Eating Disorders: An Overview

By Jeanne Rust, Ph.D.

 

Her skin was cracked, dry, and flaking. The hair from her head had begun to fall outin clumps. Her eyes were becoming duller and she had to put out her hand tofind her way. She tried harder and harderto conceal her limp when she walked. Even making the effort to speak was nearly impossible. The words had frozen in her throat. She fell. She lay frozen and immobile on the hard ground.

Night was falling and the moon began to rise. The light from the moon illuminated her bodyand in the background the bodies of her sisters lay near her, frozen andimmobile on the hard ground.

This is the graveyard of our sisters. These lifeless bodies are those of our familymembers, our friends and our patients with eating disorders, no matter whetherthey are bulimic, anorexic, or are compulsive eaters.

Eating disorders ultimately kill.
– JNR (2006)

During the past thirtyyears, the incidence of eating disorders has increased dramatically in the United States. Eating disorders affect a significant numberof Americans, an estimated 5 million every year (Kreipe, Golden, Katzman,Fisher, Rees, Tonkin, et al., 1995). These disorders include anorexia nervosa, bulimia nervosa, binge-eatingdisorders and several other variants. Although these disorders are more common in adolescent girls or youngwomen, approximately 7 million girls and adult women struggle with eatingdisorders and approximately 1 million boys and men will struggle with eatingdisorders this year (Katz, 2003).

Anorexia nervosa, bulimianervosa, and binge eating disorder are prevalent and often chronic in ourcurrent culture (Halmi, Eckert, Marchi, Sampugnaro, Apple, & Cohen, 1991;Herzog, Keller, Lavori, & Gray, 1992). One-half of one percent to 3.7% of females will suffer from anorexianervosa in their lifetime; 1.1% – 4.2% of females will suffer from bulimianervosa in their lifetime; 2% – 5% of the American population will experiencebinge eating disorder; 10%-25% of all those battling anorexia will die as adirect result of the eating disorder; up to 19% of college aged women inAmerica are bulimic; and 10% of all persons with eating disorders are male(Katz, 2003). Eating disorders have oneof the highest mortality rates of any psychiatric diagnosis (Agras, 2001).

Prevalence of EatingDisorders and Body Dissatisfaction

Eating disorders can be viewedon a continuum from normal to pathological and the numbers of people, men aswell as women, with clinical eating disorders, disordered eating, and bodydissatisfaction is growing. Statisticsshow that 42% of first through third grade girls wantto be thinner (Collins, 1991) and that 81% of 10-year-olds are afraid of beingfat (Mellin, McNutt, Hu, Schreiber, Crawford, & Obarzanek, 1991). Ninety-one percent of women recently surveyedon a college campus had attempted to control their weight through dieting. Twenty-two percent dieted “often”or “always” (Kurth, Krahn, & Nairn, 1995). Thirty-five percent of “normal dieters” progressto pathological dieting. Of those, 20-25% progress to partial or full-syndromeeating disorders (Shisslak & Crago, 1995). Because of the secretiveness andshame associated with eating disorders, many cases are probably not reported.In addition, many individuals struggle with body dissatisfaction andsub-clinical disordered eating attitudes and behaviors. For example, 80% ofAmerican women are dissatisfied with their appearance which is indicative of acultural condition (Smolak, 1996).

Effectivenessof Current Treatment

Despite the increasingprevalence of eating disorders worldwide, the treatments in mainstream practiceare not highly effective (Steinhausen, 1995; Eckert, Halmi, Marchi, Grove,& Crosby, 1995). Agras (1993) reportsthat only 32% of all patients with eating disorders who have had treatment willbe eating disorder free after one year. Over time, “40% of anorexics recover, one-third improve, and 20% have achronic course” (Steinhausen, 1995, p. 235). The statistics are somewhat better for bulimics with approximately 15%completely recovering after year 1 and 67% after year 8 (Fallon, Walsh, Sadik,Saoud, & Lukasik,1991). HoweverFairburn (1985) states that cognitive therapy for bulimia nervosa, which isprobably the most extensively studied form of treatment, is associated withonly a 40% remission rate. Thisinformation reminds us that recovery from an eating disorder is a process, notan event. It is not like an addictionsuch as alcoholism where the addict simply does not pick up a drink. An eating disorder patient, because of thecomplexity of the disorder, progresses in recovery . 

Cost of Treatment

Many patients need repeatedhospitalizations. Treatment for eating disorders is extremely expensive, sinceeating disorders tend to take such a prolonged and chronic course, and havevery high rates of comorbid medical and psychiatric disease in all thedisorders. Large numbers of victimsrequire extensive medical monitoring and treatment, and therapy generallyextends over two years to ten years or more. The cost of inpatient treatmentcan be $30,000 – $60,000 or more a month. The cost of outpatient treatment, including therapy and medicalmonitoring, can extend to well over $100,000 or more (Striegel-Moore, Leslie, & Petrill, 2000).

Eating Disorder Research

Overthe last 30 years an impressive amount of research has been done on the causes,diagnoses, etiology, outcomes of various treatments for eating disorders aswell as predictors of outcome. However, research in the eating disorder fieldhas been relatively problematic. Researchers have had problems when they have attempted to compareresults from different clinical outcome studies (Agras, 1993). There exists a disparity in assessmentmethods as well as in the number of theoretical concepts about eatingdisorders, which generates very different clinical approaches, resulting indifferent clinical results (Eckert et al., 1995). As a result of the disparity in assessmentand clinical methods, research results have been inconsistent. There are numerous conflicting studies ofetiology, predictors, treatment, and outcome. 

Etiological FactorsIncluding Stress Response

Etiologicalfactors and events that precipitate eating disorders are widely varied but theyall can be said to involve elements of environmental stress. A high frequency of environmental stressorseither in the form of life events or daily hassles (including etiologicalfactors and precipitating events) can lead a person to resort to ineffectivecoping mechanisms, which in turn can lead to eating disordered behavior as away of coping in response to stress (Shatford & Evans, 1986).

Overall,the eating disorder spectrum includes anorexia nervosa, bulimia nervosa, pica,rumination disorder of infancy, and eating disorder not otherwise specified(DSM IV, American Psychiatric Association, 1994). Pica is commonly seen inchildren with low socio-economic status and/or developmental delays. Ruminationdisorder is characterized by repeated regurgitation and significant weight lossin infancy. These two eating disorders are commonly seen in early childhood andare not related to anorexia nervosa, bulimia nervosa, eating disorder nototherwise specified or binge-eating disorder, the primary emphases of thisstudy.

Theprimary concern of people with anorexia or bulimia is fear of gaining weight,while the term anorexia literally means ‘loss of appetite’ (Williamson, Barker& Norris, 1994). The probleminherent in anorexia does not result from appetite loss, but fromself-starvation. Bulimics, in contrast, tend to maintain normal weight, but doit by binge eating, followed by inappropriate and often harmful compensatorybehaviors, such as laxatives, vomiting, diuretics, and excessive exercise(Wilson & Fairburn, 1993).

Individuals whobinge-eat without compensatory behaviors tend to eat large amounts of foodwithout feeling hungry, which results in marked weight gain. It is thereforeimportant that one understand the underlying reasons that impel the person tomake the choice to engage in such extreme behaviors, which are largely theovert result of the perception of ‘fatness’ (Wilson & Fairburn, 1993).Obesity, another eating problem, is not considered an eating disorder,according to the DSM-IV, since it has been shown to be influenced by genetic,psychosocial and nutritional factors (Williamson, Barker & Norris,1994). When Robert Spitzer prepared theDiagnostic and Statistical Manual III (DSM III, 1980), he concluded thatobesity should be classified as a somatic disorder and not a psychiatric diagnosis;hence, obesity is not included in the DSM. However, Kelly Brownell (1995) states that with the focus on medicalrisk, other important facets of a person’s life are ignored with the resultbeing nearly all treatment programs for obesity do not attempt to assess orintervene with psychological and social issues. This classification is unfortunate because there are many clinicians whobelieve that many people who are obese have eating disorders.

Etiology of EatingDisorders

PsychologicalCharacteristics

Certain psychologicalcharacteristics are more prevalent in persons with eating disorders, includingimpulsivity, feelings of ineffectiveness, manipulativeness, alienation, chroniclow self-esteem, perfectionism, and over-compliance (particularly in anorexianervosa) (Herzog, Nussbaum & Marmor, 1996; Wilson & Fairburn, 1993).Two recent studies by Wilfley (2002) and Johnson (2002) found an increased riskin eating disordered patients for anxiety disorders, cardiovascular symptoms,chronic fatigue and pain, depressive disorders, infectious diseases, insomnia,neurological symptoms, and suicide attempts during early adulthood. These symptoms have likewise been implicatedas potential risk factors for eating disorders (Pryor & Wiederman, 1996).

Familial Factors

Eatingdisorders occur more frequently in biological relatives of patients who presentwith the disorder. Cooper (1995) states that firstdegree relatives of a person with an eating disorder are three times morelikely to develop an eating disorder in their lifetime. Evidence from familystudies shows that there is a significantly greater occurrence of affectivedisorders and that there is a high rate of substance abuse as well (Cooper,1995).

Severalfamily studies have examined psychiatric conditions that co-occur with anorexianervosa (AN), bulimia nervosa (BN), or binge-eating disorder (BED). With regardto major mood disorders, studies of anorexic patients have found that there isa risk in the range of 7% to 25% that a family member will either have ordevelop a mood disorder in his or her lifetime. In control groups this risk isonly a range from 2.1 to 3.4% (Kaye, 2000). Studies of bulimic patients haveshown that their first degree relatives are several times more likely to havemood disorders than are relatives of control groups (Kaye, 2000).

