Abstract
Bulimia nervosa is a disorder of complex etiology that tends to occur in young women. These individuals binge eat and purge by vomiting or other means, and often have depression, anxiety, substance abuse and extremes of impulse control. It is thought that binge eating and purging behaviors are, at least in part, a means of coping with dysphoric mood states and interpersonal stress. Bulimic symptoms are not likely to abate without development of new coping skills and behaviors. In the past 25 years, considerable progress has been made in developing specific psychotherapies and medication for the treatment of bulimia nervosa. Despite this progress, many individuals have partial responses to therapy and may remain chronically ill. This complex illness often requires a multidisciplinary team of professionals for effective management and, despite significant advances in treatment, bulimia nervosa continues to present major challenges for providers of care.
Bulimia nervosa (BN) first received attention as a disorder in 1979 in a paper entitled ‘Bulimia nervosa: an ominous variant of anorexia nervosa’ [1]. Patients are generally within a normal weight range or higher, often rendering their illness undetectable by others. As defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition [2], BN is characterized by three major clinical features including binge eating (consumption of an unusually large amount of food accompanied by feeling a loss of control), regular engagement in inappropriate compensatory methods to influence shape and weight (including purging by vomiting, abuse of laxatives and diuretics, strict dieting and excessive exercise), and self-evaluation influenced largely by body shape and weight. The disorder typically develops in late adolescence or early adulthood. Approximately 90% of those seeking treatment are women [3,4]. The prevalence is estimated at approximately 1–3% [4], with an additional 2–5% of young women presenting with eating disorder syndromes that are subthreshhold for diagnosis of BN.
The etiology of eating disorders appears to involve a complex interaction of environmental, biological and genetic factors, with more recent research often focusing on the latter two areas. Clinical presentation of BN is frequently preceded by conditions such as anxiety disorders, poor self-concept, affective regulation problems and impulsivity [5]. These commonalities support the possibility that anxiety disorders and certain personality traits are vulnerability factors for developing BN [6]. Twin and other familial studies provide substantial evidence that BN shares some common genetic etiological factors with anorexia nervosa (AN) [7]. Genetic ties are further supported by the fact that both diagnostic groups tend to have some similar traits, such as perfectionism and obsessionality, and cross-over from one disorder to another during a patient's lifetime is not uncommon. However, those with BN often display a higher degree of impulsivity than those with the nonpurging type of AN [8], illustrated by greater levels of substance abuse, self-injurious behaviors and risky sexual activity (of note, the BN group also has a higher rate of childhood sexual abuse than control populations) [9]. Another difference between the disorders involves the patients' own perceptions of their illness. In the nonpurging type of AN, patients frequently do not experience their symptoms as foreign or disturbing, whereas individuals who binge eat and purge are more likely to express significant distress as the disorder progresses [10]. As such, it is well known that BN patients are often more amenable to intervention than their AN counterparts [10]. Nonetheless, intense shame and guilt about their behaviors can make those with BN reluctant to seek treatment, leading to a chronic and severe course of illness [11]. In our clinical experience, the sensing of an empathic provider is key to the BN patient's presenting for and engaging in treatment.
The need for effective treatments has generated much research interest in recent decades, resulting in a substantial body of literature. A range of controlled studies has shown efficacy for psychotherapy and pharmacotherapy in BN [12]. This paper will review the literature on treatment of BN, highlighting recent developments, and discuss the practical management of patients with BN.
Psychotherapeutic interventions for adults with bulimia nervosa
Accumulated research [13] suggests that the gold standard for psychotherapy is a structured and manual-based form of cognitive behavioral therapy (CBT) specifically developed for BN (CBT-BN) by Fairburn and colleagues [14]. It is focused on symptoms, maladaptive cognitions and the direct process of change rather than underlying psychological issues [15]. Used in an outpatient setting with either individuals or groups of patients, CBT-BN was designed as a three-phase approach delivered in 15–20 sessions over 4–5 months. The first phase aims to educate patients regarding BN and encourage normalized eating patterns. The second phase focuses on identifying cognitive distortions and utilizing behavioral strategies to reduce tendency toward dietary restraint. The third phase emphasizes maintenance and relapse prevention.
