Introduction
Eating Disorders as Mutidetermined
Diagnostic Groups
Anorexia Nervosa
Bulimia Nervosa
Eating Disorders Not Otherwise Specified
Relationship Between Different Eating Disorders
Eating Disorders as Multi-determined
Incidence and Prevalence
Gender Considerations
Developmental Considerations
Have Eating Disorders Changed Over Time?
Medical Complications of Eating Disorders
Psychological Disturbance Secondary to Weight Loss or Starvation
Psychological Disturbance Secondary to Chaotic Eating Patterns
Are Symptoms Positively Reinforcing?
Physical Activity and Self-Starvation
Significance of Eating Disorder Symptoms in Athletes
Treatment, Course and Outcome
The Rising Costs of Inpatient Treatment and the Need to Find Alternatives

Introduction
Eating disorders have been described in the psychiatric literature for many years; however, it has only been in the past two decades that they have commanded widespread interest in mainstream psychology, psychiatry, and allied professions (Theander, 2004). One reason for this interest is recognition of their significant health consequences. Anorexia nervosa is estimated to be the third most common chronic medical illness in girls aged 15-19 years old (Lucas, Beard, O.Fallon, & Kurklan, 1991). It is associated with significant medical complications (Becker, Grinspoon, Klibanski, & Herzog, 1999; Sharp & Freeman, 1993) and mortality rates exceed the expected incidence of death from all causes among women 15-24 years of age by 12-fold (Sullivan, 1995). Mortality rates for bulimia nervosa are much lower but still not insignificant (Keel, Dorer, Eddy, Franko, Charatan, & Herzog, 2003; Nielsen, 2003). Moreover, eating disorders during adolescence are associated with elevated risk for a broad range of physical and mental health problems during early adulthood.

Eating Disorders as Mutidetermined
Another reason responsible for the growing attention to eating disorders is that they provide a model of the complex interaction between cultural, somatic, and psychological factors in abnormal functioning. The varied psychological features associated with eating disorders contribute to their heterogeneity on presentation. Finally, there are potentially serious physical and psychological consequences of starvation that must be addressed in the understanding and the treatment of the eating disorders. These factors, as well as the increasing prevalence of eating disorders among adolescent and young adult women, have led to a rapid increase of research on eating disorders and their treatment. The last two decades have led to divergence in etiological formulations as well as convergence of opinion regarding the utility of certain practical intervention principles (Garner & Garfinkel, 1997). In spite of these advancements, current knowledge has yet to yield conclusive support for any one theoretical viewpoint or treatment modality.

Eating Disorders: An Overview© is adapted and summarized from material published previously. For references and text, see:
1. Garner, D.M., Vitousek, K., and Pike, K. (1997). Cognitive-Behavioral Therapy for Anorexia Nervosa. Handbook for Treatment of Eating Disorders. D.M. Garner & P.E. Garfinkel (Eds.). New York, NY: Guilford Press, 94-144.
2. Garner, D.M. (1997). Psychoeducational principles in the treatment of eating disorders. In: Handbook for Treatment of Eating Disorders. D.M. Garner & P.E. Garfinkel (Eds.). New York, NY: Guilford Press, 145-177.
3. Garner, D.M., Rosen, L., and Barry, D. (1998). Eating Disorders in Athletes. In: Child and Adolescent Psychiatric Clinics of North America, 7, New York: W.B. Saunders, 839-857.
4. Garner, D.M. (2004). The Eating Disorder Inventory-3 (EDI-3). Professional Manual. Odessa, Florida: Psychological Assessment Resources.
5. Garner, D.M. and Magana, C. (2005). Cognitive Vulnerability in Anorexia Nervosa. In L.B. Alloy & J. H. Riskind (Eds.) Cognitive Vulnerability to Emotional Disorders. Lawrence Erlbaum Associates, Inc., 365-403.

Diagnostic Groups
The prevailing diagnostic systems, namely the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV-TR; American Psychiatric Association [APA], 2000) and the International Classification of Diseases, 10th revision (World Health Organization [WHO], 1992), formally define as well as distinguish the two major eating disorders, anorexia nervosa and bulimia nervosa. By drawing the boundaries for these eating disorders, the current DSM-IV-TR diagnostic criteria have substantial implications for clinical care and research.