Some studieshave also found a familial connection between substance abuse and eatingdisorders. Researchers have reported an increased risk for alcohol abuse amongfirst-degree as well as second-degree relatives of bulimics (Bulik, 1987;Hudson, Pope, & Jonas, 1987; Hudson, Pope, Jonas, & Yurgelun-Todd,1983). On the other hand, Kaye (2000) states that recent studies have shownthat there is little correlation between a genetic susceptibility to alcoholismand the genetic factors in influencing susceptibility to BN. Kaye, Klump, Frank & Strober (2000) alsostate that there is new evidence of independent genetic transmission ofobsessive-compulsive disorders, anorexia and bulimia. Kaye et al. (2000) havefound preliminary data that point to a connection between AN andobsessive-compulsive personality disorder which combines rigid perfectionismwith asceticism.

Developmental Factors

In addition to familial connections, attachment processes andtheir role in the development of psychological disturbances underlyingdysfunctions of eating behavior have assumed an important place in the study ofeating disorders. The attachment relationship, according to attachment theory,is an interplay between the baby’s attempts to stay close to the primarycaregiver (initially the mother) in times of threat and the ability of theprimary caregiver to respond to the infant’s needs.

Attachmenttheory proposes that secure attachment occurs when babies develop theexpectation that the parent will be available, will meet their needsappropriately, and let them know they are cared for (O’Kearney, 1996). Insecureattachment happens when the expectation that the parent is available is notrealized and personal self-worth does not develop because of perceivedinaccessibility or inappropriate responsiveness of the caregiver (Bruch, 1973).Hilde Bruch describes abnormal attachment patterns quite simply. Abnormalattachment involves “a mother who superimposes on her infant daughter her ownconcept of the infant’s needs such that the infant’s needs and impulses remainpoorly differentiated” (Bruch, 1973, p. 7). This results in a sense ofseparateness and ineffectiveness that underlies the development of the eatingdisorder.

Social Influences

Understanding the natureof social influence on the individual has been important from the beginning ofthe study of group processes. Sherif (1935) and Asch (1951, 1956) (as cited inForsyth, 1983) conducted conformity experiments to begin to understand theinfluence of society on individuals. Research in attitude formation and group process can contribute to anunderstanding of socially mediated effects in eating disorders.

Power Struggles inRelationships

One tradition which grewfrom the early conformity work of Sherif and Asch was the types of power inrelationship roles (Eagley & Chaiken, 1993). Power struggles in eatingdisordered clients are well known and appear to be evidence of a replicatedearly power-struggle between the client and parents (Conrad, Rutgers, &Camden, 1992).

Conformity

A second tradition thatarose from the social influence literature has grown from group research, whichhas centered mainly on the cognitive mediation of conformity. Especially foreating disorders, research on minority influence, such as super models ortelevision stars, has been especially salient, indicating that minorityinfluence exercises discernable persuasive power. Harrison(1997) found a strong relationship between interpersonal attraction to thin mediapersonalities and disordered eating in a sample of 232 college students.

Cultural Influences andPressures

Much hasbeen written in magazines, journals, and books on the current culturalpressures on women to be thin (Polivy & Herman, 1993). Since the 1960s,western society has placed increasing demands on women to be thin. Being thinis believed to symbolize competence, success, control, beauty, and sexualattractiveness. Today, being thin and physically fit has become a salientcultural ideal. The mass media has been relentless in portraying this image towomen via magazines and television. The idealization of thinness has beenaccompanied by a high degree of body-image dissatisfaction among women of allages (Smolak, 1996). For many, self-worth has become tied to thinness. Thus,the pressure on women to be thin has contributed to the prevalence of dietingin American society. By fourth grade, nearly 80% of all girls have already beenon a diet. Chronic dieting has been cited as a contributing factor in thedevelopment of eating disorders, particularly binge eating disorder and bulimianervosa (Polivy & Herman, 1993).

Social Influence

A third line of researchis on social influence mediated by group polarization, which is the tendencyfor the decisions and attitudes that groups produce to be more extreme thanthose produced if the members acted alone (Eagley & Chaiken, 1993). Manyeating disorders may begin with a normal attempt to diet in order to loseweight, but for young women who may be predisposed to psychological disorders,such as low self-esteem, interpersonal distrust and maturity fears, dieting canexaggerate psychological dysfunction and precipitate an eating disorder (Polivy& Herman, 1993). There is a negative correlation between self-esteem andchronic dieting and between self-esteem and body dissatisfaction (Heatherton& Polivy, 1992). A person may experience chronic dissatisfaction with lifeand as a result be psychologically triggered to begin to either restrict or tobinge and/or purge (Heatherton & Polivy, 1992).

In summary, preconditionsfor an eating disorder include sociocultural influences (such as the culture ofthinness), familial influences, biologic and genetic factors, personality andpsychological variables (such as low self-esteem) and chronic behavioral patterns(such as dieting), that have developed in response to cultural and individualpredispositions (Wilson & Fairburn, 1993). The preconditions set the stagefor a possible precipitating event that will lead a high-risk person to developa clinical eating disorder.

Anorexia Nervosa

Themost prominent feature of anorexia nervosa (AN) is an intense fear of “fatness”(Williamson, Cubic & Gleaves, 1993). It has a 5% to 25% lifetime mortality rate in some studies (Katz,2003). It is characterized by majorbehavioral, psychological and physiological problems, including a body weightthat is less than 85% of normal weight, fear of gaining weight even thoughunderweight, preoccupation with food and abnormal consummatory behavior(Williamson, Barker & Norris, 1994; DSM-IV, 1994).

Individualswith AN view themselves as overweight even though they are dangerously thin.The process of eating in and of itself becomes an obsession. Compulsive eatinghabits predominate, such as avoiding food and meals, picking out a few foodsand eating these in small portions, or carefully weighing food. Anorectics mayrepeatedly check their body weight, and many exercise careful control over weightloss via other techniques, such as intense exercise, or purging by means ofvomiting and abuse of laxatives, enemas, and diuretics.

Uponphysical examination, individuals with AN exhibit decreased heart rate, bloodpressure and overall metabolic rate, as well as increased serum cortisol. Theyalso have profound decreases in estrogen production in females or testosteronein males (Soykan & McCallum, 1998). Over two decades of research, clinicaloutcome studies indicate that this problem is likely not a result of any singlebiological or psychological factor, but instead appears to be a product of aninteraction of biological, psychological and social influences (Cash &Deagle, 1997; Kipman, Gorwood, Mouren-Simeoni, & Ades, 1999; Wonderlich, Brewerton,Jocic & Dansky, 1997).

Someindividuals with AN recover after a single episode, a few have a fluctuatingpattern of weight gain and relapse, while many others experience a chronicallydeteriorating course of illness over many years. The mortality rate amongpeople with anorexia has been estimated at 0.5 percent per year, which is about12 times higher than the annual death rate due to all causes of death amongfemales ages 15-24 in the general population. The most common causes of deathare complications of the disorder, such as cardiac arrest, electrolyteimbalance, or suicide (Nielsen, 2001).

Table A1
Summary of Diagnostic Criteria for Anorexia Nervosa, DSM IV

1. Refusal to maintain minimal body weight at or above a minimally normal weight

2. Intense fear of weight gain, even though underweight (< 85% of normal weight)

3. Body-image disturbance, including denial of the seriousness of the current low body weight

4. Amenorrhea of 3 consecutive cycles

5. Restricting type: During current episode person does not engage in binge eating or purging.

6. Binge eating/purging type: During episode person engages in binge eating and purging

Manyproblems exist in the current DSM diagnostic system for AN. Amenorrhea has beencriticized as part of the diagnostic system, partly because it may not exist inpersons with AN (Garfinkel, Lin, Goering & Spegg, 1996), but mostly becauseit exists, when it does exist, embedded within a spectrum of physiologicalproblems which result from prolonged starvation (Stoving, Hangaard, Hansen-Nord& Hagen, 1999). These medical problems are severe andlife-threatening.

Table A2
Medical Findings Associated with Anorexia Nervosa*

Increased lanugo hair (fine body hair)

Acrocyanosis (decrease in the amount of oxygen delivered to the extremities)

Deceased subcutaneous fat

Dry skin (often scaly)

Decreased core body temperature

Heart rate below 60 BPM

Amenorrhea

Arrhythmias

ECG abnormalities

Congestive heart failure

Constipation

Crampy abdominal pain

Acute pancreatitis

Peripheral edema

Anemia

*from Soykan & McCallum, 1998.

 
Prevalence

Currentevidence estimates the prevalence of AN at approximately 0.28% (Crisp,Callender & Halek, 1992). For females only, AN has a lifetime prevalence ofabout 1% but appears to be higher in specialized populations, e.g., AN (andbulimia) is reported to be more frequent in dancers than in the general public(Stice, Telch, & Rizvi, 2000). Onestudy found a 6.5% rate of AN in dancers (Garner & Garfinkel, 1980). In asurvey conducted at UCLA using the Eating Disorders Inventory, 716 femalestudents were interviewed to determine prevalence data for different subgroups(Kurtzman, Yager, Landsverk, Wiesmeier, & Bodurka, 1989),. The meanprevalence for the entire group was 2.1% with a lifetime prevalence of 4.8%. Ina study by Szweda & Thorne (2002), other at-risk subgroups include femalehealthcare students (20% prevalence rate) and arts students (21%).

Etiology of Anorexia Nervosa

Explanationsof the development and maintenance of anorexia nervosa center aroundpsychological, biological and psychodynamic perspectives with sociologicalinfluences.

Cognitive-Behavioral Viewpoint

Oneof the most striking characteristics of anorexia nervosa is the intensity andimportance of the patients’ dysfunctional beliefs and values concerning theirweight and shape. The patients’ beliefsand values can be viewed and understood in cognitive terms. AN may be seen as a behavioral coping skill.Like all coping skills, it develops out of attempts to deal with life events.The maladaptive coping skills that develop may include fears of maturity and autonomy (separating fromparents, having a boyfriend, sexuality); feelings of ineffectiveness,helplessness, and poor self-esteem; and often disturbed relationships withpeople close to the individual. The anorectic becomes preoccupied with food andweight to distract the self from overwhelming feelings of anxiety, fear, anddepression, triggered by these events. Then, the habit of food restriction andrituals become so entrenched that this set of behaviors can become split offfrom their original stimuli and exist in and of themselves and maintainthemselves as functionally autonomous behaviors (Kleinfield, Wagner &Halmi, 1996).

