REVIEW

New and emerging treatment options for adolescent bulimia nervosa

Daniel le Grange† & While bulimia nervosa (BN) was first described more than 25 years ago, and a substantial James Lock body of literature has been accumulated describing both psychosocial and †Author for correspondence pharmacological treatments for adults with this disorder, little is known about effective University of Chicago, Department of , interventions for adolescents with BN. The scope of this article is to review recent Section for Child and developments in the treatment of adolescent BN. We highlight our current knowledge of Adolescent Psychiatry, the epidemiology of this disorder among adolescents, emphasize the importance of early 5841 S. Maryland Ave., MC3077, Chicago, recognition of this disorder, and summarize the existing treatment literature for IL 60637, USA adolescent BN. The gaps in our knowledge regarding adolescents are accentuated by Tel.: +1 773 702 9277 referring, in contrast, to the efficacious treatment for adults with BN. We conclude by Fax: +1 773 702 9929 E-mail: legrange@ underscoring the need to conduct large-scale randomized trials testing efficacious uchicago.edu treatments for adolescent BN by borrowing from what is known in the treatment of adults with BN, as well as adolescents with anorexia nervosa.

Bulimia nervosa (BN) was first described by adolescents. Some studies have indicated that Gerald Russell more than 25 years ago [1]. Since between 2 and 3% of adolescents present with then, a substantial body of literature has been significant bulimic symptoms, even though established describing the treatment of adults these do not meet full threshold criteria [10,11]. with BN. By comparison, very little has been Moreover, a recent study demonstrated that accomplished in terms of either the clinical almost half (48%) of the adolescent patients description of adolescents with BN or how to who presented with BN symptoms at a tertiary treat adolescents with BN. The purpose of this treatment program did not meet full criteria for article is to succinctly summarize what is known BN [7]. The only difference between those who about the epidemiology of BN in adolescents, met full versus those who met partial criteria and to review new and emerging treatment for BN was that patients with BN reported sig- options for this clinical population. As substan- nificantly higher frequencies of objectively large tially more is known about treatment for adults binge eating and purging. with BN [2], we will make reference to this body BN among adolescents can be a source of psy- of literature in order to put the state of our chiatric morbidity and result in impaired func- knowledge for adolescent BN in perspective. tioning in many areas of life. The most common psychiatric include Epidemiology of adolescent with suicidality, disorders and substance bulimia nervosa abuse [12,13]. While physical health is often main- The key characteristics of BN are binge-eating tained despite binge eating and purging behav- episodes, followed by inappropriate compensa- iors, medical morbidities are not uncommon. tory behaviors, for example, self-induced vom- These include, but are not limited to, complica- iting, laxative or diuretic misuse, fasting and tions such as electrolyte disturbances, parotid excessive exercise. Feelings of loss of control, gland swelling, gastrointestinal (GI) symptoms guilt and remorse usually accompany these epi- such as GI perforation and loss of dental sodes of overeating [3]. As is the case with anor- enamel [14,15]. Keywords: adolescents, anorexia nervosa, bulimia exia nervosa (AN), patients with BN overvalue Outcome among adolescents with BN nervosa, cognitive–behavioral shape and weight, and often present with remains unknown. However, among adults, therapy, eating disorders, repeated attempts to lose weight [4,5]. Pre- approximately half of those with partial BN go family-based treatment, , pharmacotherapy, menarchal onset of BN is relatively rare [6], on to develop a full disorder. For psychosocial treatment while peak age of onset among adolescents is instance, Tozzi and colleagues found that between 16 and 18 years [7–9], with a point among patients with an initial diagnosis of BN, part of prevalence of approximately 1–2% [10,11]. Par- 27% crossed over to AN within 5 years of their tial BN cases are clinically common among first diagnosis, posing additional medical and

10.2217/14750708.4.6.841 © 2007 Future Ltd ISSN 1475-0708 Therapy (2007) 4(6), 841–849 841 REVIEW – le Grange & Lock