More than 50 randomized, controlled trials have tested the psychotherapeutic treatment of BN, with more than 20 of these on CBT [13]. Studies conducted with CBT have shown a decrease in the frequency of binge eating and purging episodes averaging 60–80%, with a cessation rate of 30–50% [16]. These outcome rates for CBT-BN are significantly higher than those seen in various comparison treatments (e.g., behavior therapy, nutritional counseling, focal psychotherapy, hyp-nobehavioral therapy, stress management training, psychodynamically oriented psychotherapy, supportive psychotherapy and various forms of exposure with response prevention) or waiting lists for treatment. In interpreting the results of these studies, it should be considered that dropout rates were as high as 30% [17], even though study patients probably had relatively higher levels of motivation than many BN patients who present for treatment in clinical practice [18]. In addition, study subjects had fewer psychiatric comorbidities than is usual for BN patients. When these factors are viewed together, it is reasonable to presume that studies of typical BN patients in the actual clinical setting have yet to be performed.
Early behavior change has been shown to predict the best treatment outcome and maintenance at follow-up [19,20]. It is important to note that even with full recovery from pathological eating behaviors, a substantial number of patients display residual features including excessive weight-related concerns, dysphoric effect, social discomfort and personality traits indicative of perfectionism [21,22].
For patients with poor initial response to CBT, interpersonal therapy (IPT) has been suggested as a second-level treatment [23]. IPT is also a manual-based treatment, involving 12–20 individual sessions over 3–5 months. IPT for eating disorders was devised based on compelling evidence that eating-disordered individuals typically have interpersonal factors contributing to the development and maintenance of their illness. Discussion of food and weight issues is minimized with the premise that improved interpersonal functioning will result in a reduction of eating disorder symptomatology. The patient gains insight as the therapist targets four specific types of interpersonal problems, namely, role disputes, role transitions, grief and interpersonal deficits [24]. Fairburn and colleagues compared CBT, IPT and behavior therapy and found that while IPT appeared to have been less effective overall than CBT when patients were assessed at the end of the treatment period, those treated with IPT and CBT had similar results at long-term follow-up (mean of 5.8 years) [14]. Similarly, a second multicenter study found that CBT produced results more rapidly and showed greater efficacy at the end of the 20-week treatment, while the effects of IPT were delayed until follow-up [25]. More recently, use of CBT and IPT as sequential treatments in BN was studied in a randomized trial. The trial had a high attrition rate, with patients more likely to drop out of IPT than to achieve symptom remission. The authors concluded that lengthy sequential treatments appear to have little value [26]. With a study duration of approximately 8 months, it is possible that some attrition may have been attributable to patients tiring of the rigors of a research trial. Nonetheless, their suggestion that alternative models for therapy need to be tested appears well founded.
Developments in the area of CBT-based self-help and guided self-help (GSH) manuals for BN have suggested that these programs may be effective alternatives for patients who otherwise do not have access to treatment [27]. While relatively few patients make a full and lasting recovery from GSH alone, it is recommended as a possible first step (when CBT-BN is not feasible) in the management of patients with less severe BN [4]. To date, most controlled studies of GSH have occurred at or in conjunction with eating disorder treatment centers. However, for many patients with eating disorders, lack of access to specialty care remains a barrier to treatment, and attempts to broaden dissemination of effective therapies is an emerging area of research. As patients with BN often initially present to primary-care providers, three recent randomized, controlled trials have examined whether GSH treatment may be delivered effectively by nonspecialists in primary-care settings.
