Imp. 69-73

SPECIAL ARTICLE
The multimodal treatment of eating disorders
KATHERINE A. HALMI
Eating Disorder Program, Weill Medical College of Cornell University, New York Presbyterian Hospital, Westchester Division, White Plains, NY 10605, USA The treatment of eating disorders is based on a multimodal model, recognizing that these disorders do not have a single cause or a pre-dictable course. The treatment strategy is determined by the severity of illness and the specific eating disorder diagnosis. For the treatmentof anorexia nervosa, the key elements are medical management, behavioral therapy, cognitive therapy and family therapy, while phar-macotherapy is at best an adjunct to other therapies. In bulimia nervosa, the treatment of choice is cognitive-behavioral therapy, but agreater improvement in mood and anxiety occurs when antidepressant therapy is added. In binge eating disorder, cognitive-behavioraltherapy and interpersonal therapy produce substantial and long-lasting changes and pharmacological treatment has often a useful role. Key words: Eating disorders, multimodal treatment, cognitive-behavioral therapy, pharmacotherapy, nutritional rehabilitation
The treatment of eating disorders is based on a multi- one is at a greater risk for developing an eating disorder.
modal model. This model recognizes that eating disorders Individual biological factors include being mildly over- do not have a single cause or a predictable course. They begin with dieting or restrained eating behavior. Often the Genetic studies have shown a significant linkage on dieting is for the purpose of becoming thinner and more chromosome 1 for restricting anorexia nervosa (3) and a attractive, or it may follow a severe stress or physical ill- significant linkage on chromosome 10 for bulimia nervosa ness. Behaviors and influences antecedent to the dieting (4). There have been no consistent and replicated findings experience can be categorized as problems of biological for polymorphisms of specific genes with association stud- vulnerability, psychological predispositions, family distur- ies. Possible genetic vulnerabilities include a predisposi- bances, and environmental-societal influences. The inte- tion to a particular personality type, predisposition to a grated effect of these disturbances on dieting behavior pro- psychiatric disorder (affective or anxiety disorders) or pre- pels the individual person into developing an eating disor- disposition to neurotransmitter dysfunction. Thus, a ge- der. As the dieting continues, starvation effects, weight netic predisposition and vulnerability may become manifest loss, nutritional effects, and psychological changes occur.
under adverse conditions such as inappropriate dieting A sustaining cycle of core dysfunctional eating behaviors develops with both psychological and physiological rein- Biological vulnerability can include dysfunction of neu- rotransmitters such as serotonin, dopamine and norepi- For the anorexia nervosa patient, psychological rein- nephrine that regulate eating behavior. Studies have shown forcement occurs as the patient develops a sense of securi- that all three of these neurotransmitters are dysfunctional ty and effectiveness while she realizes that dieting and los- in eating disorder patients (5-7). Aberrations of neuropep- ing weight is something she can control very well with tides regulating eating behavior are also present in eating consummate skill. Physiological changes in the dopamin- disorders. Such changes are present in neuropeptide Y, opi- ergic, serotonergic and opioid neurotransmitter systems oids, leptin, cholecystokinin, ghrelin, melanocortins, adipo- most likely aid in reinforcing the starvation behavior.
nectin, agouti-related protein, and brain derived neuro- For bulimia nervosa, the psychological reinforcement comes about when the patient realizes that binge eating Signals from the periphery, including gut-related pep- can alleviate anxiety in a manner similar to alcohol or tides and adipokines, interact with hypothalamic peptides drugs. During the process of self-induced purging, do- in the regulation of feeding behavior and body weight.