psychiatric risks [15]. In terms of prognosis for Early identification & early treatment adolescent BN, we will have to turn to the adult Physicians who see children and adolescents do data, as no information is available for the not always consider a diagnosis of an eating dis- younger age group. Relatively little is known order [27]. Consequently, there is often a delay about the natural course of eating disorders from the time that a child or adolescent presents among women in the community [16]. In a with clinically significant symptoms and is meta-analysis, Keel and Mitchell examined 88 referred to a specialty clinic. The detrimental outcome studies and demonstrated that 50% of impact of such a delay on the outcome in AN patients achieved full recovery, while another has been argued quite persuasively [28], as it has 20% continued to meet diagnostic criteria for been shown that there is little difference in sever- BN 5–10 years after the onset of the illness [17]. ity between prepubertal and postpubertal onset The remaining 30% of patients relapsed, in that AN [29]. In fact, Gowers et al. have argued that their bulimic symptoms reappeared. While prepubertal onset AN may represent a particu- treatment appears to be effective in the short larly pernicious form of the disorder, heighten- term, the long-term impact remains rather lim- ing the argument for early intervention [30]. The ited [18], and with substantial relapse rates in same argument also applies to subthreshold pre- some studies, the enduring benefits of evidence- sentations of AN. Partial AN can be associated based treatments for adults with BN should still with morbidities such as growth retardation, be viewed as tentative. pubertal delay and deficiencies in bone mineral acquisition [31]. Psychological distress and Detrimental impact of bulimia nervosa impairment can often be as significant and Eating disorders, and BN is no exception, are equivalent to levels seen in AN [32]. associated with significant burdens on several Information about early-onset BN and sub- domains, for example, personal, familial and soci- threshold presentations of BN is not available to etal. BN in adolescents, especially if it is the same degree that it is for AN. BN may be untreated, may persist into adulthood [19–21]. The more difficult to identify than early-onset AN as risk for the development of secondary physical bulimic symptoms are easier to hide, especially and mental disorders increases with increased given the guilt and shame associated with these duration of illness [22]. In a longitudinal birth symptoms. Notwithstanding, two recent cross- cohort study of 15 psychological disorders in sectional studies of adolescent eating disorders young adults aged 21 years, Lewinsohn et al. have demonstrated that approximately half of found that BN was among the disorders with the the referrals to a specialist adolescent service are highest levels of impairment [20]; 46% sought subthreshold cases of BN. The general and eat- treatment, 23% attempted suicide and 23% ing disorder profiles of the sub- reported use of psychotropic medication. The threshold cases were similar to their full- onset of illness was during adolescence for more syndrome BN counterparts [7,33]. In keeping than 90% of these cases. Similarly, Striegel-Moore with expectations, the difference between the and colleagues argue that psychosocial adjustment two groups was limited to significantly higher for young adult women who had BN as adoles- frequencies of binge eating and purging in the cents is compromised and associated with impair- full syndrome cases. This finding is somewhat ments at various levels, for example, health, self- misleading. While BN patients reported signifi- image and social functioning [23]. Moreover, the cantly more objectively large binge-eating epi- burden associated with BN extends beyond the sodes than subthreshold cases, the latter reported sufferer to the care provider, – that is, the parents almost double the amount of subjectively large or partners of patients. The care of an individual binge-eating episodes as compared with their with BN is likely to be as challenging and stressful BN counterparts. Consequently, the purge fre- as it is for the carers of individuals with AN. In a quency for both groups matches the combined large cohort of adolescents with BN, parents and binge frequency (objective plus subjective binge partners report significant practical, emotional episodes); this suggests that the salient binge fea- and interactional difficulties in relation to their ture may not be frequency but a sense of loss of young family member with BN [24,25]. Focusing control. The latter seems equally prevalent in on the mental health and caregiving experience of these diagnostic groups. Eating psychopathology relatives of adolescents with BN, Winn and col- was also similar for BN and subthreshold cases, leagues found that more than 50% reported men- suggesting that distress and impairment might tal health difficulties [26]. be equivalent between these two groups.

842 Therapy (2007) 4(6) futurefuture sciencescience groupgroup New and emerging treatment options for adolescent bulimia nervosa – REVIEW

The uncertainty about diag- NICE recommended that treatment modalities noses has been addressed by the American Acad- be graded A–C. Grade A implies strong empiri- emy of Pediatrics. The Academy has provided cal support from several well-conducted ran- diagnostic codes for clinically significant eating domized trials, while grade C implies expert disorder presentations in children and adoles- consensus. By far the majority of the more than cents that do not meet Diagnostic and Statistical 100 recommendations that were made received Manual of Mental Disorders 4th Edition only a grade C. The only treatment for adult BN criteria [34]. The Academy has also issued a policy to receive grade A was cognitive–behavioral ther- statement on the identification and treatment of apy (CBT). CBT for adult BN has a relatively eating disorders. This recent report urges pedia- solid body of research demonstrating its superi- tricians to be aware of the clinical and prognostic ority to any other psychological or pharmaco- significance of eating disorder-not otherwise logical treatment. More specifically, the NICE specified (ED-NOS) in adolescence [14]. While it guidelines recommend that the majority of seems appropriate not to differentiate between adults with BN should receive 16–20 sessions of full syndrome and ED-NOS BN cases, it CBT. As an alternative, especially when patients remains far from certain as to what the appropri- do not respond to CBT, a course of interpersonal ate treatment for this clinical population should (IPT) may be considered. How- be. In the following section, we will attempt to ever, patients should be informed that this treat- review recent advances that have been made ment might require more time in order to towards establishing efficacious treatments for achieve similar outcomes as CBT. The fact that adolescent BN. the NICE guidelines made no specific recom- As we have indicated in our review so far, mendation for the treatment of adolescents with adolescent BN has the potential for serious BN reflects the fact that, to date, systematic physical, emotional and social developmental research in the treatment of BN has focused consequences, including a chronic and severe almost exclusively on adults. This is true despite course [35], and might constitute the majority of the fact that both binge eating and purging begin those presenting for treatment [36]. It is also clear during adolescence, and that many cases of BN that treatment is probably quite complex in that start in adolescence [39,40]. it requires attention to several key aspects of the Significant progress has been made in under- disorder, such as psychiatric, medical, nutri- standing a range of efficacious treatments for tional and developmental aspects [37]. By far the adults with BN, including CBT, IPT and anti- majority of treatment inquiry has focused on depressant medications. By contrast, other than adults with BN. In fact, more than 70 con- case series data on CBT and two randomized, trolled treatment trials have been conducted in controlled trials (RCTs) that were recently com- adults with BN. By contrast, only two random- pleted [41,42], these treatments have not been ized trials for adolescents with BN have been studied with an adolescent population. The aver- concluded. An obvious dilemma is that findings age age of participation in research studies of from adult studies are not necessarily generaliz- these treatment approaches was 28.4 years, the able to children and adolescents. Consequently, duration of the disorder approximately little is know about efficacious treatments for 10 years [43–45], and the cutoff age for entry adolescents with BN. Therefore, to provide a 18 years. In the largest clinical trial of psycho- context for the limited adolescent data, our logical treatments for BN to date (n = 220) [43], review will include the most seminal work the mean age of participants was 28.1 years among the treatment trials for adults with BN. (standard deviation: 7.2) years. In the absence of substantial published data of established treat- Treatments for adolescent ments for adolescent BN, the main focus of our bulimia nervosa review is on treatments for adults with Clinicians have relatively few clinical treatment BN [43,46–49]. guidelines for eating disorders at their disposal. It was therefore particularly timely when the National Institute for Clinical Excellence Cognitive–behavioral therapy (NICE) in the UK recently took the meaningful The cognitive–behavioral model of BN assumes step to summarize guidelines for eating disorders that the maintenance of the disorder is based on based on a comprehensive review of the dysfunctional attitudes toward body shape and literature [38]. Upon completion of their review, weight. These beliefs lead to overvalued ideas of futurefuture sciencescience groupgroup www.futuremedicine.com 843 REVIEW – le Grange & Lock

