Anorexia Nervosa in Adults: The Urgent Need for Novel Outpatient Treatments that Work

Abstract

Anorexia nervosa (AN) is a serious psychiatric disorder that often follows a protracted course, and continues to confound those who attempt treatment once the patient has reached adulthood. Several randomized clinical trials for adults with AN have tested well-known therapies such as cognitive behavior therapy, supportive psychotherapies, or focal psychodynamic therapy, all of which have delivered frustratingly few helpful treatment strategies. Perhaps a different path could be pursued where we do not aim to cure all patients with one or two of these well-trodden therapies, Instead, a more targeted alternative, testing several novel approaches, could collectively reach a larger cohort of patients suffering from AN, the most lethal of all psychiatric disorders.

Anorexia nervosa (AN) is a serious psychiatric disorder and often follows a protracted course (Wonderlich et al., 2012). As a result, it continues to present as a treatment conundrum, leading a recent review to be entitled “The enigmatic persistence of anorexia nervosa” (Walsh, 2013). This is especially true for adults with this debilitating disorder, as most randomized controlled trials (RCTs) of outpatient psychotherapy or pharmacotherapy for adults with AN have provided few treatment guidelines, and more often inconclusive findings (e.g., Pike et al., 2003; McIntosh et al., 2004; Walsh et al., 2006). This state of affairs has left investigators exasperated, highlighting the significant challenges around treatment acceptance and completion in AN, and even cautioning colleagues to reconsider embarking on such endeavors without an improved understanding of these challenges (e.g., Halmi et al., 2004. Adolescents with AN seem to be somewhat more responsive to outpatient therapy, especially when parents are involved (family-based treatment: FBT). Moreover, engaging the patient on his/her own has not been entirely fruitless either. For instance, adolescent focused therapy (AFT), as an alternative to FBT, has yielded moderately satisfactory outcomes (Le Grange et al., 2014; Lock et al., 2010).

Since the publication of Halmi and her colleagues’ paper, urging caution before embarking on further RCTs for adult patients with AN, a firmer footing around these challenges have been established. For example, two RCTs, comparing quite well established treatments for adult AN, i.e., cognitive behavior therapy (CBT), specialist supportive clinical management (SSCM), or focal psychodynamic therapy, have since been published (Touyz et al, 2013; Zipfel et al., 2013). It is reassuring that our field can succeed in such challenging endeavors. For instance, Touyz and his colleagues retained well over 85% of their cohort through 12-month follow-up, even though there were no differences in outcome between CBT and SSCM. Zipfel and his colleagues randomized 242 patients with AN, making this the largest RCT for eating disorders to date. Like Touyz, their study also stopped short of being able to ‘crown’ one treatment as statistically or clinical superior to the other in this 3-way comparison of CBT-E, focal psychodynamic therapy, and treatment as usual. While that in itself is probably not ‘bad’ news, it does ‘rob’ us of establishing any one psychotherapy as the first-line outpatient treatment for adults with AN. More important, though, is that these studies probably underscore that this disorder, especially when it has established a more protracted course as often is the case in adults, does not respond uniformly to at least one or two of these well tested psychotherapies (as well as pharmacotherapies, for that matter). So, it would seem that in the treatment of AN ‘one size does not fit all’!¹

This conundrum should not come as a surprise. AN, although initially viewed to be quite uniform or homogeneous in presentation, has become much more diverse in presentation in part as our understanding of the genetics, temperament, personality and neurobiology, to name but a few, of this disorder has improved. Therefore, targeting specific eating disorder behaviors, traits, or cognitive styles, that may be unique in some patients, but not necessarily in others, or utilizing resources not available to every patient, e.g., a significant other, might be a helpful ways to develop more targeted treatments for sub-sets of the AN clinical population. Personalized medicine, perhaps? So, where to go from here would seems an optimal question.