Anorexia Nervosa
The requirements for diagnosis of anorexia nervosa according to the DSM-IV-TR (2000) are summarized as follows: 1.) Refusal to maintain a body weight over a minimally normal weight for age and height (e.g., weight loss leading to maintenance of a body weight less than 85% of that expected or failure to make expected weight gain during a period of growth). 2.) Intense fear of gaining weight or becoming fat, even though underweight. 3.) Disturbance in the way that body weight, size, or shape is experienced. 4.) Amenorrhea in females (absence of at least three consecutive menstrual cycles).
The DSM-IV-TR (APA, 2000) divides anorexia nervosa into two diagnostic subtypes: 1) restricting type and 2) binge eating/purging type. The restricting type is defined by rigid restriction of food intake without bingeing or purging, while the binge eating/purging type is defined by stringent attempts to limit intake, followed by episodes of binge eating as well as self-induced vomiting and/or laxative abuse. This definition differs from earlier descriptions of anorexia, in which the disorder was simply subdivided based on the presence or absence of binge eating. Patients who regularly engage in bulimic episodes report greater impulsivity, social/sexual dysfunction, substance abuse, general impulse control problems, family dysfunction, and depression as part of a general picture of more obvious emotional disturbance than patients with the restricting subtype of anorexia nervosa. Those with the restricting type of anorexia nervosa are often characterized as overly compliant, but also obstinate, perfectionistic, obsessive-compulsive, shy, introverted, interpersonally sensitive and, stoical.

Bulimia Nervosa
The criteria for the diagnosis of bulimia nervosa according to the DSM-IV-TR (APA, 2000) are summarized as follows: 1.) Recurrent episodes of binge eating (a sense of lack of control over eating a large amount of food in a discrete period of time). 2.) Recurrent inappropriate compensatory behavior in order to prevent weight gain (i.e., vomiting, abuse of laxative, diuretics or other medications, fasting or excessive exercise). 3.) A minimum average of two episodes of binge eating and inappropriate compensatory behaviors per week for the past three months. 4.) Self-evaluation unduly influenced by body shape and weight. 5.) The disturbance does not occur exclusively during episodes of anorexia nervosa. Bulimia nervosa patients are further subdivided into purging and non-purging types based on the regular use of self-induced vomiting, laxatives or diuretic (APA, 2000).

Eating Disorders Not Otherwise Specified (EDNOS)
The DSM-IV-TR (APA, 2000) delineates a large and diverse diagnostic category, "Eating Disorder, Not Otherwise Specified” (EDNOS), for individuals with clinically significant eating disorders who do not meet all of the diagnostic criteria for anorexia nervosa or bulimia nervosa. Unfortunately, the term "not otherwise specified" could be interpreted as meaning that these eating problems have minor clinical significance. This assumption is incorrect since the clinical picture for many individuals with EDNOS can be every bit as complicated and serious as the two main eating disorders. For example, in order to qualify for this diagnosis, binge eating must occur, on average, at least twice a week for a six month period. This means that patient who is normal weight and who does not engage in objective binge-eating episodes but vomits 20 times following subjective binges (small amounts of food), would be classified as EDNOS rather than bulimia nervosa. Many patients complain that having a diagnosis of “not otherwise specified” discounts their level of suffering and makes them feel like they do not have a “real” problem.

Relationship between Different Eating Disorders
Current systems of diagnosing the two main eating disorders into mutually exclusive categories, based largely on body weight, present some limitations. This distinction overlooks the overlap between these eating disorders. Bulimia and anorexia nervosa can have virtually identical cognitive, clinical, psychological and behavioral symptoms. Binge eating is the hallmark for bulimia nervosa; however, it is also a common symptom in anorexia nervosa. The major difference between the disorders is that in anorexia nervosa (binge eating/ purging subtype), these symptoms occur at a low body weight. Nevertheless, the body weight threshold used to define anorexia nervosa is arbitrary. Some patients move between the diagnostic categories of anorexia and bulimia nervosa at different points in time based only on variations in body weight; thus, it is counterintuitive to suggest that these diagnostic entities are not shared to some extent. Moreover, the extraordinary variety within each of the anorexia and bulimia nervosa diagnostic subclasses emphasizes the need to be careful about generalizations based only upon diagnosis. Body weight is not likely to be the ideal marker for differentiating between eating disorders (Garner, Garner and Rosen, 1993).


Extraordinary variability exists within each of the diagnostic subgroups, in terms of demographic, clinical, and psychological variables. People with anorexia and bulimia nervosa can move between the diagnostic categories at different points in time. For example, some patients may alternate between the two subtypes of anorexia nervosa (restricting and binge eating/purging types). However, it is more common for restricters to move toward bulimia (and purging) than for bulimic patients to move to an exclusively abstaining mode. Some women move back and forth between the two subclasses for years.

Although distinctions between diagnostic subgroups have been emphasized in research, recognizing that the different subclasses share numerous features is important. For example, even though persons with anorexia are differentiated into "restricter" and “binge eating /purging" subtypes, virtually all eating disorder patients "restrict their food intake", "diet" and probably "fast" for abnormally long periods of time. Some do this in association with binge eating, some with vomiting, and/or purging, and others with neither of these symptoms. For some persons, these symptoms occur at a statistically "normal" body weight (bulimia nervosa) and for others it occurs well over the body weight norms (e.g. binge-eating disorder).