thinness and increased body dissatisfaction, and helping the patient make specific changes in are typically followed by attempts to control identified ‘problem areas’. CBT was compared shape and weight by excessive dieting. This with IPT in a large multisite trial of adult patients excessive dieting causes a sense of both psycho- with BN [43]. In this trial, CBT was superior to logical and physiological deprivation, and some- IPT at the end of treatment, but at 1 year follow- times increases depressed moods. In addition, up, no differences were found between the two because of dietary restriction, hunger is increased treatments. These studies suggested that BN was and this, in turn, leads to an increased probabil- responsive to IPT as well as CBT, but that the ity of binge eating. Binge eating, because of the improvements associated with IPT were slower to fears of weight gain as a result of eating a large develop. IPT for adolescent BN is unexamined. amount of usually calorie-dense food, is eventu- ally followed by purging as an attempt to allay CBT for adolescent BN these [50]. CBT has been tested in There have been two published case series of numerous controlled studies and has been found adolescent subjects with BN treated with CBT to be the most effective psychotherapeutic designed for adolescent use (CBT-A), both at approach to the treatment of BN. CBT has been Stanford University in California, USA [55,56]. found to be more effective than other treat- These modifications included: ments, including no therapy, nondirective ther- • Increased contact between the therapist and apy, pill placebo, manualized psychodynamic the adolescent in early treatment to promote therapy (supportive–expressive), stress manage- therapeutic alliance; ment and treatment [43,46–49] . Of • Involvement of parents in supporting treatment; those completing CBT, approximately 40% of • Use of concrete examples to illustrate points; patients with BN are abstinent, while another 20% are much improved. Fairburn and col- • Exploration of adolescent developmental issues leagues at the University of Oxford (Oxford, (e.g., autonomy concerns) in the context of BN. UK) followed bulimic patients treated with CBT The results of these two case series provide for 5 years post-treatment [51]. Nearly 60% of preliminary evidence that CBT-A is acceptable patients studied had no eating disorder, and a and feasible as a treatment for adolescents with further 20% had a subclinical disorder. The BN. In addition, there is one RCT that was con- remainder were unrecovered, with a small per- ducted in London, UK, which compared cogni- centage diagnosed as having AN. Hence, relapse tive–behavioral guided self-care (CBT-GSC) rates for the successfully treated patients appear with [42]. CBT-GSC was manual- to be low, and the benefits are long-lasting. In a ized with an accompanying workbook for the meta-analysis of nine double-blind, placebo-con- patient. This treatment comprised of ten weekly trolled medication trials (870 subjects) and 26 sessions, three monthly follow-up sessions and randomized psychosocial studies (460 subjects), two optional meetings with a close other. The CBT for BN was found to produce significantly initial focus of the treatment is on the function larger-weighted pooled effect sizes for binge eat- that bulimic symptoms may have on the individ- ing (95% CI: 1.09–1.47), purge frequency (95% ual’s life. Self-monitoring is introduced followed CI: 1.06–1.39), depression (95% CI: by problem-solving exercises with behavioral 1.10–1.51) and eating attitudes (95% CI: experimentation. As is the case in regular CBT, 1.12–1.58) than comparison treatments [52], homework is a central part of this treatment. strongly supporting the view that CBT is the Results suggest that CBT-GSC was an acceptable treatment of choice for BN in adults. In addi- and feasible treatment for adolescent BN, with a tion, a recent study directly evaluated the CBT treatment drop-out rate of 29% and an absti- model using structural equation modeling and nence rate (from both binge eating and purging) found that the factors of self-esteem, overcon- at 6-month follow-up of 36%. These abstinent cern with weight and shape, and dietary restraint rates were comparable with those found in more accounted for 97% of the variance in outcome of recent adult studies of CBT [2]. It is noteworthy binge eating and purging [53]. These results pro- that CBT-A used in the Stanford studies includes vide support for the CBT model of BN. direct parental involvement, and is therefore a CBT has been compared with IPT, modified different model from the therapist-led guided for BN [54], which focuses on the interpersonal self-care model that was studied by the London context within which the eating disorder devel- group. It is instead more similar to typical CBT oped and is maintained, with the principal aim of used in treatment studies of adults with BN.