In 2003, a study in the UK comparing CBT-based GSH provided by general practitioners (GPs) with outpatient care received at eating disorder specialty clinics concluded there was no significant difference in outcome between the two patient groups, and the findings supported the idea that BN can be treated in general practice [28]. This trial has been criticized on the grounds that the specialty care was far below standard, since the mean number of sessions attended by patients during the 6-month trial was less than five, and therefore drawing conclusions would be difficult given this questionable comparison condition [20]. Another point is of practical relevance. Following completion of the trial, only 11 of the 32 participating GPs said they would be willing to use the approach again, the main reported difficulty being lack of time –despite the fact that all but two of them were treating just one study patient. Given that therapy for BN may last for months or more, the absolute number of patients who could receive treatment exclusively in a primary-care setting appears quite limited.
The following year, Walsh and colleagues reported very different results in a trial conducted in two primary-care clinics in the Northeastern USA [29]. The comparison was among various treatment options that might be available in this setting (fluoxetine alone, placebo alone, GSH plus fluoxetine and GSH plus placebo). Fluoxetine was superior to placebo, though rates of remission were low (consistent with results of prior medication trials) and the authors concluded that GSH provided no apparent benefit. Most striking was the fact that two-thirds of the 91 patients did not complete the trial, compared with the 20–35% dropout rate usually seen when BN patients receive GSH interventions at specialty clinics. Of note, providers received only very brief training prior to participation, physician follow-up included 15-minute medication appointments, and GSH occurred in 30-minute sessions with nursing staff – a model, the authors believed, that more closely resembled true field conditions in the typically busy primary-care practice [20].
Most recently, in a trial with 109 patients, investigators in Australia compared GSH delivered by GPs with a waiting-list control group and reported very favorable results, stating they found decreases in binge/purge symptoms and overall remission rates of the treatment group comparable to those of standard CBT-BN delivered in specialty treatment settings. Perhaps the most important aspect of this study involved the nature of the providers and the extent of services delivered. The GPs were recruited by advertisement, and 13 of the 16 who agreed to participate ‘had a special interest in eating disorders in general practice’. They received more extensive training than in either of the above studies, apparently treated multiple patients over time, had access to at least occasional expert supervision and saw the patients on a relatively frequent basis [30].
The data above appear to support two significant findings. First, where primary-care physicians have the available time and interest to receive additional training in GSH techniques and closely follow individuals with BN, a subset of these patients might be manageable in primary care. Second, fluoxetine 60 mg/day appears to be a potential treatment option for BN patients seen in primary care, and while it will rarely produce remission, this may be a useful intervention for patients who are awaiting or unable to receive specialty care.
Another treatment modality to consider is therapy involving family members of the BN patient [31]. Despite the paucity of research in this area, it is recommended that family therapy be considered as an adjunct whenever feasible, especially for adolescents, college-age or other patients still living with parents, older patients with ongoing conflicted interactions with parents, and married/partnered patients experiencing relationship difficulties [27]. This may be particularly helpful for those from families that lack cohesion or have difficulty expressing emotion [32]. Clinicians also should be mindful of potential deleterious effects on children of patients with BN. Mothers with active eating disorders may be more prone to use food for non-nutritive purposes (e.g., as a rewarding stimulus) and to exhibit high levels of concern over their children's weight [33], increasing the risk of development of eating disorders in offspring.
Recent developments in pharmacotherapeutic interventions
Numerous controlled trials in the 1980s and early 1990s established antidepressants as more effective than placebo in the treatment of BN [34]. Efficacy has been shown across a variety of antidepressant medication classes, including tricyclics, monoamine oxidase inhibitors and, more recently, specific serotonin reuptake inhibitors (SSRIs). Success was most commonly measured by decrease in frequency of binge eating and purging episodes and, though studies varied widely, many achieved reductions of 50–75% in one or both areas [27]. However, despite significant progress in the field, it is important for the clinician to realize that most patients who responded to medication continued to binge eat and purge at a frequency that would meet criteria for a diagnosis of BN at the end of the studies. Only a minority of patients attained remission of binge/purge symptoms [35].