The “classic” case of anorexia nervosa has historically been known for the restricting eating pattern rather than for binge eating and some have emphasized the relative similarity of this “restricting anorexia nervosa” (RAN) subgroup. The RAN subgroup has been described as perfectionistic, obsessional, inflexible, socially introverted, emotionally reserved, conflict avoidant, rule-minded and overly compliant. The RAN subgroup also has been distinguished from individuals with more temperamental personalities who have the “typical” bulimic symptom pattern. These people are generally characterized as having greater impulsivity, interpersonal conflict, and mood instability, regardless of whether they are at normal body weight in bulimia nervosa or as a suboptimal weight in anorexia (binge-eating/purging subtype). However, as mentioned earlier, inferring specific psychological traits from the behavioral symptoms of an eating disorder is overly simplistic and does little to explain meaningful distinctions between patient subgroups (Garner et al., 1993).

Eating Disorders as Multi-determined
During the past several decades, the idea that eating disorders are caused by a single factor has been replaced by the view that eating disorders are "multi-determined". More than 25 years ago, Garner & Garfinkel (1980) proposed a model in which symptom patterns represent final common pathways resulting from the interplay of three broad classes of predisposing factors shown in Figure 1. Accordingly, environmental, cultural, individual (psychological and biological) and familial causal factors combine with each other in different ways leading to the development of eating disorders.

Cumulative evidence in recent years has specified in greater detail the respective roles of cultural, individual (developmental, psychological, biological, and genetic) and familial risk factors that contribute to the expression of eating disorders (Jacobi, Hayward, de Zwaan, Kraemer, & Agras, 2004; Stice, 2002). Important predisposing factors include (a) female gender; (b) living in Western society; (c) adolescence and early adulthood; (d) low self-esteem; (e) perfectionism; (f) depression; and a family history of any type of eating disorder, obesity, depression, or substance abuse. Significant precipitating factors include: (a) dieting to lose weight; (b) occupational or recreational pressures to be slim; (c) critical comments about weight and shape; and (d) sexual abuse. Key perpetuating factors are the psychological, emotional, and physical effects of starvation. Many risk factors do not fit into a simple paradigm as a predisposing, precipitating, or perpetuating factor and the precise mechanism of action for many known risks remains elusive. The period of vulnerability for certain risk factors is fixed because of their nature (e.g., gender or birth complications) whereas others may exert their influence at multiple points in the development and maintenance of the disorder (e.g., family dieting or a genetic liability for perfectionism). Although the strength of evidence for different risk factors varies considerably, the variables that have received the most research interest in recent years are listed below.

Table 1: List of Risk Factors with Some Research Support

Table 1: List of Risk Factors with Some Research Support

SOCIO-CULTURAL
Living in Western Society
Sports Emphasizing Slimness

INDIVIDUAL / PSYCHOLOGICAL
Female Gender
Ethnicity
Adolescence
Higher Premorbid Weight
Pubertal Status and Timing
Restrictive Dieting
High Levels of Exercise
Alcohol / Drug Abuse
Drive for Thinness
Body Dissatisfaction & Weight Concerns
Body Misperception
Low Self-Esteem or Ineffectiveness
Depression
Perfectionism
Asceticism / Self-Sacrifice
Obsessive Compulsive Traits
Anxiety / Worry
Interpersonal Insecurity / Alienation / Attachment Deficits
Fears of Psychobiological Maturity
Harm-Avoidant Temperament /
Low Interoceptive Awareness
Emotional Regulation Problems (over or under-control)
Anger
Reasoning Errors Information Processing /
Attentional Bias
Neuropsychological Deficits

GENETIC FACTORS

INDIVIDUAL DEVELOPMENTAL
Early Childhood Feeding Problems
Pregnancy Complications / Premature birth
Childhood Anxiety Problems
Childhood Obesity
Early Puberty
Age: Adolescence

ADVERSE LIFE EVENTS
Physical Abuse or Neglect
Sexual Abuse
Bullying and Teasing
Physical Illness
Weight Loss in Adolescence

PARENTAL RISK FACTORS
Obesity
Dieting
Mother with an Eating Disorder
Critical Comments about Weight
High Levels of Exercise
High Performance Expectations
Over-Concern / Hypervigilance
Depression
Low Contact / Neglect / Conflict
Substance Misuse

Based on summary data from: Jacobi, Hayward, de Zwaan, Kraemer, & Agras, 2004; Stice, 2002; Fairburn, & Harrison 2003; Ricciardelli & McCabe (2004).