844 Therapy (2007) 4(6) futurefuture sciencescience groupgroup New and emerging treatment options for adolescent bulimia nervosa – REVIEW

Family-based treatment weight, treatment in BN is more collaborative Family therapy was first implemented for adoles- between the adolescent and her/his parents. cents with BN in a small case series [57]. This Third, the secretive nature of BN, as well as the study demonstrated significant reductions in guilt and shame that are caused by these symp- bulimic behaviors through educating the family toms, make it more probable that the illness, and about the eating disorder and helping the parents its severity, can be overlooked by the parents. In to disrupt binge-eating and purging episodes. AN, on the other hand, the patient’s emaciated More recently, le Grange and colleagues at the state is usually more obvious, and allows for the University of Chicago (IL, USA) [58] described therapist to help the parents remain focused on the progression of an adolescent in family-based the seriousness of her/his condition. Finally, treatment for BN (FBT-BN). FBT-BN is derived treatment in BN is more often complicated by from the approach found to be effective for ado- comorbid illnesses, which are probably more lescents with AN originally developed at the prevalent in BN than in AN. Maudsley Hospital in London [59–61], and subse- The first two RCTs for adolescents with BN quently manualized [62] and examined in treat- and partial BN have recently been concluded. ment studies in the USA [63,64]. although In the first of these, and as alluded to earlier in modified from the approach employed with ado- this review, the Maudsley group in London lescent AN, this treatment shares many charac- compared family therapy with CBT-GSC teristics with the original Maudsley family among participants aged 12–20 years treatment model and has recently been manual- (mean: 7.6 years; standard deviation: 0.3) [41]. ized [65]. For example, FBT-BN assumes that the At 6-month follow-up, no statistical differences secrecy, shame and dysfunctional eating patterns on abstinence rates were found between family associated with BN have negatively affected ado- therapy and individual CBT-GSC. lescent development in the bulimic patient and At around the same time, a controlled study have confused and disempowered parents and was conducted by a group at the University of other family members. In addition, parental Chicago. Participants, aged 12–19 years guilt about having possibly caused the illness and (mean: 16.1 years; standard deviaiton: 1.6), anxiety about how best to proceed have further were allocated to either manualized FBT-BN; disabled them. FBT-BN is agnostic as to the or individual supportive psychotherapy cause of BN, but assumes that the usual progress (SPT) [42]. At post-treatment and 6-month fol- through adolescence is negatively affected by the low-up, significantly more patients in FBT-BN disorder. To target these problems, FBT-BN were binge/purge abstinent compared with employs a three-stage treatment. In the first SPT. Family therapy as described by the Lon- stage, treatment aims at empowering parents to don group resembles FBT-BN, however, one disrupt binge eating, purging, restrictive dieting key difference between FBT-BN and family and any other pathologic weight-control behav- therapy is that ‘family’ is defined as any ‘close iors. It also aims to externalize and separate the other’, rather than just a parent. This definition disordered behaviors from the affected adoles- of family occurred in approximately a quarter cent to promote parental action and decrease of cases, and was likely utilized, in part, because adolescent resistance to their assistance. Once the mean age of the subjects in this study was abstinence from disordered eating and related almost 18 years. This is close to adulthood, behaviors has been achieved, the second stage of especially in the UK, where the age of consent treatment begins wherein parents transition con- is 16 years. While defining family as a close trol over eating and weight-related issues back to other may fit well with this older age group, the adolescent under their supervision. The third this might not be the most effective way to stage is focused on the ways the family can help approach FBT-BN with younger adolescents, to address the effects of BN on adolescent devel- where parental authority is key to the success of opmental processes, both on the adolescent and FBT-BN. This point is emphasized further by the family as a whole. the rate of treatment uptake for the UK study, Taken together, the FBT-BN approach differs wherein 25% of eligible participants refused the from that in adolescent AN in some significant study because they did not want their families ways. First, the emphasis is not on weight resto- involved in treatment [24]. By contrast, in the ration, but rather on regulating eating and cur- younger adolescents studied by the Chicago tailing purging. Second, whereas in AN the group, 11% dropped out of treatment and none parents take charge of restoring the adolescent’s of these reported involving the family as the futurefuture sciencescience groupgroup www.futuremedicine.com 845 REVIEW – le Grange & Lock