With their relatively benign side-effect profiles, SSRIs are considered first-line pharmacological treatment of BN [4,27]. The SSRI fluoxetine is the only medication that specifically has an indication for BN that was approved by the US FDA. The indication was granted in 1996, after large, multicenter, controlled trials convincingly demonstrated that fluoxetine was more effective than placebo in the treatment of BN. It was found that a relatively high dose of 60 mg/day was superior to 20 mg/day (which did not separate from placebo) [36], and that a starting dose of 60 mg/day was generally well tolerated in this patient population [37].
While it is presumed that any of the SSRIs would be useful in treatment of BN, trials with those other than fluoxetine have been scant. Nonetheless, study drug has proven superior to placebo, except in the case of fluvoxamine, where results have been mixed. While two small, controlled trials have shown efficacy [38,39], in a much larger randomized trial, involving 267 patients in the UK, the investigators reported that fluvoxamine performed no better than placebo for short- or long-term (1 year) treatment of outpatients with BN. It was noted that the dose range of 50–300 mg may have been too low to show treatment effect with this illness. However, adverse events were a problem for many patients, and titration to higher doses may not have been tolerated [40]. Overall, therefore, little evidence exists for use of fluvoxamine as an SSRI of choice in BN.
Recent results with sertraline are promising, although only one small randomized trial, with 20 patients, has been reported. Interestingly, significant improvement versus placebo occurred with only 100 mg/day, in contrast with the relatively high doses required for fluoxetine [41]. The reason for this finding is unknown, but replication on a larger scale would be beneficial, given that some patients in clinical practice do have significant side effects on 60 mg or more of fluoxetine.
Only one randomized, controlled trial has compared various SSRIs head-to-head. The 6-week study, with 91 patients, was designed to explore a genetic basis for drug response. No such basis was discovered, but the authors reported that fluoxetine, fluvoxamine, citalopram and paroxetine all performed superiorly to placebo. It is noteworthy that this study involved inpatients, a relatively unusual treatment environment for BN [42].
Bupropion is another antidepressant worth mention. In 1988 it was found to be effective in reducing BN symptoms in one small, controlled study; however, 4 out of 69 patients receiving the drug experienced seizures [43], resulting in contraindication by the FDA for use in eating disorders. Although no study has attempted to replicate these findings using the lower doses and longer-acting forms of the drug more commonly prescribed today, clinicians in the eating disorders field generally avoid use of all formulations of bupropion in BN patients.
Although antidepressants were initially proposed for treatment of BN owing to frequent accompanying mood symptoms, these medications have also shown efficacy in BN patients without comorbid anxiety and depressive disorders, and thus appear to independently target symptoms of binge eating, purging and preoccupation with shape and weight [44]. This result, along with a generally more rapid onset of action and, for fluoxetine, a higher dosage requirement than is seen with treatment of depression, suggests that drug action for BN symptoms may occur by a different mechanism, perhaps relating in part to underlying variance in the serotonergic abnormalities found in BN patients versus those with major depression [45]. Such possibilities have lead to case reports and small trials of novel medications, most notably topiramate and ondansetron.
The anticonvulsant topiramate has been linked to appetite suppression and weight loss [46]. Results from the first randomized, double-blind, placebo-controlled trials of topiramate in BN have been promising [47,48]. In 2003, a 10-week trial of 69 patients began with a dose of 25 mg/day and titrated to a maximum of 400 mg/day (median: 100 mg/day). Topiramate was more effective than placebo in decreasing frequency of binge/purge behaviors (to a similar degree as seen in antidepressant trials) and in improving other psychological measures. The authors stated that adverse events were generally mild-to-moderate, resolved with time or dose reduction, and did not usually lead to withdrawal from the study. Another 10-week controlled trial, conducted in Germany in 2004, had similar findings with regard to reductions in binge/purge behaviors and side effects, and also reported a decrease in weight for the topiramate group. There was an unusually low dropout rate among the 60 participants, possibly because the study enrolled only women with moderate cases of BN, or because the medication was titrated slowly and doses were relatively low (25–250 mg/day) [49]. Despite the report of few intolerable side effects in these small studies, adverse reactions to topiramate are common in clinical psychiatric practice and it should be considered for use in BN only when other medications have been ineffective [27]. Additional concerns include lack of data on longer-term safety and efficacy, as well as the physical and emotional complications likely to result if its use leads to weight loss in the normal- or low-weight BN patient. Larger, multicenter trials are needed.