Incidence and Prevalence
Incidence rates are defined as the number of new cases in the population per year, whereas prevalence rates refer to the actual number of cases in the population at a certain point in time. Estimating incidence from primary care practice records has provided rates of 8.1 per 100,000 persons per year for anorexia nervosa and 11.5 per 100,000 persons per year for bulimia nervosa (Hoek et al., 1995). Prevalence estimates for eating disorders vary widely depending on the methods used and definitions of the disorder. Self-report methods have generally produced higher prevalence estimates than interview-based methods. In contrast, data from epidemiological studies have led to the conclusion that current diagnostic criteria may be overly restrictive, due to the fact that many of those who fail to meet certain diagnostic thresholds have similar psychiatric morbidity to those with the full syndrome (Sullivan, Bulik, & Kindler, 1998).

Gender Considerations
Epidemiological studies have consistently found that eating disorders are more common among women than men. In clinical samples, only 5% to 10% of patients are men (Hoek et al., 1995). Studies in the U.S. suggest the incidence of anorexia nervosa among males may be as little as 0.02% per year and the prevalence of current bulimia nervosa among males is between 0.1% and 0.5%. The differences in gender rates are generally attributed to cultural factors; however, when they occur in men, the clinical presentation is very similar to that observed in women (Andersen, 1990).

Developmental Considerations
From the earliest descriptions, there has been agreement that anorexia nervosa is a disorder primarily of adolescence and that the core psychopathology often relates to conflicts that emerge with pubertal changes in body shape. Theander (1996) reviewed the topic of age of onset; he concluded that the average age of onset is at approximately 17 years old and that approximately 30% of cases have an onset at 14 years old or younger. Selection bias yields a younger age at onset in pediatric treatment facilities. Theander summarizes the distribution of age of onset by stating that: 1) there are occasional patients with an early age of onset (7 to 10 years old); 2) the number of cases rises steeply at 11 and 12 years old; 3) most patients develop in the teenage years (13 to 19 years old); 4) there is a gradual decline in onset from 20 to 30 years old; and 5) there are fewer cases with an onset after 30 years old. Some have found that age of onset tends to be a bit younger for males than females; however, this point remains controversial.

Have Eating Disorders Changed Over Time?
There is evidence that eating disorders may manifest differently, depending on the level of emphasis the culture places on dieting at a particular point in time. Casper (1983) suggested a historical shift in the psychopathology of anorexia nervosa from earlier accounts of “ascetic motives” to the more modern motivational themes of "drive for thinness" and “fear of fatness.” Russell (1985) expanded on this observation by noting that bulimia nervosa had become part of the diagnostic nomenclature due to growing pressures on women to diet to meet unrealistically thin standards of beauty.

Support for this view comes from reports indicating that eating disorders may be earlier in “their evolution” in certain Asian countries. Lee, Ho, and Hsu (1993) have identified a subgroup of “non-fat-phobic” patients with anorexia nervosa in China who are distinguished by low scores on the EDI Drive for Thinness and Bulimia subscales but show significant psychopathology on the other EDI subscales (Lee, Lee, & Leung, 1998) and a better intermediate outcome (Lee, Chan, & Hsu, 2003).

Medical Complications of Eating Disorders
The medical complications of eating disorders are significant, potentially irreversible, and life-threatening; thus, a medical evaluation is essential as part of a comprehensive assessment. Medical complications can involve almost all organ systems and can be viewed as consequences of starvation, vomiting, laxative abuse, binge eating, and exercise (Becker, Grinspoon, Klibanski, & Herzog, 1999; Pomeroy & Mitchell, 2002; Rome & Ammerman, 2003). Table 1illustrates some of the medical complications associated with starvation and the symptoms used to control weight.

Certain medical complications can occur as a direct or an indirect result of treatment. For example, in AN, “refeeding syndrome” can lead to delirium and death and is associated with oral parenteral intake of highly caloric nutrients, particularly those high in glucose (Rome & Ammerman, 2003). Similarly, patients with AN may “water load” before being weighed to provide false information about their weight. Others may consume excessive water under the misguided belief that it “purifies the body” or reduces appetite. Water intoxication can lead to cerebral edema, ataxia, seizures, coma, and death (Jacquin, Ouvry, & Alvin, 1992). Many of the medical complications are reversed with weight restoration and the cessation of eating disorder symptoms; however, some are irreversible, such as myocarditis that can occur as a result of emetine toxicity from ipecac abuse; renal damage; and osteopenia that is increased the longer a patient remains underweight (Hotta, Shibasaki, Sato, & Demura, 1998). Pregnancy in patients with eating disorders leads to higher rates of miscarriage, obstetric complications, and postpartum depression (Franko & Spurrell, 2000). Finally, having an eating disorder as an adolescent greatly elevates the risk for physical, as well as mental, disorders during early adulthood (Johnson et al., 2002).