Thedistorted beliefs, values, and behaviors or cognitive distortions of thepatient are more than just symptoms and can assume primary importance in themaintenance of the condition. Aprerequisite for full recovery is to change these cognitive distortions. For this reason, Fairburn (1981) developed acognitive-behavioral model of treatment for anorexia nervosa and bulimianervosa. This model utilizes behavioral interventions and formal cognitiverestructuring.

Wilfley,Pike, & Streigel-Moore (1997) elaborated on the cognitive-behavioral modeland postulated their own extension of the cognitive-behavioral model thatintegrates sociocultural and biological influences which may lead to theinitiation and maintenance of AN habits. They proposed that predisposingfactors were genetic and nutritional factors, affective disorders and familialproblems, along with personality variables. Their model includes obesity and binge eating. In their model, theweight gain from binge eating leads to the perception of being fat, and AN maydevelop out of extreme weight control behaviors in response to the fatperception. One prediction from the model is that AN is a weight phobia.

Biological Theories

Thehigh incidence of eating disorders in families has led researchers tohypothesize the existence of a genetic influence in the etiology of eatingdisorders. Studies have shown that 58% to 76% of the variance in AN can beattributed to genetic factors (Klump, 2000; Lilenfeld, Kaye, Greeno,Merikangas, & Plotnicov, 1998). The prevalence rate of AN in sisters ofpatients is estimated at 6%. Genetic evidence in the etiology of AN comes fromtwin studies, in which the twins are raised apart. These studies reported substantially higherconcordance rates for monozygotic than for dizygotic twin pairs (Strober,1991). While these data are suggestive, conclusive data for genetictransmission of the disorder are not yet available.

Klumpet al. (2001) identified, in a literature review, shared geneticsusceptibilities to major depression, substance abuse, anxiety disorders andpersonality disorders. Similarly, there is growing evidence to suggest that AN(and BN) likely share the same genetic risk factors. Family and twin studiesindicate that relatives are at increased risk for AN and BN because of a sharedfamilial component between the two disorders. It appears that there may be afamilial phenotype with possible genetic predispositions.

Outcomestudies on anorexia nervosa have been available in the literature for at leastfour decades, from the 1950’s. Steinhausen (1995) has published a review of 68 outcome studiespublished between 1953 and 1989 concerning a total of 3,104 patients. Thefollow-up results show a large variation in the main features of anorexia –weight, menstruation, and eating behavior. Steinhausen (1995) found the following:

The means indicate that in close to 60%of anorexics weight and menstruation may normalize, whereas somewhat fewer(49%) show normalized eating behavior. The distinction between recovery,improvement, and chronicity (which to a great extent is based on a popular butcrude differentiation between good fair, and poor outcome) shows that, on theaverage, more than 40% of anorexics recover, one-third improve, and 20% have achronic course. (p. 235).

Eckert et al. (1995) published a 10-year follow-up studyof 76 severely ill females with anorexia nervosa. Five patients had died, whichshowed a mortality rate of 6.6%. Only 18 (23.7%) were fully recovered. Sixtyfour per cent developed binge-eating at some time during the course of theirillness, 57% at least weekly. A little over 25% of the patients studied had agood outcome as they were able to re-establish normal weight and regularmenstrual cycles, even though they still presented with abnormal eating andweight control behaviors and/or significant body image disturbance. Forty-fivepercent were unable to establish normal weight and menstrual cycles, nor tonormalize other anorexic symptoms (Eckert et al., 1995).

Psychosocial Influences

Theincreased prevalence of anorexia nervosa has been connected to the moderncultural values for females to be thin, and it appears that this emphasis onslimness has increased significantly during the past several decades(Martinez-Gonzalez, Gual, Lahortiza, Alonso, Esteves & Cervera, 2003).These influences have been previously discussed in the introduction of thispaper.

Developmental Factors

Becauseanorexia nervosa typically begins during adolescence, developmental issues arethought to play an important etiological role (Lewinsohn, Striegel-Moore &Seeley, 2001). Kotler, Cohen, Davies, Pine, and Walsh (2001) found early eatingconflicts such as struggles with food and unpleasant memories surroundingmeals. Most authors agree that the challenges during adolescence include theneed to establish independence, a well-defined personal identity, fulfillingrelationships, and clear values and principles to govern one’s life (Kail, 2002). For adolescents, it is typicaland normal to have problems with family and peers. However, it is not clearwhat is normal for adolescents with eating disorders. Disturbed parental-child relationships arereported in the literature (Kriepe & Birndorf, 2000). Eating disorders may mirror certaindysfunctional roles, conflicts, or interactional patterns within a family(Minuchin, Rosman, & Baker, 1978).

Depression hasbeen implicated as a nonspecific risk factor, and higher levels of depressivesymptoms as well as anxiety and self-consciousness are seen in adolescents withAN (Kriepe & Birndorf, 2000). Anumber of psychiatric traits are commonly found in patients with AN and BN,including mood disorders. Many of thesetraits are developed prior to the development of the eating disorder and manydeveloped secondary to malnutrition. Studies have shown that there is a risk of 7%-25% for developing majordepression in families of eating disorder patients (Klump, 2000). Anxiety disorders, personality disorders, andsubstance abuse are also common psychiatric conditions in the eating disorderpopulation.

BulimiaNervosa

Thediagnostic criteria for bulimia nervosa (BN) have undergone little change sinceRussell’s initial description of the condition in 1979 (Russell, 1979). “Thequality of Russell’s initial clinical description (of bulimia nervosa) isapparent from the fact that subsequent changes in it have been modest”(Stunkard, 1997, p. 182). The most outstanding feature of bulimia nervosa isthe frequent occurrence of binge eating occurrences.

Episodesof binge-eating are associated with feelings of loss of control. Once begun,the individual feels unable to stop eating until an excessive amount of foodhas been consumed. This loss of control is only subjective, since mostindividuals with bulimia nervosa will abruptly stop eating in the midst of abinge episode if interrupted.

Afterovereating, individuals with bulimia nervosa engage in some form of purging orother extreme behavior in an attempt to avoid weight gain. Most patients withthis syndrome report self-induced vomiting or the abuse of laxatives. Othermethods include misusing diuretics, fasting for long periods, and exercisingextensively after binge eating. It is important to note that complications fromBN (and AN) can potentially affect every organ system.

Thecurrent DSM-IV criteria for bulimia nervosa are in Table A3.

Table A3
Summary of Diagnostic Criteria for Bulimia Nervosa, DSM IV

Recurrent episodes of binge eating

Recurrent inappropriate compensatory behavior in order to prevent weight gain, such as self-induced vomiting; misuse of laxatives, diuretics, enemas, or other medications; fasting; or excessive exercise.

The binge eating and inappropriate compensatory behaviors both occur, on average, at least twice a week for 3 months

Self-evaluation is unduly influenced by body shape and weight

The disturbance does not occur exclusively during episodes of Anorexia Nervosa.

Someresearchers believe that AN and bulimia nervosa (BN) are simply twomanifestations of the same disorder (Williamson, Barker & Norris, 1994).However, individuals with BN are usually within normal weight limits and arenot as restrictive in caloric intake. The problem appears to begin later inlife than AN, and may co-occur with a history of AN.

InDSM-IV, a subtype was introduced for classification into either the purging orthe nonpurging type of BN. This scheme was introduced for several reasons.First, those individuals who purge are at greater risk for the development offluid and electrolyte disturbances such as hypokalemia and are at greater riskfor psychiatric illness. Most of thepublished literature on the treatment of bulimia nervosa has been based onstudies of individuals with the purging type of this disorder (Powers &Santana, 2002).

Aswith AN, BN has a myriad of medical problems associated with the disease.However, unlike AN, bulimics see their problems as out of control and will tendto try to stop (vs. trying to hide it as in AN). The psychological problems aresimilar to that of AN; problems with families and expression of feelings, lossof control, low self- esteem and depression (Williamson, Barker & Norris,1994). Physical findings with patients of BN are important in diagnosis and intreatment, as with AN.

Table A4
Medical Findings Associated with Bulimia Nervosa *

Bruised or abraded knuckles (Russell sign)

Facial Ecchymoses

Conjunctival hemorrhages

Pharyngitis

Dental enamel erosions

Menstrual irregularities

Esophagitis

Esophageal erosions and ulcerations

Esophageal rupture

Delayed gastric emptying

Crampy abdominal pain

Acute gastric dilation (with binge)

Constipation

Pancreatitis

Hypokalemia

Hypomagnesemia

Hyponatremia

Brachycardia

Low voltage, prolonged QT interval

Parotid and salivary gland swelling

*fromhttp://home.mdconsult.com/das/book/body/165772355/959/362.html

Prevalence

Approximately10% of women in western countries will be diagnosed with an eating disorder atsome point in their lives, making it one of the more prevalent psychiatricproblems faced by women (APA, 1994). Thelifetime prevalence for BN is approximately 2% for women (Stice, Telch, &Rizvi, 2000). Current estimated lifetime female risk for bulimia is about 8%(Kendler, MacLean, Neale & Kessler, 1991). However, like AN, there arestudies that show sharply increased prevalence rates for special populations,such as college students (4.5% to 18%) (Turnbull, Ward & Treasure, 1996) or(19%). The reports of relapse rates are mixed. Onestudy reported a 63% probability of relapse after 78 weeks (Keller, Herzog,Lavori & Bradburn, 1992).

Themorbidity and mortality associated with BN are significant. While AN has thehighest premature mortality rate of any psychiatric disorder, the prematuremortality rate of bulimia nervosa comes in second with a rate of about 5% at 10years. Among survivors, morbidity is significant. Furthermore, many of thesepatients have co-morbid psychiatric disorders including obsessive-compulsivedisorder, mood disorders, and substance use disorders. Preliminary studies ofbulimia nervosa have found that 5 to 10 years after initial presentation, 20%continue to meet full diagnostic criteria (Keel, 1997).

