16 Research Agenda for Eating Disorders

Research on eating disorders has produced a substantial base of knowledge regarding the definitions of eating disorders, their treatment, and their prevention. However, there are significant gaps in our understanding of eating disorders, especially among adolescents. Most of the research literature on eating disorders has focused on adults, and the findings may not apply to younger individuals. The research agenda regarding eating disorders among adolescents is large. Future studies, such as those outlined below, could take advantage of the increase in available technology (e.g., smartphones) and use of social media, which are of particular relevance to this age group.

In light of the minor alterations to the diagnostic criteria for anorexia nervosa (AN) and bulimia nervosa (BN), the inclusion of binge-eating disorder (BED), and rearticulated diagnoses (e.g., avoidant/restrictive food intake disorder), it will be important to understand whether the changes described in DSM-5 are helpful in reducing the use of residual feeding or eating disorder categories among adolescents without a loss of diagnostic utility. Further, information is needed about adolescents within novel DSM-5 categories, in particular BED and avoidant/restrictive food intake disorder, as little is known about the clinical characteristics of youth with these diagnoses. Finally, the utility of other diagnostic schemes (e.g., Broad Categories for the Diagnosis of Eating Disorders; Walsh & Sysko, 2009) should be considered in younger populations. A broader system of classification could offer an advantage for adolescents, because it may be able to accommodate the range of symptom presentations noted among youth and is more straightforward for assessing individuals with eating disorders who present to primary care settings, where a large proportion of adolescents with eating disorders are initially seen.

In 2010, the National Institute of Mental Health introduced the Research Domain Criteria (RDoC), a novel framework to support research on clinical symptoms and their linkage to underlying neurobiology (Insel et al., 2010). Several problematic issues related to eating disorders illustrate the rationale for RDoC. The eating disorders as defined by DSM-5 are characterized by a high degree of comorbidity and overlapping symptoms; for example, significant depression is extremely common in AN, BN, and BED, and binge eating occurs among individuals in all three categories. In addition, many individuals with clinically significant eating problems fail to meet DSM-5’s diagnostic thresholds, such as a minimum average of one episode of binge eating per week over 3 months for both BN and BED, and therefore receive a “residual” diagnosis. The goal of the RDoC approach is to avoid such problems by eschewing DSM-5 categories, by examining symptoms dimensionally and by linking them to underlying neurobiological constructs such as reward processing and impulse control.

Despite its many deficiencies, the DSM system has well-established clinical utility and broad acceptance, and the RDoC system is, as its name suggests, designed to augment research rather than clinical care. However, the fundamental principles underlying the RDoC framework are sound and provide an important foundation for novel probes of the development and persistence of eating disorders.

As sample sizes for genome-wide association studies (GWAS) of individuals with eating disorders grow, all indications are that the discovery trajectory will mirror that observed in schizophrenia. Future genomic work will identify genes and gene sets ripe for functional study using rapidly developing technologies and methodologies available to neuroscientists. Elucidation of the biology of eating disorders via genomics and other “-omics” will open avenues for the development and testing of novel pharmacological agents that directly target the core biology of the illness.

Emerging GWAS results will also permit the development of genetic risk profiles and the exploration of their ability to identify at-risk children and to predict the course and outcome of eating and comorbid disorders. In the longer term, such risk profiles may be applied in the clinic to inform the personalization of treatment. Finally, the common mechanisms underlying comorbidity identified via genomics will translate immediately into the pilot testing of novel pharmacological agents to tailor and improve treatment for eating disorders.

More large studies are needed to understand the true prevalence of eating disorders among adolescents. As described above, changes to the diagnostic criteria should result in some increase in the prevalence of AN, BN, and BED and a reduction in the frequency of less-specific diagnostic labels, but only large, nationally representative samples will provide information on the impact of these DSM-5 changes on rates of feeding and eating disorders. Future studies should also examine the prevalence of specific types of other specified feeding and eating disorders (e.g., subthreshold AN, purging disorder), which will facilitate comparisons of results across studies. Large prospective studies would provide clarification about the age of onset of DSM-5 disorders. Research is needed to better understand substantial racial/ethnic and gender differences in the prevalence of disordered eating. Alterations to the method of assessment in epidemiological studies may be needed, as a growing number of studies document weight and shape concerns and related behaviors that are more common in males and not measured in existing studies.

Psychiatric comorbidities are common among both adolescents and adults with AN and BN and include mood disorders, anxiety disorders, and substance use disorders. The outcome of adolescent patients with AN who receive treatment soon after the onset of illness appears better than that of patients who do not; however, those patients who remain ill have high rates of psychiatric comorbidity and are at risk for premature death (Franko et al., 2013). Data on the course and outcome of adolescent BN are very limited.

Diagnostic migration occurs frequently from AN-restricting subtype to AN-binge purge subtype, and from AN-binge purge subtype to BN, and it is not currently possible to identify those patients likely to migrate. Future studies should include individuals with comorbidities, such as substance use disorders, to aid in developing treatment strategies for these dual-diagnosis conditions.