At least three novel therapies for adults with AN have now been developed and undergone some preliminary testing. In the first, Bulik and her colleagues have identified that a sizable minority of adult patients are in fact partnered, and instead of ‘allowing’ these patients to fend for themselves, their significant other is engaged in a treatment program called Uniting Couples (in the treatment of) Anorexia Nervosa or UCAN (Bulik et al., 2011). Therefore, UCAN is a couple-based treatment that addresses problems with sexual functioning, relationship challenges, communication difficulties, as well as the eating disorder. It integrates CBT for AN and cognitive-behavioral couples therapy, and is an adjunctive treatment (20 conjoint sessions) to standard management of adults with AN, i.e., individual therapy, nutritional counseling, and medical follow-up.

One may very well ask why this is a novel approach, after all, utilizing family members in the care of their loved ones is hardly unique in just about any branch of adult medicine, e.g., organ transplant, bariatric surgery, oncology, to name a few. However, AN is probably one of very few, if not the only, ego syntonic disorder, which does set it apart from depression, organ failure, cancer, or probably most other medical or psychiatric diagnoses. Nevertheless, as a field we often remain ‘stuck’ in the idea that a patient with AN should be ‘allowed’ to struggle with this illness on his/her own (I appreciate that frequently we have no alternative clinical avenue to pursue). Therefore, in the context of decades of insisting (sufferers and professionals alike) that the adult patient with AN should fend for him/herself, Bulik and her team’s approach to bring the patient’s partner on board is indeed novel. The adolescent literature has been quite encouraging to engage parents to support their adolescent with an eating disorder (Le Grange et al., 2007; 2015; Lock et al., 2010). It should therefore be rather feasible that the partners of adult sufferers can be equally helpful in treatment, albeit on somewhat different relationship premises. Still, the challenge remains how to convince those patients whose illnesses have such a firm grip on their lives to invite or allow support from a loved one, whether it is a partner/spouse, or in fact a parent.

The second novel approach originates from Tchanturia’s work in cognitive remediation therapy (CRT) (Tchanturia et al., 2007). This body of research suggests that many adults with severe and enduring AN, even after weight restoration, still present with quite specific neurocognitive inefficiencies. That is, these individuals show biases in favor of an overly detailed cognitive processing style and/or also struggle to set-shift in a timely and efficient way. These neurocognitive inefficiencies can therefore be seen as endophenotypes and to play a significant role in the maintenance of this disorder (Roberts et al., 2012). Lock and his colleagues adapted CRT as a relatively brief (eight sessions) treatment to serve as a precursor or preparation for a course of standard CBT for AN (Lock et al., 2014). CRT aims to improve neurocognitive function by practicing very specific cognitive skills, before embarking on a course of CBT.

What makes CRT unique or novel, at least in its application as described here, is the fact that the goal of this initial phase of the larger treatment package (CRT + CBT) does not focus on the eating disorder per se. In fact, CRT as proposed by Lock and his colleagues emphasizes task process rather than outcome. While the treatment provider is monitoring weight and medical stability, the therapist engages the patient in a series of cognitive exercises or tasks to challenge the deficits in being overly detailed focus or unable to set-shift efficiently. Because these inefficiencies are seen to reinforce and/or maintain anorexic-like behaviors and obsessions, they are also viewed to inhibit the patient’s capacity to engage and utilize a treatment such as CBT for AN. While the therapist and patient are engaged in these CRT activities, there is no attention to behaviors such as food choices or other eating disorder symptoms. The latter is more specifically targeted in the ensuing CBT.

Most clinicians and researchers can probably agree that most patients with AN are usually unwilling to change, and therefore quite reluctant to engage in treatment. Whereas UCAN relies to a large extent on the patient’s partner to assist in these treatment challenges, CRT acknowledges the need to establish a collaborative therapeutic relationship and facilitate a willingness to change. This is achieved, in part, because of the focus on neurocognitive tasks rather than eating disorder symptoms. That said, learning new cognitive skills through moving outside one’s comfort zone, which facilitates change (and then engaging in a course of CBT), can be especially helpful for patients with severe and enduring presentations of AN. Perhaps limiting though is the fact that this approach might only be useful for patients who are highly functioning and neurocognitively intact, as the authors point out (Darcy et al., this volume).