Etiology of Bulimia Nervosa

BNnormally begins later in life than AN (around ages 17 to 25) and may occur inindividuals with a history of AN (Williamson, Barker & Norris, 1994).Williamson et al. (1990) have proposed a model, which describes behavioral,cognitive and emotional factors in BN. In this model, overeating causes extremeanxiety, which is relieved by purging. The purging ‘undoes’ the effects of thebinging, much like compulsive rituals of handwashing, and is therefore,negatively reinforced in the same manner as compulsive rituals. The cycle thenbecomes habitual. However, many other studies have found significantpsychopathology in BN subjects and their families. Bulimic patients tend tocome from families with conflict resolution problems and tend to experience lowcontrol within the family system (Herzog, Keller, Lavori & Gray, 1992).Fairburn et al. (1993) found that residual attitudinal disturbance,particularly in regard to body image, predicted treatment outcome. They foundthat, the greater the disturbance, the more dismal the outcome and a clearlinear relationship existed between the two.

Biological Factors

Evidenceof underlying biological problems is, to date, mostly circumstantial butcompelling, and tends to correlate with studies of AN. Twin studies of eatingdisorder attitudes, such as body dissatisfaction, weight concerns and body andweight preoccupation, indicate that between 32% and 72% of the variance inthese attitudes can be accounted for by genetic factors (Rutherford, McGuffin,& Katz, 1993). However, outcome data for these studies is compromised by comparisonof results across studies with different definitions of what constitutesoutcome, diagnostic criteria, proper duration for follow-up, assessmentmethods, etc. Bulimics do tend to recover over time, however (Herzog, Nussbaum& Marmor, 1996).

Psychosocial Influences

Asin AN, BN appears to be partly a socially mediated phenomenon. In a studyconducted in Norway,Wichstrom (1995) used Logistic Regression to identify predictors of eatingbehavior in adolescent girls and found that constructs such as bodydissatisfaction, adopting idols with perfect bodies, low self-worth, strongfeminine sex-role orientation, permissive parenting and early puberty wereassociated with onset of eating problems, especially BN. In a more recentstudy, Martinez-Gonzalez, Gual, Lahortiga, Alonso, Irala-Estevez, and Cervera(2003) found a positive relationship between mass media influences and alleating disorders. As in AN and all other eating disorders, cultural influencestend to underlie the pathogenesis of the problem (Mason & Chaney, 1996).

Developmental Issues

Ithas been suggested that childhood sexual abuse is a specific risk factor forthe development of bulimia nervosa and is much more common in bulimics than inrestricting anorexics (Fallon & Wonderlich, 1997). Scientific support for this hypothesis isweak. The best studies to date have found that compared with women withoutpsychiatric illness, women with bulimia nervosa do, indeed, report increasedfrequencies of sexual abuse. In a questionnaire study with binge eaters andnon-binge eaters, Fallon & Wonderlich (1997) found that binge eaters, whencompared with non-binge eaters, reported higher rates of sexual abuse.

Lacey,Coker, and Birchnell (1986) reported that 72% of bulimics in their study hadsexual conflicts such as being raped, feeling pressured into sex, or othertypes of sexual abuse. However, therates of abuse are similar to those found in other psychiatric disorders(Butterfield & Becker, 2002). Thus,whereas early abuse may predispose an individual to psychiatric problemsgenerally, it does not appear to lead specifically to an eating disorder, andmost patients with bulimia nervosa do not have histories of sexual abuse.

Binge-Eating Disorder

Lessis known about the course of binge-eating disorder (BED) when compared with ANand BN. BED has roughly the same clinical picture of psychopathology as BN.However, the patient is at increased risk for morbidity and mortality becauseof the problems associated with being overweight and obese. Many of thepatients are or become morbidly obese (Heatherton, Mahamedi, Streipe, Field& Keel, 1997).

Fairburn,Doll, Welch, Hay, Davies, & O’Connor (1998) identified some risk factorsassociated with BED, including certain aversive childhood experiences, parentaldepression, vulnerability to obesity and repeated exposure to negative commentssurrounding shape and weight. Compared to other psychiatric disorders, thosewith BED experience significantly more childhood obesity and more exposure tonegative comments about weight and overeating. 

Table A5
Summary of Diagnostic Criteria for Binge-Eating Disorder, DSM IV

Recurrent episodes of binge eating

The binge-eating episodes are associated with three or more of the following:

a. Eating much more rapidly than usual

b. Eating until feeling uncomfortably full

c. Eating large amounts of food when not feeling physically hungry

d. Eating alone because of being embarrassed by how much one is eating

e. Feeling disgusted with oneself, depressed, or very guilty after overeating

f. Marked distress regarding binge eating is present

g. The binge-eating occurs at least 2 days per week for 6 months

h. The binge-eating is not associated with the regular use of inappropriate compensatory behaviors (e.g., purging, fasting, excessive exercise) and does not occur exclusively during the course of Anorexia Nervosa or Bulimia Nervosa.

Prevalence

The prevalence of binge eating disorder is still beingdetermined. Some researchers estimatethat approximately 25% of obese individuals have episodes of binge eating (http://www.pbs.org/perfectillusions/eatingdisorders/bingeeating.html). Estimates indicate that about 60% of peoplewith binge eating disorder are women and 40% male. People who struggle with binge eatingdisorder may be obese, heavier than average weight, or normal weight.

Mostparticipants who participate in clinical trials are overweight. BED has been reported in 30% of participantsin weight loss programs, 70% of member of Overeaters Anonymous, and 50% ofthose who seek bariatric surgery (Fairburn & Harrison, 2003; deZwaan,2001).

PsychiatricComorbidity

Patients with binge eating disorder frequently presentwith mood disorders, substance abuse, anxiety, impulse control difficulties,personality disorders, and particularly depression (Agras, 2001).

Theoriesof BED

The causes of BED are unknown, but biological, familial,and psychosocial factors are probable factors. Genetic factors are being studied by researchers. Some of the familial factors include parentaldepression and obesity (Fairburn, Doll, Welch, 2003). Psychosocial factors may include physicaland sexual abuse, bullying by peers, and discrimination because of beingoverweight (Striegel-Moore, Dohm, Pike, 2002).

Asexpected, BED is related to obesity. In one sample of 19 surgical candidates,with a Body Mass Index (BMI) range of 35 to 58, a strong relationship was foundbetween BMI and questions on the Binge Eating Disorders questionnaire (Kral,2001). BED affects about 2% of the general population and about 30% of obesepatients in medical treatment, with about 60% of those patients being female(Powers & Santana, 2002). Binge eating is also associated with depressionand low self-esteem (de Zwaan, 1994).

Peoplewith Binge Eating Disorder eat an amount of food that is definitely larger thanmost people would eat during a similar period of time in similarcircumstances. BED is also characterizedby loss of control as well as eating more rapidly than normal, eating untilfeeling uncomfortably full, eating large amounts of food when not hungry,feeling distressed and guilty after binge eating, and eating alone because ofbeing embarrassed by how much one is eating (Fairburn, 1993). People who are obese without BED do not havethese characteristics.

Inthe 1970s and 1980s the discovery that body weight is determined by neural,hormonal, and metabolic factors, and that genetics play a role in determiningbody weight and shape led to the belief that weight control is extremelycomplex. At this point medicine emerged asthe leading force in treating obesity (Brownell, 1995). Very-low-calorie diets and surgery wereused. The 1990s have seen the birth of astrong anti-dieting movement which claims that dieting is never effective andmay cause more difficulties than it solves (Brownell, 1995). Current thought is that pharmacotherapy willbe able to offer individuals an effective way to lose weight. A complete discussion of psychopharmacology isbeyond the scope of this paper and can be considered a delimitation of the research.

Itis interesting to note that obesity research is funded by institutes orientedtoward physical diseases, such as the National Institute of Diabetes andDigestive and Kidney Diseases (Brownell, 1995). There are some consequences of the focus on medical risk. In my view it is virtually impossible for aperson who is obese not to have any psychological consequences even if it isjust the discrimination against obese people that is inherent in our society.

EatingDisorder Not Otherwise Specified

Almosthalf the patients who eventually become identified as having eating disordersdo not meet strict criteria for AN or BN. Some of these patients have BED, andothers do not meet the criteria for any of these syndromes. For example,patients who continue to have menses but meet all the other criteria foranorexia nervosa would be classified as having EDNOS.

Walshand Garner (1997) believe that there is still uncertainty in regards to thesubtyping found in clinical diagnosis. In the DSM-IV the frequency of bingingand purging is not specified, nor is the exact definition of binge eating givenfor a person who is anorexic and severely underweight. There is also no cleardistinction between those anorexic individuals who binge and do not purge andthose who purge but do not binge.

Adifficulty for clinicians is determining the boundary between anorexia nervosaand bulimia nervosa. For example, if a client is an anorexic who also bingesand purges, gains weight, begins to menstruate, but continues to binge andpurge, is she now partially recovered from anorexia nervosa or does she nowhave the diagnosis of bulimia nervosa? The difference between the twodefinitions is ambiguous. The DSM-IV gives guidelines,

but the finaldetermination is left to the clinician. It is important to remember thatdiagnosis and criteria for the respective disorders are independent of theetiologies of eating disorders.

Individualswith EDNOS can be seriously ill and can require immediate clinicalintervention. Some clinicians may make the mistake of minimizing theseriousness of the diagnosis. Eating Disorder NOS is a category for eatingdisorders which does not meet the criteria for any specific eating disorder.

Table A6
Summary of Diagnostic Criteria for Eating Disorder NOS, DSM IV

Eating Disorder, NOS is a criteria for disorders of eating that do not meet the criteria for specific eating disorders.

For females, all criteria for anorexia nervosa met with the exception of regular menses.

All criteria for anorexia nervosa are met with the exception that, despite significant weight loss, the individual’s weight remains within normal limits.

