifically tailored to the
cycles by teaching the person to eat small amounts of food more needs of the patient as opposed to a more standardized treatment
regularly. The “cognitive” element of the treatment is aimed at format (Ghaderi, 2006).

Emma’s completed worksheet: Identifying permissive thoughts

Situation Feelings and sensations Permissive thoughts

Friday, at college, alone, had a free Anxious I might as well keep eating now I’ve
period, thinking about my Heavy started.
assignment (how difficult it was Blank
going to be). Ate a bar of chocolate, I can make myself sick afterwards—
knew I was going to binge. Got on the so it doesn’t matter—I can have what
bus to town, went to Burger King— I want and I won’t gain weight.
had a burger, two portions off ries, a
milkshake, one big bar of chocolate, I might as well carry on until my
another smaller bar of chocolate. money has run out. I’ve got to eat
more and more.

• When was it? • What feelings did you • What were you saying to yourself
• Where were you? have? that made it easier to keep eating?
• Who were you with? • What body sensations • Identify and circle the hot thought.
• What were you doing? did you notice? This is the thought that makes it
most likely that you will binge.
• What were you thinking about?

figure 9.4
Cognitive Worksheet.
Source: Reproduced from M. Cooper, G. Todd, and A. Wells, Bulimia Nervosa: A Cognitive Therapy Programme for Clients with per-
mission from Jessica Kingsley Publishers. Copyright © 2000 Myra Cooper, Gillian Todd, and Adrian Wells.

314  CHAPTER 9 eating disorders and obesity

M09_BUTC4286_16_SE_C09.indd 314 10/06/13 4:43 PM

 One of the newest developments in the treatment of eating there were no significant differences between the groups with
disorders is to adopt a transdiagnostic approach to treatment. As regard to remission from binge eating. However, at 2-year follow
we have already mentioned, the majority of patients with eating up, people who had received either IPT or guided CBT were
disorders have a mixed clinical picture. What this means is that doing better than those in the behavioral weight loss group.
they show some symptoms of anorexia and some symptoms of What is also noteworthy is that the dropout rate was much lower
bulimia, combined in a variety of ways. Fairburn and colleagues for people in the IPT group (7% dropped out) than it was in
(2009) have reformulated cognitive behavior therapy for bulimia the guided CBT (30%) or behavioral weight loss groups (28%).
nervosa so that it is now a relevant treatment for pathological This is important because, overall, the dropout rate for minori-
eating, no matter what the diagnosis is. One form of the treat- ties in this study was very high (approximately one-third). The
ment is quite focused, targeting eating issues as well as concerns findings therefore suggest that for racial and ethnic minorities
about shape and weight, extreme dieting, purging, and binge with BED, interpersonal psychotherapy might be a particularly
eating. The other form of the treatment is broader and also suitable treatment approach.
addresses such things as perfectionism, low self-esteem, and rela-
tionship problems. in review
Results from a randomized control trial show that patients • What factors make eating disorders (especially anorexia nervosa)
who receive these treatments do much better than those assigned so difficult to treat?
to a waiting list condition. At the end of both treatments, • Describe the main features of the Maudsley approach for the
regardless of their initial diagnosis, more than half of the sample treatment of anorexia nervosa. For what kind of patients does
had responded well. There were also some hints that patients this approach yield better results?
who had more problems did a little better when they received • What treatment approaches have been shown to be helpful for
the broader treatment, whereas for more straightforward cases binge eating disorder?
the focused form of treatment was very successful. It is impor- • Why do you think cognitive-behavioral therapy is so beneficial
tant to mention that very–low-weight patients with anorexia for patients with bulimia nervosa?
were excluded from this study. Overall, however, the results sug-
• What is a transdiagnostic treatment approach?
gest that the form of the treatment that a patient receives may
not need to be tailored to a very specific DSM diagnosis. Rather,
“transdiagnostic” treatments may have a lot of promise for The Problem of Obesity
patients with eating disorders, broadly defined.
Humans have evolved to be able to store surplus energy as fat.
This has obvious advantages: It serves as a hedge against periods
Treatment of Binge Eating Disorder of food shortage and makes survival more likely during times of
famine. But in our modern world, access to food is no longer a
BED has attracted a lot of attention from researchers, and a
problem for millions of people. The food supply is stable, and
number of different treatment approaches have been suggested.
large amounts of energy-dense foods are readily available. Not
Due to the high level of comorbidity between binge-eating dis-
surprisingly, most of us are getting heavier. For some people, the
order and depression, antidepressant medications are sometimes
problem becomes even more extreme and results in obesity. Con-
used to treat this disorder. Other categories of medications, such
sidered in this way, obesity can be regarded as a state of excessive,
as appetite suppressants and anticonvulsant medications, have
chronic fat storage (Berthoud & Morrison, 2008).
also been a focus of interest (McElroy et al., 2010). Sibutramine,
Worldwide, obesity is now a major public health problem.
a medication that inhibits the reuptake of serotonin and norepi-
From 1980 to 2002 the prevalence of obesity almost tripled in
nephrine, has been shown to reduce the frequency of binges and
adults in the United Kingdom. In China, rates of obesity in
to be associated with more weight loss than a placebo medication
preschool children were 1.5 percent in 1989 and 12.6 percent
in a one clinical trial (Wilfley et al., 2008).
just 9 years later (Flegal et al., 2002). The most recent estimates
In one of the largest and most scientifically rigorous studies
show that one-third of adults in the United States are obese
to date, Wilson and colleagues (2010) randomly assigned 205
(Flegal et al., 2010). Figure 9.5 on page 316 shows how much
overweight or obese men and women who met diagnostic crite-
more prevalent obesity has become in the past two decades.
ria for BED to one of three different treatments. Some people
received interpersonal psychotherapy (IPT), which is a therapy
sometimes used in the treatment of depression (see Chapter 7). Medical Issues
Others received CBT in the form of a self-help book (Overcoming Obesity brings with it increased risk for many health problems.
Binge Eating, Fairburn, 1995), guided by a therapist. People in These include high cholesterol, hypertension, heart disease,
the third group were assigned to a behavioral weight loss treat- arthritis, diabetes, and cancer (Malnick & Knobler, 2006). It has
ment that involved exercise and moderate restriction of calories. been estimated that, by 2030, the cost of treating these prob-
So how did everyone do? At the end of 6 months of treatment, lems in the United States alone will exceed 850 billion dollars

eating disorders and obesity CHAPTER 9  315

M09_BUTC4286_16_SE_C09.indd 315 10/06/13 4:43 PM