Another novel agent proposed for treatment of BN is the anti-emetic ondansetron, a drug that acts on the serotonergic system. Ondansetron was found to be more effective than placebo in reducing binge eating and vomiting in a 4-week randomized trial of 25 patients with severe, chronic BN [50]. More information regarding potential side effects, likelihood of compliance with multiple daily dosing, cost effectiveness and impact of the drug on psychological features of BN is necessary to assess the clinical utility of ondansetron for BN.
Finally, with BN, as is the case with many other psychiatric disorders, few controlled trials have studied long-term efficacy of pharmacotherapy. Where such attempts have been made, interpretation of results has often been clouded by significant dropout rates. In the largest trial, the 150 responders (≥50% decrease in weekly vomiting episodes) to an 8-week course of fluoxetine 60 mg/day were randomized to continued drug treatment or placebo for a 1-year follow-up period [51]. Findings did include a lower rate of ‘relapse’ for the fluoxetine group, but the authors noted a worsening on all measures of efficacy over time. Considering that most of these patients were still binge eating and purging after the first 8 weeks, and therefore were in need not only of ‘maintenance’ but also further treatment of symptoms, pharmacotherapy alone may not be an adequate treatment for many patients.
Treatment combining psychotherapy & medication
At least six controlled trials have assessed direct comparisons of outcome for patients treated with psychotherapy, pharmacotherapy or a combination [27,52]. In general, results showed a greater decrease in the frequency of binge/purge episodes with CBT than with antidepressant medication when each was utilized alone. With treatments used in combination, the results to date have been mixed. Although several trials indicated that medication conferred no significant benefit beyond that achieved with psychotherapy, on balance, study results slightly favored the addition of medication to psychotherapy for many patients [53]. For example, one well-designed trial [54] found that adding an antidepressant to CBT conferred a modest advantage in symptom reduction [55]. The potential benefits of medication may be greatest for patients who fail to achieve an early reduction in binge eating and purging with evidenced-based psychotherapy, such as CBT or IPT, or for those who appear to have a depressive syndrome or anxiety disorder independent of BN [54,56]. Antidepressants may also be recommended when psychotherapy has been only partially successful. Interestingly, a recent meta-analysis reported that combining an antidepressant with psychotherapy significantly reduced the acceptability of psychological treatment, rendering antidepressants relatively more acceptable to the patient [57]. The clinical relevance of this finding is uncertain, but may indicate that a subset of BN patients prefer the perceived simplicity of taking medication to the more time- and effort-intensive nature of psychotherapy. Although more research is needed on combined treatment, there is a general consensus within the clinical community that an approach including both psychotherapy and medication is worth considering for many patients [27].
Treatment team & levels of care
BN is a complex illness, often affecting many aspects of life, and for the typical patient a collaboration of professionals from a variety of disciplines is necessary for successful care and treatment. American Psychiatric Association practice guidelines suggest the interdisciplinary team include a registered dietitian, primary-care physician, psychologist, psychiatrist and social worker [27].