All of the criteria for bulimia nervosa are met except for the binge-eating episodes and compensatory behaviors occur less than twice a week, or for a duration of less than 3 months.

All criteria for bulimia nervosa met, with the exception that compensatory behavior occurs after eating only small amounts of food (i.e., two cookies).

Repeated chewing and spitting out, but not swallowing, of large amounts of food occur.

Schwitzer(2001) conducted a study among college women in an eating disorder program,examining eating-related behaviors, associated features and developmentalissues. The most common pattern found in this study was one of regular bingeeating, paired with daily exercise and occasional purging. The collegepersonnel exhibited pathological overconcern about body image. The women in thestudy also displayed developmental issues of perfectionism, longstandingdepression and problems in parental relationships including power struggleswith parents.

Treatment Outcome in the Eating Disorders

Inthis article, the literature on eating disorders, particularly etiology andepidemiology, have been summarized according to DSM-IV diagnostic criteria (anorexianervosa, bulimia nervosa, binge eating disorder and eating disorder NOS). As an extension of this discussion, we willdiscuss how treatment success has been defined and reported by variousresearchers and what exactly constitutes a successful clinical outcome fortreatment of these disorders. Examination of the factors that predict who willbenefit from treatment will likewise be discussed.

Treatmentsuccess has been defined and reported in different ways by differentresearchers. Some have emphasized total weight loss or weight gain, whileothers focused on psychological, medical, or psychosocial improvements withtreatment. Some researchers have hypothesized that a patient needs to becompletely symptom free in order to be deemed a treatment success. Otherresearchers (Norring & Sohlberg, 1993) state that outcome can be classifiedinto three categories: “ill (has an eating disorder diagnosis), improved (hasno eating disorder diagnosis but has substantial remnants of the weight and/orshape preoccupation), and well (has no eating disorder diagnosis or remnants ofthe weight and/or shape preoccupation)” (p. 438). Steinhausen, Rauss-Mason, andSeidel (1991), in a review of 68 outcome studies published from 1953 to 1989,describe outcome categories in terms of improved as well as completelyrecovered. Other earlier clinical studies equate success with being in”remission” from the “disease” (

The narrow definition that equates success with being in”remission” from the “disease,” comes from the notion that eating disorders areaddictions much like alcohol and drug abuse (Wilson, 1995). Wilson reports the similarities betweeneating disorders and alcohol and drug abuse: strong urges or “cravings,” sense of loss of control, using thesubstance to cope with stress, thinking of the substance many times throughoutthe day, repeated attempts not to “use,” denial, and the presence ofpsychological and social problems as a result of the disorder. Overeaters Anonymous, an organization modeledafter Alcoholics Anonymous, believes that certain overeaters should not eatcertain foods, such as white flour and any sugar. Overeaters Anonymous maintains that thesesubstances are chemicals that can cause “chemical dependence.”

Wilson(1995) states that there is little in the literature that supports theaddiction model. He states that the addiction model suggests that people mustpractice extreme dietary restraint, avoiding particular foods, and engaging inhighly structured eating patterns while admitting a sense of powerlessnesswhich reinforces dichotomous thinking pattern (Miller, 1996). The patients aretold that they are always recovering, that they will never be recovered fromthe “disease.”

Manyprominent treatment centers in the country, such as Sierra Tucson, RemudaRanch, and the Meadows subscribe to this model. On the other hand, othertreatment centers such as Mirasol and Monte Nido, subscribe to the theory thata person with an eating disorder can fully recover from the disorder. Thetreatment goals of Mirasol and Monte Nido attempt to eliminate the very ideasthat addiction model centers promote. Clients at these treatment centers learn how to eat normally. They learn to tell when they are hungry andwhen they are full by developing interoceptive or proprioceptiveawareness. They learn that they can eata wide variety of foods, and they focus on empowerment rather thanpowerlessness. The elimination of blackand white thinking is addressed as a therapeutic issue.

Cognitive-behavioraltherapy (CBT), the most often used and researched model of treatment, can helpreduce dietary rigidity, change a person’s attitudes about their body image,and alter the prevalent black and white thinking in eating disorder populations(Wilson, 1995). Cognitive behavioraltherapy has been extensively studied in controlled clinical studies with strongfindings of efficacy.

CBThas been the most studied form of treatment for eating disorders. However, treatment procedures, treatmentlength, and the populations treated may differ considerably between studiesthat supposedly are measuring the same thing. W. Stewart Agras (1993), one ofthe primary researchers in the eating disorder field, has studied theeffectiveness of CBT. Drop out rates canarrange from 0% to 34%. Of the number ofpeople that remain in treatment, only 20% to 76% of them can expect to remainabstinent. This results in a 40%recovery rate. Agras (1993) goes on tosay that the results of strictly behavior therapy and nutritional counselingare similar to CBT. In a study conductedby Fairburn, Jones, Peveler, Carr, Solomon, Burton, Hope & O’Connor (1991), CBT andbehavior therapy were found to be equally effective. CBT has been found to be more effective thanfocal psychotherapy in reducing anxiety, dysthymia, and self-esteem.

Thereis no empirical evidence to support the effectiveness of the 12-step treatmentapproach, only anecdotal reports. Twelve-step treatment does not encourage psychotherapy; yet cliniciansknow the need of addressing psychological issues which can be multitudinous ineating disordered clients. Twelve-stepapproaches are popular because there are many practitioners with experience inthese programs. There are also a numberof practitioners who find the twelve-step approach antithetical to the way theythink. Twelve-step meetings are freewhich can be a benefit for certain patients. The requisite anonymity in suchprograms serves as a deterrent for empirical research.

Short-term,medium-term, and long-term outcome studies give different results. Short-termstudies are considered to be a 2 year follow-up; medium-term are from years2-6; and long-term outcome studies are from years 7-12 (Herzog, Schellberg,& Deter, 1997). Eckert et al. (1995)state that whether the eating disorder is treated or not, there is greatvariability within each disorder. They state that the research is replete withthe lack of follow-up studies that avoid selection bias, small sample size,inaccurate or unclear diagnosis, short or unequal follow-up periods, andinaccurate methods of data collection. There has been such a great disparity inassessment methods between studies that comparison across studies is extremelydifficult and potentially misleading (Eckert et al., 1995). Eating disordersare extremely difficult conditions to treat, are persistent and thereforerequire long-term follow-up. While most researchers recommend a minimumobservation time of 4 to 5 years, several studies indicate that outcome changeswith even longer follow-up (Eckert et al., 1995; Fichter & Quadflieg, 1999;Strober et al., 1997).

Whileresearch in the eating disorder field has grown in the past two decades,problems appear when results from different clinical outcome studies arecompared (Agras, 1993). There is a wide variation in assessment methods andtheoretical concepts about eating disorders, which then, in turn, generate verydifferent clinical approaches, resulting in diverse clinical results (Eckert etal., 1995). As a result of the disparity in assessment and clinical methods,clinical research results have been inconsistent (Abraham, 2002).

Treatment Outcome Studies

Outcomesfor Treatment of Anorexia Nervosa

Fichterand Quadflieg (1997) published an elegant long term study on the six-year courseof anorexia nervosa. They assessed 103 DSM-IV AN patients at four points oftime: pre-treatment, post-treatment, 2-year follow-up, and 6-year follow-up.All patients were treated in a specialized eating disorder unit. The averageage upon admission was 24.9 +/- 6.7 years.

Patientsreceived a broad range of treatment modalities including nutritionalconsultations, behavioral analysis of eating behavior, a behavioral program toinduce weight gain through reinforcement, psychoeducation, cognitive-behavioraltherapy, training in emotional expression and social skills, training inself-responsibility and self-care, and relapse prevention. Therapists attemptedto facilitate weight gain by enhancing interoceptive awareness. The length oftime the patients had had the eating disorder before admission was 6.3 +/- 4.8years. At the end of the 6-year study, six of the 103 patients had died. Themortality rate in this study was 17 times greater than for women of the sameage in the general population. The authors stated that, with the exception ofthe high mortality rates, the treatment outcome was positive.

Twoyears after treatment, 45.5% had no major eating disorder and at 6-yearfollow-up 55.4% had no major eating disorder. Nearly half of these patientsstill displayed eating disordered symptoms. Of the 56 women that did not have aclinical eating disorder at 6-year follow-up, 17 still displayed well-definedsymptoms, 14 were in partial remission, 12 had residual symptoms, and only 13could be considered symptom free or completely recovered. Thus, 34.7% had agood outcome, 38.6% an intermediate outcome, and 20.8% experienced a pooroutcome, and 5.9% had died (Fichter & Quadflieg, 1997). The anorecticpatients in the study group showed a course of illness of “high persistence.”

Outcomesfor Treatment of Bulimia Nervosa

Researchershave found that the same treatments effective in bulimia nervosa are effectivewith binge eating disorder as well (Telch, Agras, Rossiter, Wilfley, &Kenardy, 1990). At one time, cognitive-behavioral therapy was considered to bethe gold standard in the treatment of bulimia and binge eating disorder. Agras(1993) investigated the current treatment of binge eating, posing two complexquestions. The first was whether some treatments were better than others orwhether all treatments were equally effective. The second question was whethertreatments provided specific or nonspecific effects and whether they worked bythe same or different processes (Agras, 1993).

Agrasemphasized the importance of learning which factors maintain binge eating andin which way a specific therapy influenced these factors. He discussed thevariation in assessment methods that made comparisons between and acrossstudies extremely difficult. He found that studies differed in diagnosticcriteria, outcome criteria, and length of follow-up intervals. Agras (1993)found that only 32 patients out of every 100 entering treatment emerged eatingdisorder free 1 year or later.