Table 2 Medical Complications of Eating Disorders

Starvation and weight loss
Amenorrhea (primary or secondary)
Delayed puberty
Reduced metabolic rate
Cardiac arrhythmia
Prolonged qt interval
Sinus bradycardia
Cardiac muscle wasting
Mitral valve prolapse
Orthostatic changes (pulse, hypothermia, hypotension)
Reduced ventricular mass
Cerebral atrophy
Constipation
Renal failure
Lanugo hair
Delayed gastric empting, slowed gastric motility
Decreased bone density (osteopenia)
Abnormal laboratory values
Hypercholesterolemia
Hypercortisolemia
Hypophosphatemia
Increased growth hormone
Leukopenia

Self-induced vomiting
General fatigue
Weakness
Muscle weakness and cramping
Cardiac arrhythmias
Renal failure
Dental enamel erosion
Postural edema
Parotid gland swelling
Gastric reflux and irritation
Chronic metabolic alkalosis
Chronic metabolic acidosis (with laxative abuse)
Vital sign abnormalities
Calluses on the knuckles or stomach from inducing vomiting
Abnormal laboratory values
Hypokalemia
Hypocalcemia
Hypomagnesemia
Hypophosphatemia
Hyperamylasemia

Binge eating
Esophageal rupture
Acute gastric dialation and rupture

---

Sources: “Current concepts: Eating disorders,” by A.E. Becker, S.K. Grinspoon, A. Klibanski, and D.B. Herzog, 1999, New England Journal of Medicine, 340, 1092-1098; “Medical complications of anorexia nervosa and bulimia nervosa, by C. Pomeroy and J.E. Mitchell, 2002. In C. G. Fairburn and K. D. Brownell (Eds.), Eating disorders and obesity (2nd ed., pp. 278-285); and “Medical complications of eating disorders: An update,” by E.S. Rome and S. Ammerman, 2003. Journal of Adolescent Health, 33, 418-426.

Psychological Disturbance Secondary to Weight Loss or Starvation
The profound effects of weight loss on psychological functioning is illustrated in a well-known “semi-starvation” study by Keys, Brozek, Henschel, Mickelsen, and Taylor (1950) that should be familiar to all clinicians who are assessing or treating individuals with eating disorders. The study involved exposing a group of normal male volunteers to 6 months of sustained calorie restriction that resulted in the men losing an average of 23% of their former body weight (see Garner, 1997 for a detailed review). Although the experiment was described as a “semi-starvation” study, the men consumed, on average, approximately 1,550 calories per day, which is more calories than prescribed in “conservative” treatments for obesity (Stunkard, 1993).

One of the most of the striking changes that occurred in the volunteers was a dramatic increase in food preoccupations (Keys et al., 1950). The men found concentration on their usual activities increasingly difficult, because they became plagued by incessant thoughts of food and eating. There were remarkable changes in their eating habits in response to weight loss. The men spent much of the day planning how they would eat their allotment of food; they would either eat ravenously or very slowly to appreciate the taste of each morsel. During the restrictive dieting phase of the experiment, all the volunteers reported increased hunger. Some appeared able to tolerate the experience fairly well, but, for others, it led to episodes of binge eating that were indistinguishable from those seen among patients with eating disorders. The tendency toward binge eating persisted, or even became worse, during the refeeding phase of the experiment. Many of the men reported lost control of their appetites and “ate more or less continuously” (p. 843), some consuming between 8,000 and 10,000 calories a day.

In the area of eating disorders, many theories emphasize emotional factors in the etiology and maintenance of binge eating. Binge eating is often considered functional in regulating mood, numbing a person from unpleasant affect, or as otherwise expressing underlying psychological conflicts. While binge eating may serve many of these or other psychological functions, the experiences of the Keys et al. (1950) experiment indicate that the most parsimonious explanation of increased hunger and binge eating episodes is that they are directly linked to undernutrition and weight loss.

Although the participants were psychologically healthy before the experiment, most experienced significant emotional deterioration as a result of the weight loss (Keys et al., 1950). Most of the men experienced time periods during which their emotional distress was quite severe; almost 20% experienced extreme emotional deterioration that markedly interfered with their functioning. The most common emotional change was increased depression; however, elation was observed in some, and mood swings were extreme for some of the volunteers. As indicated in Table 4.2, starvation produced other emotional changes, including feelings of inadequacy and social withdrawal.

