Eating Disorders: Current Knowledge and Treatment Update

Eight years ago, DSM-5 made major changes to the diagnostic criteria for eating disorders. A major problem in DSM-IV’s criteria was that only two eating disorders, anorexia nervosa and bulimia nervosa, were officially recognized. Therefore, many patients presenting for treatment received the nonspecific diagnostic label of eating disorder not otherwise specified (EDNOS), which provided little information about the nature of the patient’s difficulties. This problem was addressed in several ways in DSM-5 (see DSM-5 Feeding and Eating Disorder list). The diagnostic criteria for anorexia nervosa and bulimia nervosa were slightly expanded to capture a few more patients in each category. But two other changes had a greater impact in reducing the use of nonspecific diagnoses.

The first of these was the addition of binge eating disorder (BED), which had previously been described in an appendix of DSM-IV. BED is the most common eating disorder in the United States, so its official recognition in DSM-5 led to a substantial reduction in the need for nonspecific diagnoses.

The second important change was the combination of the DSM-IV section titled “Feeding and Eating Disorders of Infancy or Early Childhood” with “Eating Disorders” to form an expanded section, “Feeding and Eating Disorders.” This change thereby included three diagnostic categories: pica, rumination disorder, and feeding disorder of infancy or early childhood. Pica and rumination disorder are infrequently diagnosed.

The other category, feeding disorder of infancy or early childhood, was rarely used and had been the subject of virtually no research since its inclusion in DSM-IV. The Eating Disorders Work Group responsible for reviewing the criteria for eating disorders for DSM-5 realized that there was a substantial number of individuals, many of them children, who severely restricted their food intake but did not have anorexia nervosa. For example, after a severe bout of vomiting after eating, some individuals attempt to prevent a recurrence by no longer eating at all, leading to potentially serious nutritional disturbances. No diagnostic category in DSM-IV existed for such individuals. Therefore, the DSM-IV category, feeding disorder of infancy or early childhood, was expanded and retitled “avoidant/restrictive food intake disorder” (ARFID). Combined, these changes led to a substantial reduction in the need for nonspecific diagnostic categories for eating disorders.

In the course of assessing the impact of the recommended changes in the diagnostic criteria for eating disorders, the Eating Disorders Work Group became aware of another group of individuals presenting for clinical care whose symptoms did not quite fit any of the existing or proposed categories. These were individuals, many of them previously overweight or obese, who had lost a substantial amount of weight and developed many of the signs and symptoms characteristic of anorexia nervosa. However, at the time of presentation, their weights remained within or above the normal range, therefore not satisfying the first diagnostic criterion for anorexia nervosa. The work group recommended that a brief description of such individuals be included in the DSM-5 diagnostic category that replaced DSM-IV’s EDNOS: “other specified feeding and eating disorders” (OSFED); this description was labeled atypical anorexia nervosa. The degree to which the symptoms, complications, and course of individuals with atypical anorexia nervosa resemble and differ from those of individuals with typical anorexia nervosa remains an important focus of current research.

Epidemiology

Although eating disorders contribute significantly to the global burden of disease, they remain relatively uncommon. A study published in September 2018 by Tomoko Udo, Ph.D., and Carlos M. Grilo, Ph.D., in Biological Psychiatry examined data from a large, nationally representative sample of over 36,000 U.S. adults 18 years of age and older surveyed using a lay-administered diagnostic interview in 2012-2013. The 12-month prevalence estimates for anorexia nervosa, bulimia nervosa, and BED were 0.05%, 0.14%, and 0.44%, respectively. Although the relative frequencies of these disorders were similar to those described in prior studies, the absolute estimates were somewhat lower for unclear reasons. Consistent with clinical experience and prior reports, the eating disorders, especially anorexia nervosa and bulimia nervosa, were more prevalent among women (though men are also affected). Although eating disorders occurred across all ethnic and racial groups, there were fewer cases of anorexia nervosa among non-Hispanic and Hispanic Black respondents than among non-Hispanic White respondents. Consistent with long-standing clinical impression, individuals with lifetime anorexia nervosa reported higher incomes.

Finally, when BED was under consideration for official recognition in DSM-5, some critics suggested that, since virtually everyone occasionally overeats, BED was an example of the misguided tendency of DSM to pathologize normal behavior. The low prevalence of BED reported in the study by Udo and Grilo documents that, when carefully assessed, BED affects only a minority of individuals and is therefore distinct from normality.