Fallon,Walsh, Sadik, Saoud, & Lukasik (1991) performed a 2- to 9-year follow-upstudy on 52 women with the diagnosis of bulimia nervosa according to theDSM-III-R. The patients were treated in an inpatient eating disorders unit at auniversity hospital, using behavioral methods to address eating behaviors,individual psychotherapy, pharmacotherapy (if indicated), group therapy, andfamily therapy. Of the 52 women, 46 women were interviewed, 1 had died, and 5could not be located. Of the 46 inpatient bulimic women, 39% had fullyrecovered and 41% were still symptomatic. The longer period of time afterdischarge, the greater percentage of patients recovered. At the end of yearone, 15% had recovered versus 67% at the end of year 8. They found that themean duration of recovery was more than two years.

Keel, Mitchell, Miller, Davis, & Crow (1999) found the sameresults as Fallon et al. (1991) in that the number of women who weresymptomatic for bulimia nervosa declined as the length of time post-treatmentincreases. Keel and his colleagues strictly defined eating disorder outcomewith both narrow and broad definitions. The narrow definition required asubject to be free of bulimic symptomotology for a period of six months, and inaddition, for the subject not to use body weight and shape as a way of judgingherself. The broad definition required the subject to be free of disorderedeating behaviors for at least eight weeks with no restrictions onself-evaluation of perceived weight and shape.

Womenwho had not met criteria for full remission but who no longer met DSM-IVcriteria were considered to be in partial remission. The researchers had agroup of 173 women, with a mean duration of follow-up of 11.5 +/- 1.9 years.

Thisis the longest follow-up period in the English language literature on bulimianervosa with the second largest sample size. At follow-up, one patient (0.6%)met full diagnostic criteria for anorexia nervosa and 19 women (11%) metclinical diagnostic criteria for bulimia nervosa. Thirty-two women (18.5%) metthe criteria for ED NOS, including one woman with Binge Eating Disorder. Whenresearchers reported the results according to the narrow definition ofremission, they found that 28.3% of the women were in partial remission atfollow-up and 41.6% were in full remission. When they reported the resultsaccording to the broad definition of remission, they found that 23.1% were inpartial remission, and 46.8% were in complete remission.

Five short-term studies in which patients were followedfor a year or more after treatment found that most patients do well at one yearafter treatment with at least 75% showing no bulimic symptoms (Hsu, 1995). Insix studies where patients have been followed for up to two years or more,about one-half of the patients were in remission at follow-up (Hsu, 1995). Hsuconcluded that 50% of bulimia nervosa patients were symptom free two to tenyears after intake, that about 20% of patients remained persistentlysymptomatic, and that the remainder (about 30%) have a course of illnesscharacterized by remissions and relapses or by persistent bulimic behavior thatdoes not meet clinical definition (Hsu, 1995).

In1997, Fichter and Quadflieg published the results of their 6-year long-termstudy of bulimia nervosa. They studied a group of 196 female patients whofulfilled the diagnostic criteria according to the DSM-IV for bulimia nervosa.The treatment they received focused on “biological, social, and psychologicalperspectives” (Fichter & Quadflieg, 1997, p. 362). They assessed thepatients pre-treatment, post-treatment, at 2-year follow-up and at 6-yearfollow-up. Treatment lasted for 95.5 +/- 43 days; upon admission, patientsranged 25.6 +/- 6.7 years of age; and patients had had bulimia nervosa for 8.1+/- 4.9 years. At treatment discharge, 47 patients (24.1%) were significantlyimproved, 77 (39.5%) were markedly improved, 60 (30.8%) were slightly improved,and nine (4.6%) were unchanged, one patient (0.5%) was somewhat worse, andanother one (0.5%) as decidedly worse (Fichter & Quadflieg, 1997).

Fichterand Quadflieg’s (1997) data show that patients substantially improve duringinpatient treatment, show a slight decline during the next 2 years and showfurther improvement until the last assessment at 6 years. They state that thelonger the follow-up time, the number of recovered patients increases andmortality increases (6-year mortality for bulimia nervosa is 1.1%).

The6-year outcome study based on the presence or absence of a clinical eatingdisorder showed a relatively favorable course in the treatment of bulimia nervosa.This study counted only the presence or absence of a full clinical diagnosis soit missed some of the sub-syndromal behaviors. The researchers presented the multidimensional data at four distinctpoints of time. The study showed that 59.9% of women participating in the studyhad a good outcome, 29.4% an intermediate outcome, 9.6% a poor outcome, and1.1% were deceased. “However, of the 133 females with no DSM-IV eating disorderat 6-year follow-up, 25 still showed marked symptoms, 26 had experienced nopartial remission (or were unchanged), 45 showed residual symptoms, and 37 wererated as usual self (or were eating disorder free)” (Fichter & Quadflieg,1997, p. 378).

Itis interesting to note that the above outcome studies do not remark upon whatthe clients were doing all of these years. Were they in outpatient treatment? Were they in individual psychotherapy for part of the time? Did they outgrow the disorder, indicating amaturation effect? More complete outcomestudies are needed to further refine outcome research.

Predictors of Treatment Outcome

Investigatorshave focused on a wide range of variables, differentiating those who have asuccessful treatment response from those who do not. Identifying predictivefactors can help with the decision to treat patients and the choice andduration of treatments.

Predictorsof Treatment Outcome for Anorexia Nervosa

Predictorsof poor outcome for anorexia nervosa include longer duration of illness(Herzog, Keller, & Lavori, 1988) and the presence of vomiting (Herzog etal., 1988). In more recent studies, researchers found that eating disorderduration did not predict outcome (Herzog, Sacks, Keller, Lavori, vonRanson,& Gray, 1993; Herzog, Schellberg, & Deter, 1997). Schoemaker (1997)also found that the duration of illness was not related to treatment outcome.However, Wilson, Vitousek, & Loeb (2000) found that the outcome oftreatment could be determined by the duration of the disorder in anorexianervosa. These studies, as are many others in this review, arecontradictory. In my opinion I don’tthink that researchers have found the methods with which to predict treatmentoutcome with any certainty. They areusing the obvious etiological factors but there are many things stillundiscovered about these disorders.

Wilson et al., (2000) stated that age of onset carriedprognostic significance. Several other studies found that early onset ofillness is a predictor of positive outcome (Herzog et al., 1988) while otherstudies have not found this to be true (Herzog et al., 1988). Herzog et al.(1997) found that anorexic patients who purged and had additional socialdisturbances had a significantly lower chance of recovering. Herzog et al.(1993) reported that anorexics who were also bulimic and who met full criteriafor anorexia nervosa upon admission had a more severe outcome than those whodid not meet full criteria.

Certain laboratory tests have been shown to be excellentpredictors of outcome in anorexia nervosa. Herzog et al. (1997) found thatrestrictor-type anorexia and low serum creatinine levels were predictors ofearlier recovery. They reported that patients with high creatinine levels didnot respond well in treatment. Serum creatinine levels revealed purgingbehavior in patients who tried to conceal the fact they were purging.

Wilson et al. (2000) reported that patients withpsychiatric problems and severe core eating disorder psychopathology would dopoorly with brief, low-intensity treatment. Herzog et al. (1993) hypothesizedthat comorbidity would predict a poorer outcome in eating disorders. They alsofound that anorexic patients with personality disorders had significantly worseoutcomes at 1-year follow-up than those patients without personality disorders.Herzog et al. (1993) also found in their study that percent of ideal body weight (%IBW) type of eating disorders had the strongest predictive value inregards to outcome. The researchers reported that each 10% change in %IBWcorresponded to an 18% increase in eating disorder severity: the lower asubject’s weight, the less likely she was to recover within the follow-upperiod (Herzog et al., 1993).

Researchersfound that fewer patients recovered from anorexia nervosa than from bulimianervosa or a combination of anorexia nervosa and bulimia nervosa. The type ofeating disorder was a significant predictor of outcome with anorexics being theleast likely to recover. At 1-year follow-up in this particular study, only 10%of their subjects had obtained full recovery (Herzog, et al., 1993).

Maddocksand Kaplan (1991) found that patients with a history of anorexia were morelikely to do well in treatment. Their findings do not support those reportsthat show that a previous history of anorexia is a non-significant predictor ofoutcome (Hsu & Holder, 1986). Fairburn, Kirk, O’Connor, Anastasiades, & Cooper (1987) found nodifferences in outcome in those who had a history of anorexia nervosa and thosewho had no history of the disorder. Fallon et al. (1991) reported that adiagnosis of anorexia nervosa at admission was not associated with outcome status.However, Lacey et. al., (1986) found that a history of anorexia was apretreatment indicator of poorer prognosis.

Predictorsof Treatment Outcome for Bulimia Nervosa

Outcomeresearch of bulimia nervosa is still in its infancy. A recent search of twobibliographic databases (MEDLINE and PSYCINFO) revealed that 66% of the papersabout outcome and 77% of the papers about predictors were written since 1993.Francisco Vaz (1998) published a review of prognostic indicators or predictorstreatment outcome, and compiled of table of factors (See Table 7.) that areassociated with outcome (p. 393).

Table A7
Factors Associated with Treatment Outcome Level: Area: Item

Patient:
Specific Eating Pathology
Vomiting; binge frequency; history of AN; attitudes toward weight and body shape; weight (stability/absolute values); age of onset; long history of eating disorder

Patient
Psychopathology
General psychopathology/psychiatric comorbidity; personality disorders; depression; suicide attempts; anxiety; substance abuse; family history of depression and/or alcoholism

Patient
Psychological Characteristics
Self-esteem; ineffectiveness; impulsivity/impulsive behaviors; ego functioning (ego strength/ego disturbances); alexithymia; hypnotizability

Patient’s Environment
Family; Marriage, cohabitation; Interpersonal; Social Class
Family Structure; patterns of interaction; conflicts; interpersonal relationships; social adaptation; socioeconomic level

Treatment
Response to Previous Treatment
Previous failed attempts/Initial Response

Thespecific eating pathology of any given individual has been related to theresponse to treatment with the finding that the most disturbed patients havethe poorest response. Baell & Wertheim (1992) found that pre-treatmentbinge frequency significantly predicts treatment outcome. As such, bingefrequency may reflect entrenchment in the bulimic symptoms.