reason for discontinuing treatment [42]. None- with shape and weight also show greater theless, the abstinence rates (41%) for family improvement with medication than with pla- therapy in the Schmidt et al. [41] study were cebo [69]. Several controlled studies have directly comparable with those achieved dusing evaluated the relative and combined effectiveness FBT-BN in the Chicago study. Definition of of CBT and antidepressant drug treatment abstinence was equally strictly defined for both [44,68–70]. Although antidepressant medications these studies, that is, no binge eating and purg- have been shown to be more effective than a pla- ing for the 4 weeks preceding the assessment. cebo in reducing symptoms of BN [48,50], when Schmidt and colleagues acknowledge that a added to psychological treatments (e.g., CBT or limitation of their study was the sample size, IPT), medications did not improve the out- which was likely too small to detect differences comes of core eating-related symptoms. One between two active treatments for some of their small open-label study of ten adolescents, aged outcomes [41]. In addition, the authors acknow- 12–18 years, found that 8 weeks of fluoxetine at ledge that the absence of a waiting list or atten- 60 mg/day was well tolerated in the context of tion placebo-control group prevent them from supportive psychotherapy. While the medication ruling out that improvement was simply due to was well tolerated and the findings were encour- passage of time or nonspecific effects. The Chi- aging (approximately 70% were rated as either cago study, which had a similar overall sample much improved or improved), the effectiveness size to the London study, was adequately pow- of this medication in this age group is still ered to demonstrate the potential benefits of an unknown [71]. A familiar dilemma is that we are active treatment (i.e., FBT-BN) over a non- bound to extrapolate from adult data to adoles- specific control treatment (i.e., SPT), and to cents, who may in fact have a different clinical demonstrate comparable benefits that were not presentation [72]. Taken together, the data sug- due to time effects. gests that the use of in adults with BN, although useful, offers only a marginal Pharmacotherapy advantage over CBT alone. The available treat- Antidepressant medications in the treatment of ment studies for adolescent BN are summarized BN have received intense research attention in in Table 1. adults. A series of double-blind, placebo-con- trolled trials of antidepressant medications Conclusion among adults with BN have been conducted There are no evidence-based treatments for [44,66–68]. In almost all of these controlled trials, adolescent BN. Two emerging treatments for most types of antidepressants have proven supe- this condition are family treatment (family rior to placebo in reducing binge frequency. therapy or FBT-BN), refined from the adoles- Generally, disturbance and preoccupation cent AN treatment model of the same name,

Table 1. Psychosocial and medication studies for adolescent bulimia nervosa. Study Study type n Age* (years) Type of Ref. treatment

Dodge et al. (1995) Case series 8 16.5 (1.2) FT [57] le Grange et al. Case study 1 17 FBT-BN [58] (2003) Kotler et al. (2003) Open trial 10 12–18 Fluoxetine 60 mg [71] Lock (2005) Case series 34 15.8 CBT-A [63] Schapman et al. Case series 7 16.3 (1.3) CBT-A [54] (2006) Schmidt et al. RCT 85 17.6 (0.3) FT vs CBT-GSC [41] (2007) Le Grange et al. RCT 80 16.1 (1.6) FBT-BN vs SPT (2007) *Mean (standard deviation), actual years, age range or mean age with no SD. CBT-A: Cognitive–behavioral therapy for adolescents; CBT-GSC: Cognitive–behavioral guided self-care; FT: Family therapy; FBT-BN: Family-based treatment for bulimia nervosa; RCT: Randomized, controlled trial; SPT: Supportive psychotherapy.

846 Therapy (2007) 4(6) futurefuture sciencescience groupgroup New and emerging treatment options for adolescent bulimia nervosa – REVIEW

and CBT (self-guided or adjusted for an adoles- In the near future, more definitive studies of the cent age group, which includes a parental com- most promising treatments described here ponent, i.e., CBT-A). These treatments share a (FBT-BN and CBT-A) are needed as a first step developmental perspective on the needs of ado- in advancing treatments for adolescent BN. lescents. For instance, in FBT-BN, treatment is modified to account for the usually older age of Future perspective adolescents and the egodystonic nature of BN, A large RCT will establish the treatment of choice characteristics that typically differentiate ado- for both younger and older adolescents with BN. lescents with BN from those with AN. In CBT- In addition, the first combination trial of fluoxet- A, adjustments are made to account for varia- ine and psychotherapy might indicate whether tion in motivation, cognitive ability and special this selective serotonin re-uptake inhibitor is as needs for therapeutic alliance building. Little helpful in the treatment of adolescents with BN as evidence suggests that existing medications will it has been in the treatment for adults with BN. play a primary role in the treatment of adoles- cents with BN, especially given concerns about Financial & competing interests disclosure impulsivity or other forms of self-harm in this le Grange and Lock are both supported by a grant from the age group [73], and the potential for increased National Institute of Mental Health (USA) and both receive when treated with selective royalties from Guilford Press. The authors have no other rel- serotonin re-uptake inhibitors. evant affiliations or financial involvement with any organi- The need to develop an evidence base for zation or entity with a financial interest in or financial treatments for adolescents with BN is clear, conflict with the subject matter or materials discussed in the given the medical, psychological, social and manuscript apart from those disclosed. societal costs of this disorder over the lifetime of No writing assistance was utlilized in the production of a person who develops the disorder early in life. this manuscript.

Executive summary

• Treatment studies for adolescent bulimia nervosa (BN) are limited, with no established evidence base for this disorder in younger patients.

• Applying the findings from the substantial evidence base for adults with BN to a younger patient population is inappropriate.