pamine is released in the brain and likely contributes to State related changes in nutrition and body weight influ- the physiological reinforcement of the binge/purge behav- ence cerebrospinal fluid and plasma levels and receptor Individual psychological risk factors include a perfec- tionistic-obsessional personality, which is a risk especially RISK FACTORS
for the restricting type of anorexia nervosa (10). Low self- There are several different categories of risk factors for esteem and a sense of ineffectiveness, lack of confidence developing eating disorders. Familial risk factors have and a feeling of inadequacy are risk factors for anorexia been demonstrated with family studies. If one has a family nervosa and bulimia nervosa (11). Affective disorders member or relative with anorexia nervosa, bulimia nervosa (depression), alcohol and drug abuse are risk factors espe- or obesity, then one has a greater chance of developing an eating disorder (1). Also, if one has a family member or rel- Individual behavior such as dieting or involvement in ative with depression or alcohol/drug abuse/dependence, activities or professions that emphasize weight control – such as gymnastics, ballet, wrestling, jockeys, actors and hasten weight gain without deleterious effects. Supple- models – are additional risk factors (13). mental nocturnal nasal gastric refeeding has been used for In addition to the biological and psychological risk fac- better short-term outcome in hospitalized adolescent girls tors, there is the influence of social climate. This includes with anorexia nervosa. Other centers have used voluntary cultural risk factors such as living in an industrialized nasal gastric tube feeding and have taught patients how to country, emphasis on thinness as being beautiful, and in insert their own tube. Adequate controlled trials have not the U.S. a significant general weight increase over the past been conducted for this type of intervention. 40 years (14). Stressful life events, such as death of a close For individual psychosocial interventions, there contin- relative or friend or sexual and physical abuse, can also be ues to be difficulty in recruiting and retaining patients. At risk factors for the development of an eating disorder (15).
least a one-third dropout rate or withdrawal due to relapsecomplicate the interpretation of randomized controlledtrials (21). There is some indication that cognitive-behav- TREATMENT
ioral therapy following weight gain may reduce the risk of Both the severity of illness and the specific diagnosis will determine the treatment strategy for an eating disor- Family therapy is the most effective treatment for adoles- der. Guidelines have been developed for the treatment of cents with anorexia nervosa and seems to be equally effec- different degrees of severity of illness and are established tive when administered as conjoint or as separated family from hospitalization to day programs to intensive outpa- therapy (23). Overall, 6 months of therapy also seems to be tient therapy to group therapy (16). There are no random- as effective as 12 months; however, patients with severe ized controlled trials that adequately assess the intensity of obsessive-compulsive disorder may require longer treat- treatment. The major categories of eating disorders are ment (24). There is little formal study of the efficacy of anorexia nervosa, bulimia nervosa and binge eating disor- group psychotherapy for the treatment of anorexia nervosa der. Variations of these disorders are treated similarly to or the usefulness of support groups for this disorder.
the major diagnostic category which they approximate. Two core assumptions are made about anorexia nervosa There are three reviews of new treatment research in eat- in cognitive therapy. First, food avoidance, necessary to ing disorders with critical analyses. These include the maintain a low weight, is essentially a food phobia. Sec- Cochrane reviews (17), the Australian and New Zealand ond, anorexia nervosa serves a positive function: it pro- practice guidelines for the treatment of anorexia nervosa (18) vides an escape from aversive developmental issues and and the guidelines for treatment of eating disorders by the distressing life events often of an interpersonal nature. National Institute for Clinical Excellence in London (19).
One element in cognitive therapy is cognitive restructur- ing. In this approach the patient must identify specific nega-tive thoughts, list the evidence for these thoughts, list the Treatment of anorexia nervosa
evidence against the thoughts, form a reasoned conclusion For treatment of anorexia nervosa, the key elements are and use the reasoned conclusion to guide her behavior.
medical management, behavioral therapy, cognitive thera- Another element in cognitive therapy is problem-solving. In py and family therapy. Pharmacotherapy is at best an this procedure the patient identifies a specific problem, adjunct to the other therapies in this disorder. Nutritional develops different strategies, considers the likely effective- rehabilitation and weight restoration are essential. Behav- ness and feasibility of each strategy to deal with the problem, ioral therapy is useful in managing weight gain and pre- selects the best strategy, defines the steps to carry out the vention of binge eating and purging. Cognitive therapy strategy, carries out the chosen strategy and then evaluates addresses the distorted cognitions of feeling fat, evaluating the entire problem-solving process in light of the outcome.
self-worth solely by body image and the pervasive sense of Another essential element in cognitive therapy is monitor- ineffectiveness and inadequacy. Family therapy is especial- ing. For this the patients must make daily records of food ly effective for children under the age of 18. Fluoxetine intake, including the type of food ingested, the time of inges- may prevent relapse in patients who have obtained at least tion and the environment where the ingestion occurred.