Turnbull,Schmidt, Troop, Tiller, Todd, & Treasure (1997) conducted a regressionanalysis to establish significant associations with eating disorderpsychopathology at the end of treatment and again at 18-month follow-up, andfound that binge frequency was significantly associated with outcome(p<.001). They found that a lower frequency of binging at the beginning of treatmentwas predictive of a better outcome at the end of treatment.

Wilson,Loeb, Walsh, Labouvie, Petkova, Liu, & Waternaux (1999) found that a highpretreatment incidence of binge eating and purging predicted a poorer responseto treatment. However, in contrast, Vaz (1998) has enumerated a number ofstudies that show that this relationship is questionable (Hsu & Holder,1986; Maddocks & Kaplan, 1991).

Theage of onset and the duration of the illness have produced conflicting reportsas well. In some studies, patients with an early onset had a better prognosisthan those who were older when the illness appeared (Vaz, 1998). Hsu and Holder(1986) did not find that the duration of the disorder predicted outcome;however, Turnbull et al. (1997) did find that the longer duration of illnesspredicted a better outcome. Fallon et al. (1991) found that good outcome wasassociated with a shorter duration of bulimia before hospitalization. InBossert, Schmolz, Wiegand, Junker, & Krieg (1991), the researchers reportedthat even when patients were younger at the onset of the illness, treatmentoutcome could not be predicted.

Researchersfound that the duration of previous hospital treatment(s) for bulimia nervosasignificantly predicted patients’ response to treatment. Vaz (1998) stated thatthe length of the illness does not have a significant influence on the outcomeof bulimia.

Asshown above, there are a number of predictors of outcome that have beenstudied. Out of the many predictors, afew are more apparent than many of the others.

Depression

Thedata concerning the significance of depression in the history of bulimianervosa is conflicting. Fairburn et al. (1987) reported no significance orrelationship to outcome in the pre-treatment level of depressive symptoms. In arecent study (Turnbull, et. al., 1997), depression was not one of thepredictors of outcome for short term or long term treatment for bulimianervosa. However, Bossert et al. (1991) found that at the beginning oftreatment, the patient’s level of depression was closely related to treatmentoutcome. Vaz (1998) reported that it was not certain if depression itself was afactor for affecting outcome or if it was a nonspecific factor that couldcomplicate psychotherapy, causing poor outcome. In this study, it was foundthat patients who had attempted suicide exhibited more severe symptoms ofpsychopathology and had more difficulties in the course of treatment than didother patients.

Ihave found at Mirasol that probably 98% of our patients suffer from majordepression as well as generalized anxiety disorders. I have also found that those patients withsevere personality disorders require much longer in treatment than do thosewithout any Axis II pathology.

Substance Abuse

Therelationship between alcohol and drug abuse and bulimia nervosa is mixed in theliterature. Some studies supported the idea that a history of alcohol abusecould be a predictor of poor outcome (Herzog et al., 1993; Lacey, 1983; Wilsonet al., 1999). Fallon et al. (1991) found that substance abuse was not asignificant factor in regards to recovery and treatment outcome. A surprisingvariation was that Collings & King (1994) reported that the absence of afamily history of alcohol abuse was a predictor of poor outcome. Vaz (1998)reports that families without a history of alcoholism experience a betteroutcome. Then Vaz (1998) went on toreport on another study which indicated that patients without alcoholism intheir families have better outcomes than patients with no family history ofalcoholism (p. 395).

InMirasol’s four day family program, it is the families that have a history ofsubstance abuse that are easier to treat. The rigid, perfectionistic families are much more difficult becausetheir dysfunction is not as obvious and is easier to deny than that of thealcoholic family.

Self-esteem

Oneclient characteristic that seems to be a consistent predictor of treatmentoutcome in the eating disorder literature is low self-esteem (Baell &Wertheim, 1992; Fairburn, Peveler, Jones, 1993; McKisack & Waller, 1997;White, 1999). Vaz (1998) reported that low self-esteem seemed to correlate withpoor prognosis, but not in all studies. Fairburn et al. (1987) found that aconsistent predictor of outcome was self-esteem with the patients having lowself-esteem not responding as well to treatment.

Inmy view the patient with very low ego-strength is the one who does not respondwell to treatment. Self-esteem is a partof ego-strength.

Biopsychosocial Approaches and Eating Disorders

Biopsychosocialmodels appear to be the most integrative approaches to treatment in nearly anydisease process because they posit biological vulnerabilities which interactwith modifiable behavioral and psychological factors, both of which covary withsociocultural factors. Any biopsychosocial model of illness emphasizes uniqueinteractions of biological, psychological and social factors that need to betaken into account to understand the full etiology of the illness (Chapman,Nakamura & Flores, 1999) and represents a movement away from a descriptionin purely medical or psychological terms (Engel, 1977).

Biopsychosocialmodels were developed about 1970 in the arena of chronic pain (Waddell &Turk, 1992). In Waddell and Turk’s model (1992), the cortical areas thatprocessed recent pain experiences somehow triggered emotional memories of pastpain experiences encoded in the limbic system, neospinothalamic and reticularsystems. The impetus of their theory was that it opened the door for a growingdiscussion and further research on the brain’s complex, ongoing andbidirectional influence on the body. At the same time, others argued foralternatives to the biomedical or strictly psychological models. Engel (1977)argued for a model that would bring psychological and social factors intobiomedicine. During the same decade, psychophysiology was ushered in as’empirical evidence’ for the direct physiological effects of corticalactivation on physiology (Andreassi, 2000; Basmajian, 1989). Individuals suchas David Shapiro at UCLA and Ian Wickramasekera at Stanford and the Universityof Illinois dedicated their careers from those early years until the present tothe development of what was to be called Behavioral Medicine.

Biopsychosocial Models andEating Disorders

Thedevastating problem of eating disorders highlights the need for furtherinnovation in research and treatment protocol development in healthcare. AnAmerican Dietetic Association (2001) position statement strongly supports whatclinicians and researchers in the field have come to know: the complexity ofeating disorders involves a multitude of problems which may be best served byan integrative biopsychosocial approach to the problem.

Foreating disorders, a number of studies have discussed biopsychosocial approachesconceptually. Lemmon & Josephson (2001) argue for family therapy as onecomponent in a biopsychosocial approach in eating disorder therapy but make noattempt to empirically validate the outcome of such integration. Kirkley (1986)argues for support for a psychosocial model of bulimia based on epidemiologicaldata from sociology, and Kirkley and Agras’s study (as cited in Kirkley, 1986)further suggests addition of biological predisposition for negative affectincluding depression and anxiety, and names this combination of psychologicaland biological factors a biopsychosocial model (Kirkley, 1986). Other authorsargue directly for the use of a biopsychosocial approach to eating disorders(Burge, 2001; Coupey, 1997; Gupta, 1995; Johnson & Taylor, 1996; Lucas,1981, 1986; Morley, 2002; Zeiter, 2002), while dozens of others attest to aneed for an integrative approach without naming it as such.

Despitethe presence of multiple explanatory models for the development of eatingdisorders, there are no current empirical studies to support any of the models.

Integrative Diathesis-Stress Models for theEating Disorders

Like the term biopsychosocial, diathesis-stress embodiesthe concept of multiple potential vulnerabilities that encompass centralnervous system, genetic and other biological or neurobiological predispositionsto problems that can lead to certain disorders. According to this paradigm, ifthe individual experiences multiple challenges in social functioning, hasphysical problems, family and other interpersonal problems, includingindividual psychological problems, disease may develop. Because of pre-existingvulnerabilities, followed by a host of problems, Kerns and Jacob (1995) believeit is reasonable to conclude that a particular disorder develops as anindividual’s attempt to cope with the entire experience.

Inthe eating disorders literature, there are two empirical studies that reportdiathesis-stress modeling. Joiner, Heatherton, Rudd, and Schmidt (1997) testeda diathesis-stress model on 890 women from two separate non-clinical samples,and found perfectionism to be a predictor of bulimic symptoms if women believedthemselves to be overweight. In this sample, it was the perception of beingoverweight which predicted the bulimic symptoms; the diathesis was generalperfectionism and the stress was the belief of being overweight. In 1999, Vohs,Bardone, Joiner, Abramson, & Heatherton, in a test of the same model, usedhierarchical linear regression to determine if other predictors couldempirically predict bulimia. They found that perfectionism, perceived weightstatus, and self-esteem predicted bulimic symptoms in their multiple linearregression analysis.

Despitethe presence of multiple explanatory models of eating disorder etiologies,there are surprisingly few models that integrate the various viewpointseffectively to directly support a meaningful integration of the variousfindings. Many viewpoints have been reviewed in this study; biological,genetic, family, psychological and social theories of eating disorders.However, there has yet to be a theory-driven substantial study that providesdirect etiological data. Furthermore, regardless of the ever-increasingknowledge of the multidimensional nature of the eating disorder experience,theorists have yet to develop a model that integrates neurobiological andpsychosocial perspectives.

The following is a review of the availableliterature on eating disorders and hypnotizability, negative affectivity,self-deception, catastrophizing, major hassles and minor stressors, copingstrategies, and social support. Theseconstructs were chosen because they can give us a completely different way ofviewing our patients. We find that todaywe can look at a patient according to the DSM diagnostics as well as utilizingthe following symptoms or constructs.

Eating Disorders and Hypnotizability

Hypnotizabilityin eating disordered populations is a risk factor that can be viewed as apersonality trait as well as an ability (Crawford & Gruzelier, 1992; Oakley& Frasquilho, 1998; Morgan, 1973). In addition to hypnotic performance, hypnotizability correlates witheating disorders (Oakley & Frasquilho, 1998). In eating disordered populations, bulimicsare significantly more hypnotizable than anorexics; anorexics of the purgingsubtype generally are more hypnotizable than restricting anorexics(Vanderlinden, Spinhoven, Vandereycken, & Van Dyk, 1995).

