Clinical Context
Eating disordered thoughts and behaviors onset primarily in the adolescent years (
1,
2). Anorexia Nervosa (AN) has a peak incidence rate at approximately 15 years of age and it is rare for onset to occur after the age of 25 years (
3,
4). Although the incidence of Bulimia Nervosa (BN) peaks in the late teenage years, disordered eating, including binge eating and purging at subthreshold rates, begins in the middle teenage years (
1,
5,
6). Binge eating often begins in childhood and adolescence (
7–
9), but Binge Eating Disorder (BED) is not typically diagnosed until early and middle adulthood (
10,
11). These epidemiology data relating age and eating disordered behavior underscore the importance of understanding the presentation and treatment of eating disorders in youth. Nonetheless, until relatively recently, the need to focus on this younger age group had not been a priority.
Despite the description of AN in the medical literature for several hundred years (
12,
13), it was not until 1987 that the first randomized clinical trial (RCT) examining treatments for adolescents with AN was published (
14). There was not a subsequent completed RCT focused on treatment for adolescents with AN published for another decade (
15). While there are a number of effective treatments for adults with BN based on a relatively large evidential base (
16,
17), for adolescent BN, there remain only two RCTs examining treatments, both published in 2007 (
18,
19). As for BED, there are limited evaluating treatments for youth (
20–
23) despite a comparatively robust number of studies in adults with this disorder (
24–
30).
Diagnostic Issues
One reason for the lack of emphasis on youth with disordered eating is related to the diagnostic criteria of former iterations of the Diagnostic and Statistical Manual of Mental Disorders (DSM). The criteria used for diagnostic categorization were based primarily on adult phenotypic symptom presentation without regard for symptom evolution or developmental variation (
31). This led to delays in the diagnosis of children and adolescents with eating disorders (
32). Reasons for failure to meet DSM diagnostic criteria included inaccurate strategies for assessing weight (
33–
35), differences in ability to verbalize internal states and motivation (
36), inappropriate behavioral threshold or duration of behaviors (
37), and the failure to utilize parental report in the evaluation of younger patients (
37–
39). Some of these deficiencies are corrected in DSM-5, especially for adolescent AN. In addition to these changes, the authors of DSM-5 have added a new diagnosis, Avoidant Restrictive Food Intake Disorder (ARFID) that encompasses a range of eating problems not associated with disordered thinking about appearance and weight (
40). Taken together, these changes will likely reduce the number of patients who previously were diagnosed in the vague Eating Disorder Not Otherwise Specified (EDNOS) group and (
41,
42) likely make it easier for clinicians to more accurately diagnose children and adolescents with eating disorders. Nonetheless, the ability to understand the interface of eating disorder symptoms and developmental variation in the context of cognitive, social, and emotional development is a refined skill requiring rather extensive training and experience.
In addition to the diagnostic challenges in youth with eating disorders, as suggested above, our understanding of effective treatments for younger patients is limited by the state of current research in the field. There are 11 randomized controlled trials (RCTs) that included a total of approximately 1000 adolescent participants 19 years of age or younger published for AN (
14,
15,
43–
50). In contrast to studies of BN for adults where there are many studies available (
16,
17,
51), there are only two published RCTs examining treatment for adolescents with BN. These studies include 165 adolescent participants and examined FBT for BN, CBT (self-help), and individual therapy (supportive) (
18,
19). Similarly, treatment of BED in adults has received considerable attention in RCTs (
10,
23,
24,
26,
29), but BED is relatively unexamined in adolescents (
20,
21,
52). As noted, ARFID is a new diagnosis in DSM-5, but there are no empirical studies to guide treatment for this disorder. Thus, the database is limited and conclusions about the most effective treatment approaches are tentative.
Psychosocial Treatments For Anorexia Nervosa In Adolescents
The best evidence-based treatment for adolescents with AN is a specific form of family therapy, called Family-Based Treatment (FBT). The approach was developed at the Institute of Psychiatry at the Maudsley Hospital in London and manualized at Stanford University and The University of Chicago (
53–
55). In contrast to historical approaches, which typically excluded parents by hospitalizing patients (
13) or focusing individual therapy on autonomy struggles and self-efficacy (
56,
57), FBT promotes parental management of eating disordered behaviors in their children early in treatment until the child is better able to manage his or her behaviors in at an age-appropriate level of independence. Of the 11 published studies referenced above for adolescent AN, 7 of them examined FBT (
14,
15,
43,
44,
48,
58). Although the earliest studies suffered from a range of methodological limitations, taken together the results of these support the view that FBT is an effective approach for the outpatient management of adolescents with short duration AN (
51,
59). Further, the results suggest that compared with other approaches so far studied, FBT works quickly (
60), reduces the need for hospitalization (
44,
61), produces higher rates of recovery (
14,
48), and lowers costs (
62). Despite these encouraging findings, FBT has not been rigorously compared with many other potentially effective treatments for adolescents with AN, including cognitive behavior therapy (CBT) (
66), interpersonal psychotherapy (
64), or dialectical behavioral therapy (
65), which are all potentially useful approaches.