85% of a normal weight. Atypical antipsychotics may be Monitoring also includes daily records of binge/purge useful in reducing severe anxiety and augmenting weight behavior, exercise, mood and interpersonal difficulties (25). Pharmacotherapy of anorexia nervosa is limited. Cypro- Nutritional rehabilitation programs usually employ heptadine facilitates weight gain in restrictive anorexia emotional nurturance and a variety of behavioral interven- nervosa and has an antidepressant effect. Chlorpromazine tions, which involve a combination of reinforcers that link or olanzapine may reduce severely obsessional, compul- exercise, bed rest, and privileges to target weight, desired sive and agitated behavior; a side effect is weight gain. Flu- behaviors and informational feedback. There are no ran- oxetine may reduce relapse of weight and eating disorder domized controlled trials to demonstrate the superiority of nasal gastric tube feeding over oral feeding in nutritional There are few new randomized controlled trials of phar- rehabilitation. Advocates claim that such intervention may macotherapy for anorexia nervosa. These trials have shown fluoxetine in a dose of 60 mg/day adds no benefit Cognitive-behavioral therapy is the first choice treat- to inpatient treatment of underweight anorexia nervosa ment for bulimia nervosa. It was the most effective treat- patients. Another trial of 35 anorexia nervosa patients that ment in 35 controlled studies, which showed 40 to 50% of were partially weight restored showed some indication patients abstinent from bingeing and purging at the end of that fluoxetine during weight maintenance may decrease treatment (16-20 weeks). Reduction in bingeing and purg- the relapse rate (26). A third trial, which compared three ing occurred in 70 to 95% of patients. Thirty percent with drugs, clomipramine, fluoxetine and amisulpride, in inpa- no improvement post-treatment showed improvement to tients showed that amisulpride had the best effect on full recovery one year after treatment (29). There is some evidence that treatment programs which From the recent promising pilot studies of pharma- include dietary counseling and managing are more effective cotherapy in anorexia nervosa we can conclude that than those that do not. The nutritional rehabilitation in antipsychotic medications such as olanzapine and queti- bulimia nervosa involves establishing patterns of regular, apine may be helpful during the weight restoration phase.
non-binge meals. Increasing caloric intake and expanding Citalopram may reduce depression and anxiety during macronutrient selection in meals is also important. This weight restoration. Fluoxetine is not beneficial in weight will likely correct any nutritional deficiencies that may be restoration but may decrease relapse rate in anorexia ner- vosa patients. Nutritional supplements with L-tryptophan Another psychotherapy shown to have some effect in do not increase effectiveness of fluoxetine (28). treating bulimia nervosa is interpersonal therapy. Focus There are many problems with these treatment studies psychodynamic psychotherapy was not as effective as cog- of anorexia nervosa. First, there are very few randomized nitive-behavioral therapy in short term trials. Behavior controlled trials. Second, patients are not motivated for therapy with exposure and response prevention had no treatment. This is evident in that patients do not enter tri- additive benefits over cognitive-behavioral therapy. A als and dropout rates are high. Third, medical complica- meta-analysis of 40 group psychotherapy treatment studies tions often require withdrawal from trials. Fourth, very suggested moderate efficacy. Groups that included dietary small sample sizes are present in completed trials. counseling were more effective, as were groups with morefrequent visits during treatment. Many clinicians favor acombination of individual and group therapy in the treat- Treatment of bulimia nervosa
The treatment of bulimia nervosa has several key ele- There are no randomized controlled trials of family ments. Cognitive-behavioral therapy, which can be con- therapy in the treatment of bulimia nervosa: it may be con- ducted as individual or in group format, has a psychoedu- sidered for adolescents. Self-help manuals and guided self- cational component and requires self-monitoring. Other help manuals use cognitive-behavioral techniques: there techniques are cognitive restructuring, problem-solving are limited trials with varying results; more development of and cost benefit analyses. Behavioral therapy in the treat- these self-help manuals and larger studies are needed.