Assessment of medical stability, as well as nature and severity of eating disorder symptoms and behavior, are the primary determinants used to recommend level of care. Other factors to consider include degree of social support, level of motivation for treatment and patient access to healthcare. Most patients with BN can be managed in an outpatient setting [58]. For some, this may consist of weekly individual and/or group psychotherapy targeting BN. Others will benefit from an intensive outpatient program, meeting for several hours 3 days/week. For the self-motivated patient these approaches may be preferred [59], as the individual may continue with work or school on at least a part-time basis, allowing immediate application of new coping skills and behaviors in a realistic setting. If outpatient treatment fails to ameliorate binge/purge behavior, a partial hospital program should be considered. These typically meet 5 days/week, with supervision during most meals. For the infrequent patient who needs 24-h observation to interrupt symptoms, inpatient/residential treatment may be necessary. Serious medical problems (e.g., electrolyte disturbances) requiring extended monitoring, poorly controlled co-occurring psychiatric disorders, significant suicide risk and substance use problems interfering with treatment are also indications for this highest level of care [27].
Evaluation for level of care begins with a thorough physical examination and review of systems, laboratory studies as appropriate and a general psychiatric assessment. The most serious acute medical complications of BN stem from aberrations in volume status and electrolyte levels that result from purging, periods of severe calorie and fluid restriction or excessive exercise. As such, key elements of the examination include vital signs, height and weight measurements, heart rate, rhythm and sounds, capillary refill and tests for muscular weakness. Blood chemistry, including electrolytes, complete white and red cell counts with differential, thyroid function tests and urinalysis are recommended at baseline for all patients. Electrocardiogram and other studies may be performed where indicated [27]. It should also be noted that patients who purge by vomiting require referral for dental examination and ongoing care, as tooth enamel erosion and caries are common. A detailed review of the assessment and management of medical complications of BN is beyond the scope of this paper and may be found elsewhere [4,60].
Individuals who are not in the care of a psychiatrist generally should be referred for assessment, as high prevalence rates have been reported among BN patients for major depression, as well as anxiety disorders such as obsessive compulsive disorder [61]. Some authors also describe increased incidence of post-traumatic stress disorder resulting from childhood trauma or abuse. In addition, substance use problems appear to affect up to 40% of patients with eating disorders (with the highest percentages seen in BN) [62]. Drugs of choice include alcohol, diet pills and illicit stimulants [59]. Self-injury may also be a problem for patients with BN. Comorbid diagnosis of BN and borderline personality disorder appears to increase the likelihood of self-injurious behaviors, which may in turn increase rates of attempted and completed suicide [63].
Practical issues: how to understand & deal with bulimia nervosa
While reviewing treatment options for BN is important, it is also relevant to provide some understanding of the unique symptoms of this disorder and the frustrations that patients, their families, and providers encounter when seeking and engaging in treatment.
Early in the course of illness, symptoms of BN may be highly reinforcing for the patient. Many have difficulties with communication and relationships, and binge eating can become a reliable source of comfort, providing a means of coping with dysphoric mood states and interpersonal stress [18,55]. But extreme fear of weight gain often drives compensatory purging and an existence dominated by relentless calorie counting and obsessive thoughts about food, weight and shape. The endless binge/purge cycle fosters shame, secrecy and social isolation, and by the time a patient seeks treatment the disorder may have taken over her life. Despite this frequently intense level of suffering, BN patients nearly always present with some degree of ambivalence for treatment [64], an observation that is baffling to many providers. It is helpful to recognize that symptoms may serve as a coping mechanism, albeit a dysfunctional one, and are unlikely to abate without development of new skills and behaviors to take their place [65].
For BN patients who initially present to primary-care providers, responding to their concerns with compassion and understanding is key to forging a therapeutic alliance. With those unwilling or unable to pursue a specialty treatment program, discussing the elements of healthy eating and exercise can be a place to start. For instance, eating in a healthy way includes all food groups and avoids extreme notions, such as attempting to remove all fat from the diet. Discussing exercise regimens in detail is also important, as some BN patients may ‘purge’ with workouts lasting several hours in length. For patients prone to denial and resistance, it is helpful to avoid power struggles. This can be accomplished by asking the patient to identify her own goals and then giving suggestions about how they might be achieved.