• Family-based treatment and cognitive–behavioral therapy are promising treatments for adolescent BN.

• The role of antidepressant medication in the treatment of adolescent BN is unknown.

• Determining the relative efficacy of family-based treatment for BN and cognitive–behavioral therapy for adolescents is a priority.

• The American Psychiatric Association offers a thorough review of the treatment guidelines for eating disorders at [101].

Bibliography 6. Kent A, Lacey H, McClusky SE: 10. Kotler LA, Walsh BT: Eating disorders in 1. Russell GFM: Bulimia nervosa: An ominous Pre-menarchal bulimia nervosa. children and adolescents: pharmacological variant of anorexia nervosa. Psychol. Med. 9, J. Psychosom. Res. 36, 205–210 (1992). therapies. Eur. Child and Adolesc. Psychiatry 429–448 (1979). 7. le Grange D, Loeb KL, Orman S, Jellar C: 9, 108–116 (2000). 2. Mitchell J, Agras WS, Wonderlich S: Bulimia nervosa: a disorder in evolution? 11. van Hoeken D, Seidell J, Hoek HW: Treatment of bulimia nervosa: where are we Arch. Ped. Adoles. Med. 158, 478–482 Epidemiology. In: Handbook of Eating and where are we going. Int. J. Eat. Disord. (2004). Disorders, (2nd Edition). Treasure J, Schmidt 39, 95–101 (2006). 8. Fairburn CG, Cooper Z, Doll H et al.: The U, van Furth E (Eds). John Wiley & Sons, 3. American Psychiatric Association: Diagnostic natural course of bulimia nervosa and binge Chichester, USA, 11–34 (2004). and Statistical Manual of Mental Disorders eating disorder in young women. Arch. Gen. 12. Herzog DB, Keller MB, Sacks NR et al.: (4th Edition). American Psychiatric Psychiary 57, 659–665 (2000). Psychiatric in treatment- Association, Washington, USA (2000). 9. Flament M, Ledoux S, Jeammet P et al.: seeking anorexics and bulimics. J. Am. Acad. 4. Fairburn CG, Harrison PJ: Eating disorders. A population study of bulimia nervosa and Child Adolesc. Psychiatry 31(5), 810–818 Lancet 361, 407–416 (2003). subclinical eating disorders in adolescence. (1992). 5. Steiger H, Séguin JR: Eating disorders: In: Eating Disorders in Adolescence: 13. Herzog DB, Nussbaum KM, Marmor AK: anorexia nervosa and bulimia nervosa. Anorexia and Bulimia Nervosa. Comorbidity and outcome in eating In: Oxford Textbook of Psychopathology. Millon Steinhausen H (Ed.). Brunner/Mazel, NY, disorders. Psychiatr. Clin. North Am. 19(4), T, Blaney PH (Eds). Oxford University Press, USA, 21–36 (1995). 843–859 (1996). NY, USA, 365–389 (1999).

futurefuture sciencescience groupgroup www.futuremedicine.com 847 REVIEW – le Grange & Lock