Many of the starvation-induced symptoms observed in this study took months to resolve, even after body weight had returned to normal levels (Keys et al., 1950). There are a number of important ways that these symptoms can maintain AN. Delayed gastric emptying can make the patient with AN, who is already sensitive to slight increases in weight, experience postprandial bloating (Lee, 1993). The tendency to engage in binge eating, particularly during the weight-regain process, causes guilt that can lead to compensatory symptoms such as vomiting. Starvation-induced obsessionality, depression, emotional sensitivity, and social isolation can aggravate pre-existing emotional disturbance. Minnesota Multiphasic Personality Inventory (MMPI; Butcher, Dahlstrom, Graham, Tellegen, & Kraemmer, 1989) profiles in a small minority of participants indicated the profound deterioration that can result from weight loss. One man, who scored well within normal limits at initial testing, was retested after 10 weeks of weight loss (10 pounds [4.5 kg], or approximately 7% of his original body weight), and he showed gross personality disturbances on the MMPI.

As indicated in Table 3, the men experienced numerous physical changes, including gastrointestinal discomfort, edema, decreased need for sleep, and decreased tolerance for cold temperatures. Sexual interests were likewise drastically reduced and Keys et al. (1950) observed that “many of the men welcomed the freedom from sexual tensions and frustrations normally present in young adult men” (p. 840). The fact that starvation perceptibly altered sexual urges and associated conflicts is of particular interest, because it has been hypothesized that this process is the driving force behind the dieting of many patients with AN. According to Crisp (1980), AN is an adaptive disorder in the sense that it curtails sexual concerns for which the adolescent feels unprepared. During rehabilitation, sexual interest was slow to return.

This study is significant to the assessment of psychological and physical functioning of individuals with eating disorders. Many of the attributes assumed to be primarily emotional disturbances may be interpreted as secondary to weight loss. Moreover, many of the changes observed in the men persisted for many months; therefore, the assessing clinician must be mindful of the lasting effects of weight loss on many of the psychological domains assessed by psychological measures that assess a wide range of psychological attributes.

Psychological Disturbance Secondary to Chaotic Eating Patterns
The psychological disturbances found on psychological testing may indicate factors that play a role in causing or maintaining an eating disorder; however, it is possible that their presence or magnitude may simply be by-products of the disorder. Generalizations regarding the primacy of personality features and other psychopathological traits are limited by the fact that most assessments have been performed while patients were actively symptomatic with the eating disorder. Observed traits may indicate secondary elaborations resulting from chaotic eating patterns. Studies with a range of psychological measures have shown that those with BN who respond to treatment with rapid reductions in bingeing and vomiting display a corresponding decline in depression and other signs of emotional disturbance (Fairburn, Cooper, Kirk, & O’Connor, 1985; Garner, Olmsted, Davis, et al., 1990; Wilson, Rossiter, Kleifield, & Lindholm, 1986). Garner, Olmsted, Davis, et al. (1990) found that patients with BN whose bingeing and vomiting was brought under control during treatment showed significant improvements on measures of borderline traits, social adjustment, depression, self-esteem, and other areas of functioning. Thus, the presence of an eating disorder may have a considerable influence on the assessment of psychological distress and other traits that could lead to erroneous inferences regarding the presence of personality disturbance. Recognition that chaotic eating patterns affect psychological functioning does not invalidate clinical observations made during the acute phase of the eating disorder; however, it should be the basis for caution in interpreting the results. Greater confidence that symptoms reflect primary disturbance may come from determining if they were evident before the onset of the eating disorder.

Are Symptoms Positively Reinforcing?
Many of the core symptoms in both AN and BN (e.g., extreme dieting and exercise) have been described as “ego-syntonic,” in that the person with an eating disorder may actively embrace them with little or no motivation to change. This is very common in the early stages of an eating disorder, because the patient may experience improved self-esteem and feelings of self-control without experiencing some of the more distressing symptoms that come with chronicity. Also, early on, there may be considerable social reinforcement for the initial weight loss and the self-control required to adhere to a diet. Many eating-disorder symptoms are initiated by voluntarily patient, and any attempt to suggest that they are problematic is met with resistance. Eating disorders have acquired a positive social connotation to some who actively strive for, and then cling to, an “anorexic identity,” due to its associations with celebrity status and socially desirable traits (Garner & Bemis, 1985),

Physical Activity and Self-Starvation
There is some evidence that a participation in sports or strenuous physical activity may play a pathogenic role in the development and maintenance of eating disorders by disrupting the normal mechanisms of energy balance. Epling, Pierce and colleagues (Epling and Pierce, 1996; Epling et al., 1983) advanced the “activity induced anorexia” hypothesis based on well controlled animal research showing that food intake is reduced following exercise, and that in the face of food restriction, physical activity increases, creating a weight loss cycle. Early studies indicated that about 90% of adolescent rats and 70% of mature rats whose access to food was restricted to 60-90 minutes a day and who were given free access to a running wheel, progressively increased the amount of running time and decreased food intake, sometimes literally running themselves to death (Epling et al., 1983). Epling and Pierce (1996) maintain that under-eating and over-exercising can be mutually reinforcing and self-perpetuating behaviors that are physiologically mediated by the release of endogenous opioid peptides in the brain.