An important consideration in treatment relates to BN patients' overwhelming fear of ‘fatness’, which leads many to strive for an unnaturally low weight. Acceptance of a healthy body weight is crucial to recovery, because attempts to lose weight will usually trigger BN symptoms [66]. A classic pattern in BN involves restricting calories for long periods over the course of a day in an attempt to control or lose weight, followed by strong cravings to binge eat and subsequently purge, only to lead to further restricting. The link between physiological and emotional deprivation and binge eating has been well established [67] and explaining the factors leading to overwhelming urges to binge can reassure the patient of a rational basis for her symptoms [66]. Improvement of nutritional status impacts both physical and psychological wellbeing, so a return to eating meals and snacks spaced regularly throughout the day is a vital step in recovery. In fact, self-monitoring performed by keeping detailed written records of dietary intake, with timing of meals and snacks as well as episodes of binge eating and purging along with associated thoughts and feelings, is considered a cornerstone of CBT for BN [68]. Referral to a registered dietitian who specializes in eating disorders can facilitate this process.
For patients ready to pursue more intensive treatment, referral to a qualified eating disorders specialist can help with further assessment and determination of appropriate level of care. The National Eating Disorders Association is an excellent resource of information [101]. In addition, the Eating Disorder Referral and Information Center website lists therapists and programs throughout the USA [102]. Insured individuals can gather information regarding providers within their network.
Future perspective
In light of the fact that BN was defined as a diagnostic entity only a few decades ago, researchers and clinicians alike have made great strides in developing and implementing treatments helpful to many patients. Nonetheless, our current established treatments, both psychological and pharmacological, lead to a remission rate of perhaps 50% at best over the long-term and a substantial number of patients remain chronically ill [34]. To improve success rates with currently available evidenced-based psychotherapies, future studies should examine the reasons for substantial dropout rates from therapy, as well as conditions surrounding subsequent relapse, including impact of features of the disorder that may persist despite remission of binge/purge behaviors by the end of treatment [18]. Strategies for improving amount and quality of method-specific training received by treatment providers may be another factor that is important to effective delivery of CBT and IPT [69]. In addition, efforts are underway to test enhancements of the CBT-BN curriculum itself, including greater emphasis on body image disturbances, low self-esteem, perfectionism and interpersonal difficulties [20]. Other areas to watch include the emerging adaptation to eating disorders of psychotherapeutic methods found to be useful in the treatment of patients with addictive disorders and with borderline personality disorder – readiness for change and motivational enhancement therapy in the former case, and mindfulness and dialectical behavior therapy in the latter [70,71]. The American Psychiatric Association has called for well-conducted studies of such alternative or integrated psychotherapeutic treatment approaches for nonresponders, as well as more trials including complex BN patients with comorbid psychiatric conditions who are often seen in clinical practice [27].
The number and size of pharmacotherapeutic trials for treatment of the disorder has slowed in recent years, which is unfortunate, as few notable developments in medication management have occurred in the decade since the FDA approved fluoxetine for use in adult patients with BN. As is the case with psychotherapy, there is some evidence that early response to medication predicts positive outcome [72]. Therefore, studies investigating antidepressant switching and augmentation strategies, which are nearly nonexistent for treatment of BN [18], are sorely needed. Exploration of medication classes other than antidepressants forms the core of more recent research efforts, but controlled trials have been small and relatively few. This may in part relate to the limited state of knowledge on physiological processes underlying binge/purge behavior. Elucidation of the biological aspects of eating disorders (including those impacting body image, self-esteem, obsessionality and perfectionism) holds the promise of new drug targets, such as mechanisms of hunger, satiety and taste perception [27].
Executive summary
Bulimia nervosa (BN) is characterized by binge eating, inappropriate compensatory behaviors such as vomiting, laxative abuse or excessive exercise, and self evaluation influenced largely by body shape and weight. Patients are generally within a normal weight range or higher.
Approximately 90% of those seeking treatment are women.