14. Golden NH, Katzman DK, Kreipe RE et al.: 26. Winn S, Perkins S, Walwyn R et al.: 39. Lock J, Reisel B, Steiner H: Associated Eating disorders in adolescents: position Predictors of mental health problems and health risks of adolescents with disordered paper of the Society for Adolescent Medicine. negative caregiving experiences in careers of eating: how different are they from their J. Adolesc. Health 33, 496–503 (2003). individuals with bulimia nervosa. Int. J. Eat. peers? Results from a high school survey. 15. Tozzi F, Thornton LM, Klump KL et al.: Disord. 40, 171–178 (2007). Child Psychiatry Hum. Dev. 31, 249–265 Symptom fluctuation in eating disorders: 27. Bryant Waugh RJ, Lask BD, Shafran RL, (2001). correlate of diagnostic crossover. Am. Fosson AR: Do doctors recognise eating 40. Stice E, Agras WS: Predicting onset and J. Psychiatry 162, 732–740 (2005). disorders in children? Arch. Dis. Child. 67, cessation of bulimic behaviors during 16. Newman DL, Moffitt TE, Caspi A et al.: 103–105 (1992). adolescence. Behav. Ther. 29, 257–276 Psychiatric disorder in a birth cohort of 28. le Grange D, Loeb KL: Early identification (1998). young adults: prevalence, comorbidity, and treatment of eating disorders: prodrome 41. Schmidt U, Lee S, Beecham J et al.: clinical significance, and new case incidence to syndrome. Early Interv. Psychiatry 1, A randomized controlled trial of family from ages 11 to 21. J. Consult. Clin. Psychol. 27–39 (2007). therapy and cognitive behavioral guided 64, 552–562 (1996). 29. Cooper PJ, Watkins B, Bryant Waugh R, self-help for adolescents with bulimia 17. Keel PK, Mitchell JE: Outcome in bulimia Lask B: The nosological status of early onset nervosa and related conditions, Am. nervosa. Am. J. Psychiatry 154, 313–321 anorexia nervosa. Psychol. Med. 32, 873–880 J. Psychiatry 164, 591–598 (2007). (1997). (2002). 42. le Grange D, Crosby R, Rathouz P, 18. Klein DA, Walsh BT: Eating disorders. Int. 30. Gowers SG, Crisp AH, Joughin N, Bhat A: Leventhal B: A randomized controlled Rev. Psychiatry 15, 205–216 (2003). Premenarcheal anorexia nervosa. J. Child comparison of family-based treatment and 19. Striegel-Moore RH, Leslie D, Petrill SA Psychol. Psychiatry 32, 515–524 (1991). supportive psychotherapy for adolescent et al.: One-year use and cost of inpatient and 31. Bachrach LK, Gudo D, Katzman DK: bulimia nervosa. Arch. Gen. Psychiatry 64, outpatient services among female and male Decreased bone density in adolescent girls 1049–1056 (2007). patients with an eating disorder: evidence with anorexia nervosa. Pediatrics 86, 43. Agras WS, Walsh BT, Fairburn CG et al.: from a national database of health insurance 440–447 (1990). A multicenter comparison of claims. Int. J. Eat. Disord. 27, 381–389 32. Bunnell DW, Shenker IR, Nussbaum MP, cognitive-behavioral therapy and (2000). Jacobson MS: Subclinical versus formal interpersonal psychotherapy for bulimia 20. Lewinsohn PM, Striegel-Moore, Seeley JR: eating disorders: differentiating nervosa. Arch. Gen. Psychiatry 57, 459–466 Epidemiology and natural course of eating psychological features. Int. J. Eat. Disord. 9, (2000). disorders in young women from adolescence 357–362 (1990). 44. Agras WS, Rossiter EM, Arnow B et al.: to young adulthood. J. Am. Acad. Child 33. Binford R, le Grange D: Adolescents with Pharmacologic and cognitive–behavioral Adolesc. Psychiatry 39, 1284–1292 (2000). bulimia nervosa and eating disorder not treatment for bulimia nervosa: a controlled 21. Kotler L, Cohen P, Davies M et al.: otherwise specified-purging only. Int. J. Eat. comparison. Am. J. Psychiatry 149, 82–87 Longitundinal relationships between Disord. 38, 157–161 (2005). (1992). childhood, adolescent and adult eating 34. Wolraich ML, Felice ME, Drotar D: 45. Wilson GT, Eldredge KL, Smith D, Niles disorders. J. Am. Acad. Child Adolesc. The Classification of Child and Adolescent B: Cognitive–behavioral treatment with Psychiatry 40, 1434–1440 (2001). Mental Diagnoses in Primary Care: and without response prevention for 22. Johnson JG, Cohen P, Kotler L, Kasen S, Diagnostic and Statistical Manual for Primary bulimia. Behav. Res. Therapy 29, 575–583 Brook JS: Psychiatric disorders associated Care (DSM-PC) Child and Adolescent (1991). with risk for the development of eating Version. American Academy of Pediatrics, 46. Agras WS, Telch CF, Arnow B et al.: disorders during adolescence and early IL, USA (1996). One-year follow-up of cognitive–behavioral adulthood. J. Consult. Clin. Psychol. 70, 35. Steiner H, Lock J: Anorexia and bulimia therapy for obese individuals with binge 1119–1128 (2002). nervosa in children and adolescents: a review eating disorder. J. Cons. Clin. Psychology 65, 23. Striegel-Moore RH, Seeley JR, of the past 10 years. Child Adolesc. Psychiatry 343–347 (1997). Lewinsohn PM: Psychosocial adjustment in 37, 352–359 (1998). 47. Fairburn CG, Jones R, Peveler RC et al.: young: adulthood of women who 36. Currin L, Schmidt U, Jick H, Treasure J: Psychotherapy and bulimia nervosa. experienced an eating disorder during Time trends in the incidence of bulimia Longer-term effects of interpersonal adolescence. Am. Acad. Child Adolesc. nervosa. Br. J. Psychiatry 186, 132–135 psychotherapy, behavior therapy, and Psychiatry 42, 587–593 (2003). (2005). cognitive behavior therapy. Arch. Gen. 24. Perkins S, Winn S, Murray J, Murphy R, 37. Yager J, Andersen A, Devlin M, Mitchell J, Psychiatry 50, 419–428 (1993). Schmidt U: A qualitative study of the Powers P, Yates A: American Psychiatric 48. Pope HG, Hudson JI, Jonas JM experience of caring for a person with Association practice guidelines for eating Yurgelin-Todd D: Bulimia treated with bulimia nervosa. Part 1: The emotional disorders. Am. J. Psychiatry 15, 201–228 impramine: a placebo-controlled, double- impact of caring. Int. J. Eat. Disord. 36, (1993). blind study. Am. J. Psychiatry 140, 554–558 256–268 (2004). 38. National Collaborating Centre for Mental (1993). 25. Winn S, Perkins S, Murray J, Murphy R, Health: Eating Disorders: Core Interventions 49. Wilfley DE, Agras WS, Telch CF et al.: Schmidt U: A qualitative study of the in the Treatment and Management of Group cognitive–behavioral therapy and experience of caring for a person with Anorexia Nervosa, Bulimia Nervosa and group interpersonal psychotherapy for the bulimia nervosa. Part 2: Carers’ needs and Related Eating Disorders. British nonpurging bulimic: a controlled experiences of services and other support. Psychological Society and Gaskell, Leicester comparison. J. Cons. Clin. Psychology 61, Int. J. Eat. Disord. 36, 269–279 (2004). & London, UK (2004). 296–305 (1993).