ment of bulimia nervosa is usually in conjunction with cog- Dialectical behavioral therapy focuses on training in emo- nitive therapy. In this form of treatment, restricting expo- tional regulation skills: one study showed significant sures to cues is common as well as developing alternative improvement compared to a waiting list (31). behaviors with response prevention techniques to stop A summary of pharmacotherapy in bulimia nervosa vomiting. Interpersonal therapy focuses on interpersonal concludes that all antidepressants are better than placebo relationships and classifies the type of interpersonal prob- for reducing binge eating. Over a dozen double-blind lem. Pharmacotherapy has shown best results with se- placebo controlled trials of antidepressants have been lective serotonin reuptake inhibitor drugs, which reduce conducted with a dosage similar to the treatment of binger/purge behavior. These are preferred because of the depression. Medications improved mood and preoccupa- lower side effect profile. Tricyclic antidepressants also tion with shape and weight in about 20% of patients.
reduce binge/purge behavior but have greater side effects.
Complete abstinence from bingeing and purging, however, In contrast to anorexia nervosa, treatment studies of occurred in only 20 to 30%. Some medications that have bulimia nervosa have proliferated in the past 15 years.
been effective in reducing binge frequency should not be Controlled studies of specific therapy techniques such as used in treating bulimics because of their side effects.
behavioral therapy, cognitive therapy, psychodynamic Bupropion has been associated with convulsions in bulim- therapy and psychoeducation therapy have been conduct- ic patients. Trazodone has been associated with producing ed in both individual and group therapy format. Multiple delirium in a few bulimic patients. The monoamine oxi- controlled drug treatment studies have also been conduct- dase inhibitors can cause hypertensive crises if bulimic ed. Often a variety of therapy techniques are used together patients do not follow the required restricted diet (32). in either individual or group therapies. There is no way at Fluoxetine in a dose of 60 mg/day is the drug of first present to predict which bulimic patient will respond to choice because of beneficial effects and a favorable side effect profile. If the first trial is unsuccessful there is evi- dence that another antidepressant trial may be effective.
are associated with weight loss in the treatment of binge Minimal duration of successful treatment should be 6 eating disorder. Other drugs shown superior to placebo for months. Baseline laboratory assessments such as cell binge eating disorder are phenytoin and topiramate. There blood count, serum electrolytes, liver function tests, blood is one positive open study for zonisamide. Pharmacological urea nitrogen/creatinine ratio, thyroid function and elec- treatment should be considered as an option in all cases of binge eating disorder, not just those with concomitant Ondansetron, a 5HT3 antagonist, decreases afferent mood disorders. It should be considered strongly in those vagal activity and has been shown to have a mild effect on who fail to respond to psychological treatment. One should reducing binge/purge behaviors in bulimic patients. How- consider trials of topiramate, a selective serotonin reuptake ever this drug has serious side effects of constipation, inhibitor, sibutramine, venlafaxine, bupropion (no purging headaches and abdominal pain. Topiramate has been or history of bulimia nervosa or anorexia nervosa) and zoni- shown to effectively reduce binge/purge behavior in a samide. Be prepared if necessary to conduct a minimum of double-blind placebo-controlled trial in bulimia nervosa three trials to obtain an optimal response. Choose medica- patients: this medication must be initiated in a very low tion based on patient comorbidity and preference in side dose of 25 mg/day and gradually increased with a maxi- effect profile. Use doses similar to those for approved indi- mum of 400 mg/day. A slow increase of dosage will help cations. Treat for a duration similar to that for bulimia ner- prevent side effects of fatigue, paresthesia, difficulty con- vosa or major depressive disorder: for example, 6 to 12 centrating, and influenza-like symptoms. The opiate months at a level of substantial improvement before antagonist naltrexone, in a dose of 200-300 mg/day, has attempting discontinuation of the drug. In some cases treat- been shown to reduce binge/purge behavior. However, ment may need to be continued indefinitely (36). there is a concern of liver toxicity at this dosage (33).
Three randomized controlled studies comparing cogni- CONCLUSIONS
tive-behavioral therapy and pharmacotherapy showedthat the combination was superior to medication alone.