In addition to FBT, other family approaches have been of recent research interest. One study from France compared treatment as usual (TAU) to TAU plus a family therapy targeting family psychological process (
50). In contrast to FBT, which views the family as a potentially effective resource to help their children with AN, this type of family therapy does not promote familial management of the symptoms of AN, but instead aims to address family conflicts and communication style. The study included 60 randomized female participants (30 per group) between 13 and 19 years of age. There were no statistically significant differences in outcomes between the two randomized groups using an intent-to-treat analysis, but in a completer sample, the group that received TAU plus family therapy gained more weight and had higher rates of menstrual return. The study provides limited, but systematic support for the potential utility of a type of family therapy targeting family processes rather than symptom management. Another recent study of 164 adolescents with AN directly compared FBT to a manualized form of systemic family therapy (SFT) (
66) in a multisite randomized protocol.
Moderators of outcome are of interest because any single treatment is likely to not be effective for all patients (
67). To identify moderators of treatment effects, it is necessary to compare outcomes based on baseline characteristics of participants randomized to different treatments (
68). There are few studies of moderators of treatment effect in adolescents with AN. However, one study examined whether there were different patient groups that needed more or less FBT based on their baseline characteristics (
44). Participants in this study who came from single parent or reconstituted families (remarried) did better if they received more therapy. Also, those participants with higher levels of obsessive-compulsive features related specifically to eating did better with more treatment. Another study that compared individual therapy to family therapy found that three markers of severity (level of eating related psychopathology, level of obsessive compulsive thinking about eating, and purging behaviors) moderated outcome with patients who received FBT doing better (
69). The authors concluded that those adolescents who were most severely ill needed the help of the parents to recover, while those with milder symptoms could be helped just as well in individual therapy.
While most studies of adolescent AN have focused on outpatient treatment, one important study examined the potential benefits of a highly specialized psychiatric inpatient treatment program for this age group in a large RCT. Hospitalization for the acute physiological management of the consequences of severe malnutrition is necessary (
70), but there is an ongoing debate about whether inpatient weight restoration is needed (
46,
71). Early studies found conflicting evidence, but in the first RCT to examine this question for adolescents with AN, Gower and colleagues randomized 167 adolescents (aged 12–18 years) with AN to a specialized eating disorder inpatient treatment program, outpatient CBT, or TAU at the local mental health program (
47). Overall 33% of the participants had a good clinical outcome while 27% still met criteria for full syndrome AN at follow-up. Those participants who received the highly specialized long-term inpatient treatment (about 16 weeks) had no better outcomes than those who received weekly outpatient therapy. This led the authors to conclude that there was no advantage to utilizing a specialized inpatient eating disorder treatment program as an initial treatment for adolescents with AN. Further, either of the other outpatient approaches was superior in terms of cost-effectiveness (
72). It is important to note that this study did not conclude that hospitalization for weight restoration is not effective or not needed, but suggests on average, that it is not likely to lead to superior outcomes. There is no doubt that some patients need and benefit from inpatient weight restoration programs, but we do not yet know how best to identify for whom this costly treatment is needed.
Psychosocial Treatments For Bulimia And Binge Eating Disorder In Adolescents
Turning to BN in adolescents, the two published studies find differing results. In one study, FBT was compared with individual supportive therapy. The study employed a randomized design that included 80 adolescents (mean age 16 years) with BN or clinically significant binge eating and purging (
19). At the end of treatment and at the one year follow-up, FBT was statistically superior to this comparison treatment. This study also examined moderators and found that adolescents with less severe eating-related psychopathology before starting treatment did better in FBT (
73). This finding contrasts with the finding above related to adolescents with AN for whom FBT was superior when they showed evidence of greatest symptomatic severity (
69). The second published RCT randomized 85 adolescents and young adults (mean age about 18 years) with BN and/or binge eating to either family therapy or a self-help version of CBT (
18). There were no differences in outcome between the groups. Overall about 35% no longer reported binge eating or purging. Based on these two studies, there is insufficient data to support recommending a specific therapy for adolescent BN at this time. Both family and cognitive behavioral interventions appear helpful for about a third of treatment-seeking adolescents with this disorder.
It is also important that treatments that are found to be effective in RCTs and in specialized clinical settings be studied in more generalized clinical settings. Research into dissemination and implementation of treatments for eating disorders is in the early stages. The results examining the viability of FBT for dissemination and implementation in diverse settings are promising, however. FBT has been implemented in diverse academic medical centers in the U.S. (
74) as well as in treatment programs in the U.K., Brazil, Australia, New Zealand, and Canada (
61,
74–
77). In addition, two broader implementation efforts, both in Australia, suggest that the benefits of FBT led to systematic decreases in hospitalization use. At a children’s hospital in Sydney readmission rates of adolescents with AN dropped by 50% as FBT was implemented (
64). Similarly, at another children’s hospital in Melbourne, hospitalization admissions decreased by 56% and readmissions decreased by 75%. There was also a 51% decrease in overall hospital days (
78). However, these improvements were accomplished only by overcoming some significant challenges, particularly because initially therapists expressed concerns about the FBT model and its impact on professional roles. Specifically, the more circumscribed role of registered dieticians and medical providers in FBT was initially an uncomfortable fit for many therapists. These findings echo similar reports from a qualitative report from Canada examining therapist attitudes toward FBT (
79). While these preliminary studies are promising, most aspects of dissemination and implementation remain unexplored, including barriers to adopting evidence based treatments other than FBT as well as identifying effective strategies for efficient training and also skills maintenance strategies.