848 Therapy (2007) 4(6) futurefuture sciencescience groupgroup New and emerging treatment options for adolescent bulimia nervosa – REVIEW

50. Apple RF, Agras WS: Overcoming Eating 58. le Grange D, Lock J, Dymek M: Family- 67. Walsh BT, Hadigan CM, Devlin MJ et al.: Disorders: A Cognitive-Behavioral Treatment based therapy for adolescent with bulimia Long-term outcome of antidepressant for Bulimia Nervosa and Binge Eating nervosa. Am. J. Psychother. 67, 237–251 treatment for bulimia nervosa. Am. Disorder. Client Workbook. (2003). J. Psychiatry 148(9), 1206–1212 (1991). The Psychological Corporation,TX, USA 59. Russell GF, Szmukler GI, Dare C, Eisler I: 68. Walsh BT, Wilson GT, Loeb KL et al.: (1997). An evaluation of family therapy in anorexia Medication and psychotherapy in the 51. Fairburn CG, Norman PA, Welch SL et al.: nervosa and bulimia nervosa. Arch. Gen. treatment of bulimia nervosa. Am. A prospective study of outcome in bulimia Psychiatry 44, 1047–1056 (1987). J. Psychiatry 154, 523–531 (1997). nervosa and the long-term effects of three 60. Eisler I, Dare C, Hodes M et al.: Family 69. Mitchell JE, Raymond N, Specker S: psychological treatments. Arch. Gen. therapy for adolescent anorexia nervosa: A review of the controlled trials of Psychiatry 52, 304–312 (1995). the results of a controlled comparison of two pharmacotherapy and psychotherapy in the 52. Whittal ML, Agras WS, Gould RA: Bulimia family interventions. J. Child Psychol. treatment of bulimia nervosa. Int. J. Eat. nervosa: a meta-analysis of psychosocial and Psychiatry 41, 727–736 (2000). Disord. 14(3), 229–247 (1993). pharmacological treatments. Behav. Ther. 61. le Grange D, Eisler I, Dare C, Russell G: 70. Leitenberg H, Rosen JC, Wolf J et al.: 30, 117–135 (1999). Evaluation of family treatments in Comparison of cognitive–behavior therapy 53. Byrne S, McLean N: The adolescent anorexia nervosa: a pilot study. and desipramine in the treatment of bulimia cognitive-behavioral model of bulimia Int. J. Eat. Disord. 12(4), 347–357 (1992). nervosa. Behav. Res. Therapy 32(1), 37–45 nervosa: a direct evaluation. Int. J. Eat. 62. Lock J, le Grange D, Agras WS, Dare C: (1994). Disord. 31, 17–31 (2002). Treatment Manual For Anorexia Nervosa: 71. Kotler L, Devlin B, Davies M, Walsh BT: 54. Lock J: Adjusting cognitive behavioral A Family-based Approach. Guildford Press, An open trial of fluoextine in adolescents therapy for adolescent bulimia nervosa: NY, USA (2001). with bulimia nervosa. J. Child Adolesc. results of a case series. Am. J. Psychother. 59, 63. Lock J, Agras WS, Bryson S, Kraemer H: 13, 329–325 (2003). 267–281 (2005). A comparison of short- and long-term 72. le Grange D, Schmidt U: The treatment of 55. Fairburn CG: Interpersonal psychotherapy family therapy for adolescent anorexia adolescents with bulimia nervosa. J. Mental for bulimia nervosa. In: Handbook of nervosa. J. Am. Acad. Child Adolesc. Health 14, 587–597 (2005). Treatment for Eating Disorders (2nd Edition). Psychiatry 44, 632–639 (2005). 73. Fischer S, le Grange D: Co-morbidity and Garner DM, Garfinkel P (Eds). Guilford 64. le Grange D, Binford R, Loeb KL: high-risk behaviors in treatment seeking Press, NY, USA (1997). Manualized family-based treatment for adolescents with bulimia nervosa. Int. J. Eat. 56. Schapman A, Lock J, Couturier J: anorexia nervosa: a case series. J. Am. Acad. Disord. (2007) (In Press). Cognitive–behavioral therapy for adolescents Child Adolesc. Psychiatry 44, 41–46 (2005). with binge eating : a case series. 65. le Grange D, Lock J: Treating Bulimia in Website Int. J. Eat. Disord. 39, 252–255 (2006). Adolescents: A Family-Based Approach. 101. www.psych.org/psych_pract/treatg/pg/Eating 57. Dodge E, Hodes M, Eisler I, Dare C: Family Guilford Press, NY, USA (2007). Disorders3ePG_04–28–06.pdf therapy for bulimia nervosa in adolescents: 66. Agras WS, McCann U: The efficacy and an exploratory study. Am. J. Fam. Ther. 17, role of antidepressants in the treatment of 59–77 (1995). bulimia. Ann. Behav. Med. 9, 18–22 (1987).

futurefuture sciencescience groupgroup www.futuremedicine.com 849