In conclusion, the treatment of eating disorders is based One study showed that the combination was superior to on a multimodal approach. Patients need to be treated cognitive-behavioral therapy alone, while the other two with a multidisciplinary team, including a psychiatrist for pharmacotherapy and psychotherapy, a nutritionist for In conclusion, the treatment of bulimia nervosa can be nutritional education and meal planning, an internist or summarized by stating that binge eating, purging, and core pediatrician for medical care and a family therapist for eating disorder attitudes respond best to cognitive-behav- ioral therapy. A greater improvement in mood and anxiety For the treatment of anorexia nervosa, the key elements occurred when antidepressant therapy was added. are medical management, behavioral therapy, cognitivetherapy and family therapy, while pharmacotherapy is atbest an adjunct to other therapies. In bulimia nervosa, the Treatment of binge eating disorder
treatment of choice is cognitive-behavioral therapy, but a Binge eating disorder is still considered in the category greater improvement in mood and anxiety occurs when of eating disorders not otherwise specified. This disorder is antidepressant therapy is added. In binge eating disorder, distinguished from bulimia nervosa by a lack of compen- cognitive-behavioral therapy and interpersonal therapy satory behaviors to counteract the caloric intake and produce substantial and long-lasting changes and pharma- weight gain from binge eating episodes. These patients do cological treatment has often a useful role.
not purge, exercise or engage in dieting. Randomized con-trolled treatment trials have used the same techniques asthose for bulimia nervosa. Patients with binge eating dis- References
order have responded well to cognitive behavioral therapy 1. Lilenfeld LR, Kaye WH, Greeno CG et al. A controlled family and antidepressants that have been effective in treating study of anorexia nervosa and bulimia nervosa: psychiatric disor- ders in first degree relatives and effects of proband comorbidity.
A summary of psychotherapy treatment research for binge eating disorder is as follows. Cognitive-behavioral 2. Crisp AH. The possible significance of some behavioral correlates of weight and carbohydrate intake. J Psychosom Res 1976;11:117-23.
therapy and interpersonal therapy produce substantial and 3. Grice DE, Halmi KA, Fichter M et al. Evidence for a susceptibili- long lasting changes in the specific and general psy- ty gene for anorexia nervosa on chromosome 1. Am J Hum Genet chopathology of binge eating disorder. Cessation of binge eating is associated with both weight loss and mainte- 4. Bulik CM, Devlin B, Vacanu S et al. Significant linkage on chro- nance of this loss over a 1 year period. mosome 10p in families with bulimia nervosa. Am J Hum Genet2003;72:200-7.
Double-blind placebo controlled trials of antidepres- 5. Kaye WH. Persistent alterations in behavior and serotonin activi- sants have shown that desipramine, fluvoxamine, fluoxe- ty after recovery from anorexia and bulimia nervosa. Ann NY tine, sertraline and citalopram all reduce binge eating and 6. Kaye WH, Frank GK, McConah AC. Altered dopamine activity 22. Pike KM, Walsh BT, Vitousek K et al. Cognitive behavior therapy after recovery from restricting-type anorexia nervosa. Neuropsy- in the post hospitalization treatment of anorexia nervosa. Am J 7. Kaye WH, Strober M. Neurobiology of eating disorders. In: Char- 23. Eisler I, Dare C, Hodes M et al. Family therapy for adolescent ney DE, Nestler EJ, Bunny BS (eds). Neurobiological foundation of anorexia nervosa: the result of a controlled comparison of two mental illness. New York: Oxford University Press, 1999:891-906.
family interventions. J Child Psychol Psychiatry 2000;41:727-36.
8. Monteleone P, Di Lieto A, Castaldo E et al. Leptin functioning in 24. Lock J, Agras WS, Bryson S et al. Comparison of short versus eating disorders. CNS Spectrums 2004;9:523-9.
long term family treatment for adolescent anorexia nervosa. Pre- 9. Jimerson DC, Wolfe BE. Neuropeptides in eating disorders. CNS sented at the Eating Disorder Research Society Annual Meeting, 10. Strober M. Personality and symptomatological features in young, 25. Kleifield EI, Wagner S, Halmi KA. Cognitive-behavioral treat- nonchronic anorexia nervosa patients. J Psychosom Res 1980; ment of anorexia nervosa. Psychiatr Clin North Am 1996;19:715- 11. Fassino S, Daga G, Amianto F et al. Temperament and character 26. Kaye WH, Nagata T, Weltzin TE. Double-blind placebo-con- profile of eating disorders: a controlled study with the Tempera- trolled administration of fluoxetine in restricting and restricting- ment and Character Inventory. Int J Eat Disord 2002;33:412-25.
purging type anorexia nervosa. Biol Psychiatry 2001;49:644-52.