The gold standard treatment in psychotherapy for BN is a specific, manual-based form of cognitive behavioral therapy (CBT-BN) that focuses on symptoms, maladaptive cognitions and the direct process of change rather than underlying psychological issues.
Although most patients who complete CBT-BN have some improvement in symptoms, less than 50% achieve remission of binge/purge behavior and a significant number of those later relapse. Many patients continue to be troubled by issues related to food, weight and shape.
In actual practice, many therapists incorporate elements of CBT into an eclectic approach.
Interpersonal therapy (IPT) and guided self-help (GSH) are alternative manual-based treatments for BN. GSH potentially may allow some patients access to therapy in primary-care settings, but the time and effort required of nonspecialist providers may make effective treatment difficult.
Specific serotonin reuptake inhibitors (SSRIs) are the first-line pharmacological treatment for BN. Fluoxetine has the only US FDA-approved indication and a relatively high dose (60 mg/day) is often required to reduce binge/purge symptoms. Action in BN appears to be independent of antidepressant effects. Few trials have studied other SSRIs, but most are believed to be effective.
Other classes of antidepressants have also shown efficacy, but bupropion is generally not recommended for use in BN, as patients in one trial in the late 1980s experienced seizures at a higher-than-expected rate.
The anticonvulsant topiramate has been linked to appetite suppression and weight loss and may be an option for treating binge/purge symptoms when other medications have been ineffective and when potential weight loss would not be detrimental to the patient.
Trial results have shown a greater decrease in binge/purge frequency with CBT than with antidepressant medication when each was utilized alone, but further studies and general consensus in the field support a moderate additional benefit for at least some patients when both treatments are combined.
With both psychotherapy and medication, there is growing evidence that early behavior change predicts outcome in BN so treatment of nonresponders should be switched or augmented aggressively.
In most cases a multidisciplinary team approach is necessary for successful care and treatment.
Evaluation for level of care begins with a thorough physical examination, review of systems and laboratory studies as appropriate. The most serious acute medical complications stem from aberrations in cardiovascular and electrolyte status resulting from BN behaviors.
All patients should undergo general psychiatric assessment, as BN is often comorbid with mood disorders and substance use problems that may complicate treatment.
Most BN patients can be managed on an outpatient basis, but significant psychosocial issues, low motivation for recovery or serious medical/psychiatric problems may necessitate higher levels of care.
BN behaviors often serve as a coping mechanism for dysphoric mood states and interpersonal stress, nearly always leading to at least some patient ambivalence for treatment. Symptoms are not likely to abate without development of new coping skills and behaviors.
The binge/purge cycle tends to escalate and foster shame, secrecy and social isolation.
BN patients often present to primary-care providers, and compassion and understanding is key to forging a therapeutic alliance. Elements of healthy eating and exercise should be discussed in an empathic manner. Asking the patient to identify her own goals and then giving suggestions on how they may be achieved can help avoid power struggles.
Knowing and explaining the biological underpinnings of binge cravings can reassure the patient of a rational basis for her symptoms.
Acceptance of a healthy body weight and return to a regular pattern of eating are vital for physical and psychological well-being, as attempts to lose weight will usually trigger BN symptoms.
Referral to an eating disorders specialist is often necessary. The National Eating Disorders Association is an excellent resource of information [101]. In addition, the Eating Disorder Referral and Information Center lists therapists and treatment programs throughout the USA [102].
Executive summary
Despite great strides in development of therapeutic offerings for BN, most patients remain symptomatic or relapse following treatment.
To improve effectiveness of evidenced-based psychotherapies, future studies must evaluate reasons for patient dropout and relapse. Efforts are underway to enhance the CBT curriculum, and studies of alternative and integrated therapies are also needed.
Few notable developments in medication management of BN have occurred in the decade since FDA approval of fluoxetine. Studies investigating medication switching and augmentation strategies are nearly nonexistent in BN and are sorely needed. Elucidating the biological aspects of eating disorders holds the promise of new drug targets for novel medications.