Future developments in treating adolescents with eating disorders
It is interesting to note that two treatments that have been used successfully in adults with a range of eating disorders—CBT and interpersonal psychotherapy (IPT)—have received little systematic study in adolescents. CBT has the most substantial database for the treatment of BN and BED, but there are only case series data available for the use of this treatment in adolescents (
63,
84). However, data from these cases series suggest that CBT may be useful for adolescents with AN and BN. Similarly, only case series data are available for IPT (
52), a therapy that focuses on specific interpersonal problems associated with the onset or maintenance of eating disorder symptoms. IPT has been studied in adults with BED and piloted with adolescents with BED (
20,
21,
29).
Some of the most exciting future developments for eating disorder treatment in youth are likely to arise from novel findings related to the neurobiological basis of these disorders (
87). It is possible that medications, psychosocial treatments, and cognitive treatments will be informed by a better understanding of the neural underpinnings of these disorders. Studies suggest underlying differences in the brain activation associated with the neurocircuitry related to cognitive inhibition (
86,
87), emotion regulation (
88), and reward processing (
89,
90). Treatments targeting cognitive and emotional regulation processes in patients with eating disorders are being developed. Cognitive Remediation Therapy (CRT) for eating disorders aims to improve perseverative and overly detailed thinking styles associated with them through cognitive exercises (
91–
95). The use of CRT is helpful for treating brain injury, schizophrenia, and OCD (
96–
98). These exercises do not focus on eating or weight, but instead target general cognitive style rather than content. Only one published RCT examining CRT included adolescents (over the age of 16 years) and adults (
99). The study compared CRT to CBT over two months and found that treatment dropout was lower in CRT and there were greater changes in both cognitive flexibility and detailed processing in those randomized to CRT. However, there were no differences in any eating disorder related behaviors or psychopathology between the two treatments at the end of treatment, despite these being the targets of CBT.
Dialectical Behavior Therapy (DBT) is another experimental psychosocial treatment of eating disorders and targets emotion regulation (
30). The DBT model conceptualizes eating disordered behaviors as maladaptive attempts to cope with overwhelming emotions. By overcoming the challenge of difficulties in emotion regulation, general self-regulatory behaviors (
100) broadly defined will also improve (
101,
102). In this sense, DBT attempts to help patients to modulate affect to prevent an eating disordered behavior (e.g., a binge episode) rather than targeting behavioral and cognitive strategies to change dietary restraint (as in CBT). In the context of eating disorders in youth, DBT may be particularly helpful because self-regulatory skills are linked to frontal-lobe development during adolescence and young adulthood and coincide with highest risk period for the development of eating disorders (
101,
103,
104). The results of studies in adults with BED and BN provide preliminary support for the clinical use of DBT (
28,
65,
105,
106); however, systematic studies using DBT in an adolescent sample with eating disorders are not yet available.
There has been considerable progress in the clinical management of eating disorders in adolescence in the past decade. DSM-5 has been revised to be more sensitive to the clinical importance of the early presentations of eating disorders, which should lead to earlier identification, more accurate diagnoses, and more rapid intervention for this younger population. In addition, new systematic studies of treatments now support the use of a particular form of family therapy (FBT) for adolescents with AN. Some systematic data support the use of other forms of family therapy and in some instances individual therapy for this disorder in teenagers. The role of hospitalization has been examined and data suggest that for adolescents with AN, hospitalization for weight restoration is not superior to outpatient treatments in most instances. However, where access to quality outpatient care is lacking, and in some instances even where excellent outpatient care is applied, inpatient care may be necessary. It is known that some patients need and benefit from inpatient weight restoration programs, but we do not yet know how best to identify for whom this costly treatment is needed. For adolescents with BN and BED, systematic investigations of psychosocial treatments are few and there is no substantial evidence base supporting any specific treatment. However, preliminary studies support the clinical effectiveness of FBT, CBT, DBT, and IPT for these disorders, but systematic support is lacking. Similarly, psychopharmacologic interventions have not been shown in systematic studies to be useful in adolescent eating disorders, but their clinical use may be indicated in refractory cases and when there is significant psychiatric comorbidity. Future advances in clinical diagnosis and treatment of eating disorders in adolescents will likely result from ongoing and novel studies of the neural underpinnings and psychological cognitive processes associated with these disorders.