Studies of the course and outcome of adolescent BN are needed, and early identification and intervention strategies need to be developed.

While the majority of medical complications associated with eating disorders are reversible with nutritional rehabilitation and cessation of the binge–purge cycle, there are indications that growth retardation, reduced bone mass, and, possibly, structural brain changes may not be entirely reversible. Studies probing structural brain changes in AN and their relationship to neuropsychological changes are needed. In addition, there is a pressing need to develop efficacious treatments for reduced bone mass among adolescents with AN.

A better understanding of the pathophysiology of the refeeding syndrome is required to avoid unnecessarily conservative refeeding protocols, which can lead to prolonged and costly inpatient medical hospitalization. Randomized controlled trials are necessary to examine the safety of different rates of refeeding and the effect of macronutrient composition on development of the refeeding syndrome.

A significant focus of future research should be on effective psychological treatments for adolescents with eating disorders. Family-based therapy appears useful for adolescents with AN and BN, and self-help forms of cognitive-behavioral interventions show promise for adolescents with BN. However, it is imperative to build on these initial efforts. Several future research priorities should be highlighted. It is clear that additional treatment modalities in this population should be explored. Perhaps the most obvious candidate is Enhanced Cognitive Therapy (CBT-E; Fairburn, 2008), which has been studied in other eating disorder groups. Another priority is to design randomized controlled trials to specifically identify moderators and mediators of treatment outcome so that patients can be matched with the appropriate treatment modality.

Few controlled studies have evaluated the utility and safety of pharmacological treatments for adolescents with eating disorders, although medications are frequently used in the clinical treatment of these patients. Antidepressant medications reduce binge eating and vomiting behaviors for adults with BN, but additional study will be necessary before firm conclusions can be reached regarding the use of these medications with younger patients. There is currently no evidence for the efficacy of pharmacological treatments in low-weight adults with AN; however, recent reports have suggested that atypical antipsychotic medications, such as olanzapine, may be useful, and future research should evaluate these medications in a controlled manner among younger patients.

The field has made significant strides in terms of documenting effective treatments across the range of eating disorders. Although more studies focused on adolescents have been published, more can and should be done to develop interventions for this age group. Data across disorders consistently indicate that early intervention has positive effects, and the field would benefit from continuing to gather data that provide an empirical basis for defining treatment response, relapse, and remission. Promising findings related to Internet-based strategies should be pursued, particularly given the reality that adolescents are technology “natives,” and it is likely that they will find such strategies familiar. Privacy issues are a particular concern with Internet-based strategies, especially when working with minors in such a context. However, the economic leverage and extensive reach that are feasible with the Internet represent strong advantages.

As relapse-prevention interventions develop further, it will be useful to analyze the components of the interventions to determine what is most salient and significant in preventing relapse. Relatedly, as data emerge from neuroscience, more targeted interventions informed by basic neuroscience may be important in terms of understanding who is at risk and which specific interventions prevent relapse. Studies also need to pay careful attention to distinguishing between enduring impairment and relapse.

Gone are the days when it was thought that individuals could not fully recover from eating disorders; however, we continue to have much work to do to ensure that more individuals achieve full and lasting recovery.

Adolescents have not been the focus of medication trials to prevent relapse, and studies that have included adolescent patients are not sufficient to draw conclusions about relapse prevention for these patients. It is important to determine whether response rates to continued medication regimens are similar in adolescents to the extant data for adults.

Our definition of “risk factor” has evolved in several meaningful ways over the past decade. In the early years of research, the focus was primarily on sociocultural factors, family factors, and life events. The current conceptualization of risk is much more fluid and encompasses not only environmental factors but also neurobiological, genetic, and epigenetic factors, as well as interactions and correlations across these domains, and the impact of adolescent development. Although our conceptualization of risk has evolved, our study of risk has not kept pace. Few truly specific risk or protective factors for the development of eating disorders have been definitively established, and developmentally informed risk profiles have yet to be developed, tested, and validated. Knowledge concerning males and BED is particularly limited. It is important to clarify the developmental nature of BED and determine whether childhood presentations of loss-of-control eating are on the same continuum of pathology as BED in adulthood. The use of prospective designs should be encouraged. Such research methods are a potentially powerful means of elucidating the etiological processes that give rise to eating disorders.

Although a significant amount of evidence has been generated on prevention interventions since the last edition of this book, additional work is needed to extend the extant studies. Future research should investigate whether the risk factors for symptom onset differ from those for symptom escalation and symptom maintenance. This is important because the former are germane to the design of universal and selected prevention programs, but the latter are necessary for the design of optimally effective indicated-prevention programs and treatment interventions, respectively. Further, given the national and global efforts toward obesity prevention, it is also important that none of these efforts in any way encourages behaviors or attitudes that can increase risk not only for BED but for the other eating disorders as well.