A subject of some debate and substantial uncertainty is whether the incidence of eating disorders (the number of new cases a year) is increasing. Some studies, such as that of Udo and Grilo, have found that the lifetime rates of eating disorders among older individuals are lower than those among younger individuals, suggesting that the frequency of eating disorders may be increasing. However, this might also reflect more recent awareness and knowledge of eating disorders. Other studies that conducted multiple examinations of the frequency of eating disorders in the same settings over time appear to suggest that, in the last several decades, the incidence of anorexia nervosa has remained roughly stable, whereas the incidence of bulimia nervosa has decreased. Presumably, this reflects changes in the sociocultural environment such as an increased acceptance of being overweight and reduced pressure to engage in inappropriate compensatory measures such as self-induced vomiting after binge eating.

The COVID-19 pandemic has impacted virtually every facet of life across the world and has produced severe financial, medical, and psychological stresses. Preliminary research suggests that such stresses have exacerbated the symptoms of individuals with preexisting eating disorders and have led to increased binge eating in the general population. Hopefully, these trends will improve with successful control of the pandemic.

Neurobiology

Much recent research on the mechanisms underlying the development and persistence of eating disorders has focused on the processing of rewarding and nonrewarding/punishing stimuli. Several studies have suggested that individuals with anorexia nervosa are less able to distinguish among stimuli with varying probabilities of obtaining a reward. Other studies suggest that, when viewing images of food during MRI scanning, individuals with anorexia nervosa tend to show less activation of brain reward areas than do controls. Such deficits may be related to disturbances in dopamine function in areas of the brain known to be involved in reward processing. Research based on emerging methods in computational psychiatry suggests that individuals with anorexia nervosa may be particularly sensitive to learning from punishment; for example, they may be very quick to learn what stimuli lead to a decrease in the amount of a reward. Conceivably, they may learn that eating high-fat foods prevents weight loss and produces undesirable weight gain, and they begin to avoid such foods. These studies, and a range of others, focus on probing basic brain mechanisms and how they may be disrupted in anorexia nervosa. A challenge for this “bottom-up” approach is to determine how exactly disturbances in such mechanisms are related to the eating disturbances characteristic of anorexia nervosa.

Other recent studies take a “top-down” approach, focusing on the neural circuitry underlying the persistent maladaptive choices made by individuals with anorexia nervosa when they decide what foods to eat. Such research successfully captures the well-established avoidance of high-fat foods by individuals with anorexia nervosa and has documented that such individuals utilize different neural circuits in making decisions about what to eat than do healthy individuals. These results are consistent with suggestions that the impressive persistence of anorexia nervosa in many individuals may be due to the establishment of automatic, stereotyped, and habitual behavior surrounding food choice. A challenge for such top-down research strategies is to determine how these maladaptive patterns develop so rapidly and become so ingrained.

Research on the neurobiology underlying bulimia nervosa is broadly similar. Although the results are complex, individuals with bulimia nervosa appear to find food stimuli more rewarding, and there are indications of disturbances in reward responsiveness to sweet tastes. Several studies have documented impairments in impulse control assessed using behavioral paradigms such as the Stroop Task. In this task, individuals are presented with a word naming a color (for example, “red”) but asked to name the color of the letters spelling the word (for example, the letters r, e, and d are green). Increased difficulties in performing such tasks have been described in individuals with bulimia nervosa and linked to reduced prefrontal cortical thickness.

Genetics

It has long been known that eating disorders tend to run in families, and there has been strong evidence that this in part reflects the genes that individuals inherit from their parents. In recent decades, it has become clear that the risk of developing most complex human diseases, including obesity, hypertension, and eating disorders is related to many genes, each one of which contributes a small amount to the risk. Because the contribution of a single gene is so small, the DNA from a very large number of individuals with and without the disorder needs to be examined. For instance, genomewide association studies (GWAS) in schizophrenia have examined tens of thousands of individuals with schizophrenia and over 100,000 controls and identified well over 100 genetic loci that contribute to the risk of developing schizophrenia.

GWAS examining the genetic risk for eating disorders are under way but to date have focused primarily on anorexia nervosa. The Psychiatric Genetics Consortium has collected information from 10,000 to 20,000 individuals with anorexia nervosa and over 50,000 controls and has, so far, identified eight loci that contribute to the genetic risk for this disorder. In addition, this work has identified genetic correlations between anorexia nervosa and a range of other disorders known to be comorbid with anorexia nervosa such as anxiety disorders as well as a negative genetic correlation with obesity. These data suggest that the genetic risk for anorexia nervosa is based on a complex interplay between loci associated with a range of psychological and metabolic/anthropometric traits.