Torem (1986) found that 73% of his 30 bulimicpatients were moderately to highly hypnotizable. Barabasz (1991) and Griffiths (1993) studiedhypnotizability in bulimic populations and confirmed previous findings ofhypnotizability in this population.

Intheir study of bulimic patients, Covino, Jimerson, Wolfe, Franki. & Frankel(1994) found that bulimics were highly hypnotizable when compared with otherpsychiatric patients and show high dissociative ability.

Highhypnotizables are hypersensitive to psychological and physiologicalchanges. They generally have superiorsensory memory and have a superior ability to transfer information from sensorymemory to short term memory. Thisability could be used to rapidly learn and retain operant anxiety. This is the possible reason that bulimicsfind strong emotion or feelings intolerable for them.

Pettinatiand Wade (1986) report that having the ability to experience dissociation maybe a relevant factor in regards to the high hypnotizability found in bulimicpatients. Clinicians have compared theacts of binge eating and purging to dissociative experiences.

Beumontand Abraham (1983) found that in a group of 30 bulimics, 75% had experienceddissociation. Dissociation has not beenfound to be a feature of anorexia nervosa of the restricting type (Vanderlindenet al., 1995). Vanderlinden et al. (1995) found higher scores in bulimics andpurging anorexics on the Dissociation Questionnaire (DIS-Q, Vanderlinden,Vandereycken, van Dyk, & Vertommen, 1993). Oakley and Frasquilho (1998) reported that hypnotizability anddissociation can be related to body image distortion which is common in bothanorexia and bulimia nervosa. Anindividual who is highly hypnotizable can easily absorb or internalize themessages from society that promote a slim body shape as the ideal. This is the factor that is known as suggestibility. Internalization of the thin body ideal leadsto body dissatisfaction.

Bothanorexics and bulimics have been found to have equivalent body imagedisturbances; both populations overestimate their body size (Williamson, Cubic,& Gleaves,1993). This populationutilizes restrained eating as a strategy to meet society’s ideal and also toreduce negative affect (Oakley & Frasquilho, 1998).

Manyanorexics are of low hypnotic ability and as a result they frequentlydemonstrate a hyposensitivity to psychological and physiological changes, atendency to deny psychological causation of behavior, and a propensity toremain in denial of the severity of their illness. The anorexic who is low in hypnotic ability issubject to stress disorders because he or she is relatively insensitive to ordeficient in attention to relationships between psychological states andphysiological states. They have apsychological insensitivity to changes in mood and feelings. They have a lack of proprioceptive orinteroceptive awareness which means they lack the ability to discriminatebetween different feeling states in the body, such as being hungry and beingfull. Many anorexics are alexithymic,meaning they have no words for moods.

EatingDisorders and Negative Affectivity and Self-Deception

Both of these constructs areincluded in one section because they are interrelated and it is necessary tohave an understanding of both together. As noted above, neuroticism or negative affectivity (NA) is distinguishedby the ability to remember unpleasant or aversive events that happened in thepast. High NA individuals will feelincreased discomfort at all times even without the presence of stress and willreact more dramatically to difficult situations. They report more unpleasant feelings in timesof stress and well as times of no stress (Watson & Clark, 1984). When they are high in negative affectivity,they tend to have more negative emotions such as anxiety. Anxiety facilitates classical conditioningand the gastrointestinal tract is known to be one of them most rapidlyconditionable organs in humans (Wickramasekera, personal communication). Anxiety is one of the most common complaintsof women with eating disorders.

Paaand Larson (1997) found that negative affect was a significant predictor ofrestrained eating in women. It isestimated that as many as 30% of American women may be restrained eaters(Eldredge, Wilson, & Whaley, 1990). Restrained eaters depend on dieting rules rather than interoceptiveawareness to dictate how they choose to eat. When individuals have a lack on interoceptive awareness, they lack theability to tell when they are hungry or full. This population is more likely to experience weight cycling and has agreater likelihood for developing an eating disorder than unrestrained eaters(Eldredge et al., 1990).

Grilo,Masheb, and Berman (2001) conducted a research study where they subtyped womenwith bulimia nervosa along dietary and negative affect dimensions. They found a dietary-negative affect subtypein a group of 48 women tested. 56% ofthe women in the group were found to be of the dietary-negative affectsubtype. This subtype was found to havegreater eating-related psychopathology and psychological disturbance.

Kaye, Lavia, Frak, &Molina (2001) reported that anorexia nervosa and bulimia nervosa have manypsychological characteristics in common. One of these important traits is negative affect. They hypothesize that the commonalities maybe related to serotonergic function.

Another construct thatis important to this discussion is repression that is defined as a lowneuroticism score on the Eysenk Personality Inventory (Eysenck & Eysenck,1968) and a high Marlowe Crowne score (Crowne & Marlowe, 1960). A high repression score means that patientscan block a threat from consciousness, in essence enabling patients to keep asecret from themselves but not from their bodies. When distress or negative affect is repressedfrom consciousness it frequently manifests itself in the body in the form ofsomatic complaints. Patients arecompletely unaware of many of the underlying physiological and psychologicalissues. Repressors often have a vestedinterest in seeing themselves as being perfectly well-adjusted at the expenseof their emotional lives, and frequently are not in touch with their needs aswell as their bodies. They willfrequently appear completely healthy on standard psychological tests and areunaware of the fact that “illusory mental health” may be a risk factor forphysical illness.

Eating Disorders and MajorHassles and Minor Stressors

Major hassles and minor stressorsare both factors that can precipitate as well as promote the maintenance of aneating disorder. Reiff and Reiff (1990)in their list of seven scenarios which can precipitate an eating disorder,named some of the events or causal factors, such as physical, sexual, oremotional abuse, that could not only precipitate but also contribute to themaintenance of an eating disorder. Cohen, McGowan, Fooskas, & Rose (1984) discuss the relationshipbetween life experiences and psychological disorder. They reported a relationship betweenpsychological distress and negative life events or major hassles and minorstressors. Coping strategies and socialsupport systems are adaptations to stress that may be used to buffer to hasslesand stressors.

Eating Disorders, CopingStrategies, and Social Support

Coping is important as a way ofmediating the stressors that can lead to psychological disorders, includingeating disorders. In 76% of cases,stressful life events or major hassles/minor stressors precede the onset ofanorexia nervosa (AN) and bulimia nervosa (BN) (Schmidt, Treasure, Tiller,Blanchard, & Harris, 1991).

Generally active copingstrategies, such as taking action and problem solving, are associated withlower levels of stress, anxiety, and depression, and avoidant coping strategiesare associated with increased psychological distress (Holohan & Moos,1985). Mayhew and Edelman (1989) havefound that eating disorder pathology is related to avoidant coping in studentpopulations.

In a study conducted byTroop, Holbrey, Trowler, & Treasure (1994), researchers found thatanorexics and bulimics used more avoidance coping than did controlsubjects. They discovered that bulimicsubjects relied more on wishful or magical thinking than did controls oranorexics.

Eating disorder patientsdid not use problem-focused coping as often as controls and self-blame in thispopulation was significantly higher (Troop et al., 1994). Troop et al. (1994) also found that BeckDepression scores (BDI) were related to avoidant coping so that the moredepression patients exhibited, the more often they used avoidant copingmethods. BDI scores were also inverselyrelated to problem-focused coping.

Endler and Parker (1990a, 1990b)report on another view of avoidance coping, namely that a person can avoidstress by distracting herself with a substitute task or by looking for supportfrom some other person. I believe thatthis is what happens with individuals with eating disorders: They are confronted with life stress eitherin the form of major hassles or minor stressors, and they distract themselvesthrough the use of eating disordered behavior. 

Endler & Parker(1990a, 1990b) found that avoidance coping was positively related topsychological dysfunction. There hasbeen speculation that negative body image might function as a stressor (Hawkins& Clement, 1984), thereby possibly promoting the development of an eatingdisorder through the use of negative coping strategies. Shatford and Evans(1986) reported that women with bulimia nervosa were more likely to useavoidance-focused coping to deal with stress associated with negative lifehassles or stressors.

Social support goes hand in handwith coping strategies and in some ways, it can be considered a type of copingstrategy. If an individual isexperiencing negative life stress, a way of coping with that stress could bereaching out to friends or family for emotional or financial support. Caplan (1974) suggested that the presence ofstrong social support systems was a major factor in determining an individual’sadjustment to stressful life events.

Of interest is a studydone by Cohen, McGowan, Fooskas, & Rose (1984) where researchers found that”received social support does not have a direct or stress-buffering effect onpsychological disorder” (p. 585). However, perceived social support can have a direct positive effect onstressful life events.

Grissett and Norvell (1992) studiedthe relationship between perceived social support and women with bulimianervosa. They found that bulimicstypically have impaired relationships with friends and family, reporting lessperceived social support, more conflict in their relationships, and a reducedability to be effective socially. Morefrequent binging and purging was associated with a higher level of socialimpairment (Herzog, Norman, Rigotti, & Pepose, 1986; Herzog, Keller,Lavori, & Ott, 1987).

Eating disorderedpatients are more dissatisfied with their social support and feel more anxiousand alienated from family and friends. Their lack of strong social support systems can be a factor in themaintenance of the disorder (Grissett & Norvell, 1992). The lack of social competence may alsocontribute to their inability to develop and utilize a social support system.

Itis now well established in the literature how the central and peripheralcomponents of the stress system interrelate and are the mechanism by which theycontribute to physical disease (Habib, Gold, & Chrousos, 2001; McEwen,2002). For instance, the two leading causes of death in the year 2000, heartdisease and cancer (NIH, 2003), have a known stress-related component in thedisease process (Harbuz, 2002; Opie, 2002).

Eatingdisorders have one of the highest mortality rates of any DSM-IV diagnosis(Fichter & Quadflieg, 1999). Despitethe increasing prevalence worldwide and its high mortality rate, the treatmentsin mainstream practice are not highly effective. The rates of recovery following treatment areshockingly low. These abysmal recoveryrates indicate that there is a need to explore and develop treatmentalternatives to find more effective treatments. 

For list of References, click here.