Eating disorders have steadily moved into full recognition as common, serious disorders with a high mortality rate (1). Although treatment remains pragmatic, there is increasing recognition of a genetic predisposition for, and neurobiological consequences of, eating disorders. Eating disorders challenge monistic etiological paradigms because they require the presence of abnormal eating behavior and most human behavior has polythetic contributions.
Many myths about eating disorders have been exploded in the last decade. The group of people with eating disorders clearly includes men, older women, and minorities—groups that are often overlooked because of the common stereotype of teenage girls with eating disorders (2). They are curable disorders, not necessarily life-long illnesses, a prognosis not always available to other psychiatric disorders (3). The recently published Practice Guideline for the Treatment of Patients With Eating Disorders(4) has brought a coordinated multidisciplinary approach to treatment in place of unvalidated monotherapies. There is increasing clarification that the core psychopathology, a morbid fear of fatness and/or a relentless drive for thinness, is not a delusional belief nor an ego-alien symptom but an overvalued belief—an intriguing, little studied theory.
Psychiatric disorders in general, and eating disorders in particular, have been subjected to draconian restrictions on length of hospital stay for treatment. A positive result has been the development of a spectrum of care; partial hospitalization programs now care for the majority of patients, both as a step-down from shorter inpatient care and a step up from outpatient evaluation. Recent studies (5) have validated the empirical decisions for transition to partial hospitalization.
Multiple controversies exist in the field of eating disorders. The role of medications remains enigmatic. In the treatment of bulimia nervosa, antidepressants are effective short-term competitors in treatment effectiveness with cognitive behavior therapy, but their withdrawal more often leads to relapse than with cognitive behavior therapy (6). In contrast to schizophrenia or bipolar disorder, in which the concept of meaningful connections as substantial contributors to illness lacks evidence, eating disorders usually begin as coping mechanisms in the service of solutions to developmental challenges, mood stabilization, or regulation of family conflict. The role of psychodynamics, as opposed to more validated cognitive behavior therapy or interpersonal psychotherapy (7), remains tantalizingly elusive. Finally, international viewpoints concerning the preventability of eating disorders remain sharply divided.
A relevant contribution in this issue, by Westen and Harnden-Fischer on personality profiles in eating disorders, tackles the often-unexamined distinction between axis I and axis II disorders. Most clinicians recognize that eating disorders may occur with comorbid disorders that complicate treatment. Westen and Harnden-Fischer advance beyond this pragmatic understanding with a study demonstrating that personality pathology clusters predict the severity of eating disorder symptoms, adaptive functioning, and etiological variables. Axis II components, in their view, are not merely complicating baggage but defining features. Left unanswered are the specifics of how the alliance of eating disorders with one of three personality pathology clusters modifies treatment and whether these clusters play a significant role in the onset and maintenance of the disorder or increase its complexity.
Bellodi et al. attack the subject of comorbidity in eating disorders by going one step further: suggesting that eating disorders may not be independent disorders but "should be considered as part of the obsessive-compulsive spectrum." This hypothesis is supported by their finding of a greater probability of the obsessive-compulsive state or trait in patients with anorexia nervosa. Additional support comes from Srinivasagam et al. (8), who speculated that one set of genetic vulnerabilities may predispose one to developing any eating disorder, but anorexia nervosa in particular requires the presence of a second genetic predisposition—traits of perfectionism, symmetry-seeking, or perseverance.
Problematic with the Bellodi et al. contribution is the lesson of multiple previous failed experiments attempting to demonstrate that eating disorders are a forme fruste of schizophrenia, depressive disorder, family dysfunction, or multiple personality disorder. None has established the transmutability of eating disorders into a more fundamental diagnosis but instead have repeated the experiment of the blind men and the elephant, each grasping an accurate but limited subset of the whole animal. The presence of more obsessive-compulsive features in relatives of eating disorder patients can also be viewed as a factor that puts one at greater risk.
Experienced authors from the University of Toronto, Woodside et al. illuminate the issue of men versus women with eating disorders: how are they similar or different? Their methodology circumvents the problem of studying only clinical referrals by comparing cases derived from a community-based epidemiological study. A number of surprising results were found. The prevalence of full anorexia nervosa in men is much higher (16%) than in clinically derived estimates (5%–10%), suggesting a community reservoir of undiagnosed men. Gender parity is almost established in groups with partial syndromes of anorexia nervosa (men to women=1:1.5) and in groups with partial bulimia nervosa (men to women=1:1.8). The authors demonstrate that men and women with eating disorders are substantially similar to each other, which is comforting for making diagnoses and carrying out treatment. Their contribution illuminates the total but not the individual components of gender difference considerations in eating disorders. I and others have suggested significant differences in social learning and reasons for dieting between men and women with eating disorders (9). Whatever these differences, the welcome focus on evidence-based gender comparisons begins to remediate the previous banishment of men from the capability of having eating disorders because of theoretical dogma (not having a "fear of oral impregnation") or statistical inconvenience.
A number of additional areas warrant study. In regard to the diagnosis of anorexia nervosa, at least two studies (10, 11) have shown that amenorrhea is not necessarily present in patients with eating disorders, which enlarges the category of anorexia nervosa and diminishes the overly broad "eating disorder not otherwise specified"—a source of diagnostic confusion. Additionally, the requirement of achieving less than 85% of normal weight for diagnosis is arbitrary. It is, in fact, the decrement of weight from any stable set point to a substantially lower weight that produces the starvation symptoms that are typical of anorexia nervosa.
National standards for the treatment outcome of anorexia nervosa and bulimia nervosa are within our grasp. Specifying a minimum range of weight restoration, specific percentage decreases in bulimic symptoms, and desirable changes in core psychopathology within a time-limited treatment phase are now possible. In most situations where there are concerns about gender discrimination, it is usually women who suffer most, but that is not the case regarding eating disorders. Men with eating disorders are often excluded from programs by gender alone or are treated indistinguishably from teenage girls.
Optimism in regard to the future of treating eating disorders, however, is entirely appropriate in this postmillennial age of irony. We live in a time where past promises of shorter work weeks remain unfulfilled, where technology takes precedence over human rights, and where the world is increasingly split into starving and overfed populations. It is into this complex arena that eating disorders present themselves as enigmas requiring a new paradigm but, above all, as disorders requiring an informed, compassionate, persevering, gender-specific approach to treatment.
Address reprint requests to Dr. Andersen, Department of Psychiatry, University of Iowa Hospitals and Clinics, 200 Hawkins Dr., Iowa City, IA 52242; firstname.lastname@example.org (e-mail).