Treatment

Although there have been no dramatic developments in our knowledge of how best to treat individuals with eating disorders, there have been some significant and useful advances in recent years.

For anorexia nervosa, arguably the most significant advance in treatment in the last quarter century has been family-based treatment for adolescents. In this approach, sometimes referred to as the “Maudsley method,” the family, guided by the therapist, becomes the primary agent of change and responsible for ensuring that eating behavior normalizes and weight increases. This approach differs markedly from prior treatment strategies that assumed parental involvement was not helpful or even detrimental. Family-based treatment is now widely viewed as a treatment of first choice for adolescents with anorexia nervosa and has also been adapted to treat bulimia nervosa.

Family-based treatment can be quite challenging for parents. The entire family is asked to attend treatment sessions, and one session early in treatment includes a family meal during which the parents are charged with the difficult task of persuading the adolescent to consume more food than he/she had intended. An alternative but related model, termed “parent-focused treatment,” has recently been explored in a few studies. In this approach, parents meet with a therapist without the affected adolescent or other members of the family and receive guidance regarding how to help the adolescent to alter his or her behavior following techniques virtually identical to those provided in traditional family-based treatment. Several small studies have examined this approach, and results suggest similar effectiveness. Although more research is needed, these findings suggest that parent-focused treatment may be an attractive alternative to family-based treatment for many parents and practitioners.

The COVID-19 pandemic has led to a dramatic acceleration in the provision of psychiatric care remotely, including family-based treatment. Work on providing family-based treatment via videoconference had begun prior to the arrival of COVID-19, as this specialized form of care is not widely available, and its provision via HIPAA-compliant video links would offer a substantial increase in accessibility. Several small studies suggested that remote provision of family-based treatment is feasible and likely to be efficacious. The restrictions imposed by COVID-19 on face-to-face contact have accelerated the remote delivery of family-based treatment; hopefully, new research will document its effectiveness. It should be noted, however, that, in most cases, local contact with a medical professional who can directly measure weight and oversee the patient’s physical state is required.

The treatment of adults with anorexia nervosa, who typically developed the disorder as teenagers and have been ill for five or more years, remains challenging. Structured behavioral interventions, such as those available in specialized inpatient, day program, or residential centers, typically lead to significant weight restoration and psychological and physiological improvement. However, the rate of relapse following acute care remains substantial. Furthermore, most adult patients with anorexia nervosa are very reluctant to accept treatment in such structured programs. A recent helpful development is evidence that olanzapine, at a dose of 5 mg/day to 10 mg/day, assists modestly with weight gain in adult outpatients with anorexia nervosa and is associated with few significant side effects. Unfortunately, it does not address core psychological symptoms and must be viewed as adjunctive to standard care.

There have been fewer recent developments in the treatment of patients with bulimia nervosa and of BED. For bulimia nervosa, cognitive-behavioral therapy remains the mainstay psychological treatment, and SSRIs continue to be the first-choice class of medication. For BED, multiple forms of psychological treatment are associated with substantial improvement in binge eating, and, in 2015, the FDA approved the use of the stimulant lisdexamfetamine (Vyvanse) for individuals with BED. Unlike most psychological treatments, lisdexamfetamine is associated with modest weight loss but has effects on pulse and blood pressure that may be of concern, especially for older individuals.

Also noteworthy are the development and application of new forms of psychological treatment for individuals with eating disorders. These include dialectical behavior therapy (DBT), acceptance and commitment therapy (ACT), and integrative cognitive-affective therapy (ICAT). Although only a few controlled studies have examined the effectiveness of these treatments, anecdotal information and the results of these studies suggest that such methods may be useful alternatives to more established interventions.

Conclusions

Eating disorders remain uncommon but clinically important problems characterized by persistent disturbances in eating or eating-related behavior. Cutting-edge research focuses on neurobiology and genetics, utilizing novel and rapidly evolving methodology. There have been modest advances in treatment approaches, including the COVID-19 pandemic’s acceleration of treatment delivery via video-link. Future studies will hopefully clarify the nature of ARFID and of atypical anorexia nervosa and lead to the development of more effective interventions, especially for individuals with long-standing eating disorders. ■