Davis et al. (1994) followed this line of reasoning in suggesting that participation in sports and exercise may play a fundamental role in the development of anorexia nervosa. In a study of hospitalized anorexia nervosa patients and age-matched controls they found that patients were more physically active than controls from adolescence onward. Excessive exercise prior to the onset of the disorder was reported by 78% of patients and 60% indicated that they were competitive athletes or dance. Furthermore, 75% of the patient sample claimed that physical activity steadily increased during the period when food intake and weight loss decreased the most. Thus, excessive exercise may be more than a deliberate method of calorie expenditure in anorexia nervosa; it may actually predispose the person to the progression of self-starvation. Excessive physical activity may also interfere with recovery. Strober (1997) reproted that the strongest predictor of relapse in a long-term follow-up of adolescent anorexia nervosa patients was feeling compelled to engage in exercise.

It would be naive to assume that “activity induced anorexia” can explain the development and progression of a disorder as complex as anorexia nervosa. However, this theory and it provides valuable insights into possible biological mechanisms that underlying the breakdown of normal energy regulation when food restriction is coupled with strenuous exercise. Moreover, the notion that calorie restriction and excessive activity may potentate one another must be taken seriously in sports where athletes are encouraged to be thin to meet performance or appearance standards.

Table 3: Warning Signs for Eating Disorders

Significant weight loss
Dramatic weight fluctuation
Denial of hunger
A significant reduction of the amount of food eaten
Avoiding eating, skipping meals
Dieting becomes the most important part of the person's life
Rigid eating patterns (only diet foods or foods that are perceived to be low in calories are consumed)
Dividing food into "good" (fruits and vegetables) and "bad" (meats, sweets) categories
Unusual rituals at mealtime such as cutting food into small pieces, moving food around the plate and disposal of food so that it will not have to be eaten
Preoccupation with food, collecting recipes and cooking for others while finding excuses to avoid eating
Storing or hoarding food
Binge eating
Marked fear or anxiety before eating and guilt after eating
Complaints of bloating and unusual fullness after eating small amounts of food
Signs of self-induced vomiting (e.g. leaving the table immediately after eating in order to vomit)
Signs of laxative abuse or diuretic abuse
Chewing food and spitting it out without swallowing
Excessive exercise to burn off calories
Loss of menstrual periods
Dry skin and swollen salivary glands
Muscle spasm, numbness and tingling in the arms and legs
Sore throat and erosion of tooth enamel caused by vomiting
Social isolation, low self-esteem & depression

Significance of Eating Disorder Symptoms in Athletes
Some researchers have argued that ballet students with symptoms of anorexia nervosa only superficially resembled actual clinical cases because the students experienced considerable improvement in their eating disorder after one year without medical intervention Szmukler et al. (1985). In a study of 49 female ballet students, Le Grange et al. (1994) identified 12.3% of the sample as having anorexia nervosa or a “partial syndrome anorexia nervosa” and agreed with the Szmuckler et al. interpretation that the disorders may not have the same clinical significance in the ballet as they do in clinical samples. Rather than being conceptualized as an “illness,” Le Grange et al. (1994) suggested that eating disorders in the ballet sample should be considered as “adaptive” because of the special “nutritional needs and body shape requirements” of ballet dancers. Offering further support against their deleterious role, Le Grange et al. (1994) point out that having an eating disorder did not seem to interfere with students completing their academic year.

The suggestion that eating disorders might be thought of as benign or adaptive variant in certain environments is misguided in light of what is known about the irreversible skeletal damage (including osteoporosis, stress fractures, and scoliosis. This is not to mention the well established risk of dancers developing serious eating disorders that can have devastating physical and psychological consequences. Serious questions need to be raised to the major sports regulatory bodies about the social and ethical implications of pressuring young women, many of whom are under the legal age to give consent, to lose body weight in order to meet the prevailing requirements for shape, weight and nutritional intake required in the ballet. For example, in the Le Grange et al. (1994) study, just over half of the ballet students studied were underweight, perceived themselves to be fat and controlled their food intake by dieting, vomiting, and laxative or diuretic abuse. About a third reported menstrual disturbance or delayed menarche, and as mentioned earlier, these symptoms can lead to irreversible bone loss and other health risks. Finally, Le Grange et al. (1994) reported that those ballet students who had eating disorders continued to experience serious symptoms over the follow-up period. This is similar to an earlier prospective study by Garner et al. (1987) who found that young ballet students with elevated Drive for Thinness and Body Dissatisfaction scales on the EDI were at risk for developing or maintaining anorexia nervosa during a 2 to 4 year follow-up. In sum, the real physical risks of maintaining a low body weight mentioned earlier as well as the persistence of serious eating disorder symptoms among dancers would be sufficient to counter any argument that eating disorders in the ballet are benign.