12. Wade TD, Bulik CM, Neale M et al. Anorexia nervosa and major 27. Ruggiero GM, Laini V, Mauri MC et al. A single-blind comparison depression: shared genetic and environmental risk factors. Am J of amisulpride, fluoxetine and clomipramine in the treatment of restricting anorectics. Prog Psychopharmacol Biol Psychiatry 13. DiNicola FA. Anorexia multiforme: self starvation in historical and cultural context. Part 1: self starvation as a historical 28. Barbarich N, McConaha C, Halmi KA et al. Use of nutritional chameleon. Transcult Psychiatry Res Rev 1990;27:165-96.
supplements to increase the efficacy of fluoxetine in the treatment 14. Pate JD, Pumariea AG, Hester C et al. Cross-cultural patterns in of anorexia nervosa. Int J Eat Disord 2004;35:10-5.
eating disorders: a review. J Am Acad Child Adolesc Psychiatry 29. Wilson GT, Fairburn CG, Agras WS et al. Cognitive-behavioral therapy for bulimia nervosa: time course and mechanisms of 15. Gard MC, Freeman CP. The dismantling of a myth: a review of change. J Consult Clin Psychol 2002;70:267-74.
eating disorders and social economic status. Int J Eat Disord 30. Wilson GT, Loeb KL, Walsh BT et al. Psychological versus phar- macological treatment of bulimia nervosa: predictors and proces- 16. American Psychiatric Association. Practice guidelines for the sors of change. J Consult Clin Psychol 1999;67:451-9.
treatment of patients with eating disorders, 2nd ed. Washington: 31. Telch CF, Agras WS, Linehan M. Dialectical behavior therapy for American Psychiatric Association, 2000.
binge eating disorder. J Consult Clin Psychol 2001;69:1061-5.
17. Hay PJ, Bacaltchuk J, Stefano S. Psychotherapy for bulimia ner- 32. DeZwaan M, Roerig J. Pharmacological treatment of eating disor- vosa and binging. The Cochrane Database of Systematic Reviews.
ders: a review. In: Maj M, Halmi K, Lopez-Ibor JJ et al (eds). Eat- ing disorders. Chichester: Wiley, 2003:223-86.
18. Royal Australian and New Zealand College of Psychiatrists. Aus- 33. McElroy SL, Arnold LM, Shapira N et al. Topiramate in the treat- tralian and New Zealand clinical practice guidelines for the treat- ment of binge eating disorder associated with obesity: a random- ment of anorexia nervosa. Aust N Zeal J Psychiatry 2004;38:659-70.
ized, placebo-controlled trial. Am J Psychiatry 2003;160:255-61.
19. National Institute of Clinical Excellence. Eating disorders. Clini- 34. Walsh BT, Wilson GT, Loeb KL et al. Medication and psy- chotherapy in the treatment of bulimia nervosa. Am J Psychiatry 20. Halmi KA. Eating disorders: anorexia nervosa, bulimia nervosa and obesity. In: Hales RE, Yudolfsky S (eds). Textbook of clinical 35. Agras WS, Telch CF, Arnow B et al. One year follow-up of cogni- psychiatry. Washington: American Psychiatric Publishing, 2003: tive-behavior therapy for obese individuals with binge eating dis- order. J Consult Clin Psychol 1997;65:343-7.
21. Halmi KA, Agras WS, Crow S et al. Predictors of treatment 36. Fichter MM, Uqadfileg N, Gnutzmann A. Binge eating disorder: acceptance and completion in anorexia nervosa: implication for treatment outcome over a 6-year course. J Psychosom Res 1998; future study designs. Arch Gen Psychiatry (in press).

Source: http://www.essstoerungen-aargau.ch/media/archive1/fachpersonen/therapie/angebote/halmi_multimodaltherapy.pdf