TABLE 4: The Effects of Semi-Starvation From a 1950 Minnesota Study*

Attitudes and behavior toward food
Food preoccupation
Collection of recipes, cookbooks, and menus
Unusual eating habits
Increased consumption of coffee, tea, and spices
Gum chewing
Binge eating

Emotional and social changes
Depression
Anxiety
Irritability and/or anger
Lability
"Psychotic" episodes
Personality changes on psychological tests
Social withdrawal

Cognitive changes
Decreased concentration
Poor judgment
Apathy

Physical changes
Sleep disturbances
Weakness
Gastrointestinal disturbances
Hyperacuity to noise and light
Edema
Hypothermia
Paresthesia
Decreased basal metabolic rate (BMR)
Decreased sexual interest

---

aKeys, Brozek, Henschel, Mickelsen, and Taylor (1950). Note. From “Psychoeducational principles in treatment,” by D. M. Garner, 1997. In: D.M. Garner & P.E. Garfinkel (Eds.), Handbook of treatment for eating disorders (pp. 147-177). New York: Guilford Press. Adapted with permission of the author.

Treatment, Course and Outcome
There have been major advancements in the treatment of both anorexia nervosa and bulimia nervosa in recent years. Although there are large variations across outcome studies, approximately 70% of adolescents with anorexia nervosa recover from their eating disorder. The course can be difficult for these patients, and a significant proportion remain impaired in terms of psychological adjustment and social functioning (Patton, Coffey, & Sawyer, 2003; Steinhausen, Boyadjieva, Griogoroiu-Serbanescu, & Neumarker, 2003; Strober, Freeman, & Morrell, 1997). Long-term follow-up studies of adults with anorexia nervosa indicate a less favorable outcome with approximately 50% recovery rates; however, again, there are significant variations in outcome across follow-up studies (Steinhausen, 2002). Nevertheless, there is tremendous variability in treatment outcome from different centers, suggesting that the components for effective treatment exist but are not well established or consistently applied.

There is more evidence of effective treatment for bulimia nervosa, with as many as 70% recovering within 3 months of an empirically validated outpatient cognitive-behavioral therapy (Fairburn et al., 1995). However, recent studies have emphasized the fluctuating nature of symptoms over the follow-up period with many patients who have pervasive psychological and interpersonal problems (Fichter & Quadflieg, 1997; Keel, Mitchell, Miller, Davis, & Crow, 1999; Wilson, Fairburn, Agras, Walsh, & Kraemer, 2002). This cannot be attributed simply to selective bias associated with treatment seeking, because prospective studies of representative community-based cases indicate that poor psychosocial functioning and psychological maladjustment affect a significant subgroup of patients with anorexia nervosa (Wentz, Gillberg, Gillberg, & Rastam, 2001) and bulimia nervosa (Fairburn et al., 2003). Even among those who recover from anorexia nervosa, there is evidence of elevated rates of persistent perfectionism, obsessionality, and poor social functioning (Patton et al., 2003; Steinhausen et al., 2003), as well as an increased risk for a broad range of physical and mental health problems during early adulthood (Johnson, Cohen, Kasen, & Brook, 2002).

The Rising Costs of Inpatient Treatment and the Need to Find Cost-Effective Alternatives
The rising cost of inpatient treatment has resulted in shorter lengths of stay over the past 15 years, creating a major barrier to care for those suffering from eating disorders. According to Weisman et al. (2001), the total cost for an inpatient day at New York Presbyterian Hospital in 1984 was $415 per day ($651 in 1998 dollars); this rose to $1,400 per day in 1998. Due to higher hospitalization costs, the insurance industry began decreasing lengths of stay. As shown in Figure 2, the average length of stay in 1984 was 150 days, but this dropped to 24 days in 1998 (Weisman et al., 2001). The resulting lower discharge weights indicated in Figure 1 have meant patients are being discharged at a BMI barely above the conservative cut-off for anorexia nervosa. This is a weight repeatedly shown in follow-up studies to cause relapse in anorexia nervosa (Baran, Weltzin & Kaye, 1995; Commerford, Licinio & Halmi, 1997; Fichter & Quadflieg, 1999; Howard, Evans, Quintero-Howard, Bowers & Andersen, 1999; Steinhausen, 2002; Zipfel, Lowe, Reas, Deter & Herzog, 2000). Some clinicians have advocated increased funding for treatment (Litt, 1999), but the reimbursement required to achieve weight restoration in anorexia nervosa and symptom control in treatment-resistant bulimia nervosa can pose an extraordinary burden for insurance companies and others who must pay for treatment. The goal in developing the River Centre Clinic Eating Disorder Program was to find a safe and effective alternative to inpatient treatment that could be effective for patients with anorexia nervosa and treatment resistant bulimia nervosa.