The third of these novel approaches under discussion, exposure and response prevention (ERP), borrows from behavior theory or behavior psychology. ERP is a well-trodden path in the treatment of many anxiety disorders, especially obsessive-compulsive disorders. The co-occurrence of anxiety disorders, whether as a precursor or consequence of AN, and the persistence of food specific anxieties even after weight restoration, have for a considerable period of time now received extensive research and clinical attention (Bulik et al., 1997; Pollice et al., 1997). It is this overlap in eating and anxiety disorders, or shared clinical phenomena, that is at the core of Steinglass and colleagues’ calling for the utilization of ERP for a subset of adult patients with AN (Steinglass et al., 2014).

Exposure and response prevention for anorexia nervosa, or AN-EXRP, targets specific eating or food related anxieties in adults with AN who are acutely weight restored. In other words, AN-EXRP is an adjunct to concurrent inpatient treatment (about 12 sessions). In this novel treatment, exposure sessions are initiated gradually in order to facilitate therapeutic alliance, support the patient in building self-efficacy, and cultivate the patient’s certainty in this treatment approach. Having to negotiate a hierarchy of foods that are highly anxiety provoking, and being prevented from engaging in avoidant rituals, are understandably uncomfortable or even unpleasant. Therapeutic alliance is therefore particularly challenging in AN-EXRP, where treatment is more confrontational as opposed to the more ‘fun’ CRT (as described by the therapists, c.f. Darcy, this volume), where the focus is not on eating disorder symptoms.

The three case studies presented here, drawn from early treatment development studies of novel psychotherapies for adults with AN (still missing novel pharmacological agents), allow us to view the RCT path that have been pursued until now with a fresh set of eyes. It also allows us to speculate what that path might look like going forward. UCAN incorporates the patient’s partner in treatment, while CRT targets cognitive style, and AN-EXRP focuses on food specific anxieties. Each of these treatments is unique precisely because each focuses its interventions mainly in one arena, whether it is relationships (UCAN), the way patients think (CRT), or the fact that certain foods make them highly anxious (AN-EXRP). Each treatment is also an adjunctive intervention; UCAN in the context of standard outpatient treatment for adult AN, CRT preceding CBT, and AN-EXPR in the context of inpatient refeeding. In other words, in each treatment, the therapist is tasked with identifying which patient best fits which treatment. These paths are relatively clear then; patients with partners stand to benefit from UCAN, whereas those presenting with difficulty set-shifting or overly focused on detail could be assigned to CRT, and patients with persistent food related anxieties despite acute weight restoration could benefit from AN-EXRP. Therefore, all three approaches are more patient specific or targeted interventions.

In theory, treatment providers may report utilizing some degree of treatment integration or personalization, but in practice it is likely that they offer the majority of their patients with prominent ED symptoms primarily CBT, or psychodynamic psychotherapy, or supportive psychotherapy, or one of an array of treatments. This treatment of choice is often based on personal preferences or ones professional training background. Looking at the outcomes from a dozen or so RCTs for adults with AN to date, such an approach does not appear to deliver the desired results. The far majority of these RCTs seem to underscore, in their inconclusive findings, at least one argument – any one therapy is probably not going to be the answer for all adult patients with AN. I would venture to speculate that is indeed also the message that our colleagues from UCAN, CRT and AN-EXPR have read in these past RCTs, and therefore led to the more focused or targeted approaches put forward in these three reports.

I started this Comment with the question “where do we go from here”, given that it has been such a challenge to garner from the published RCTs any one treatment that can be elevated to the first-line psychotherapy for adults with AN. Perhaps we have been barking up the proverbial wrong tree? We do not always agree on much when it comes to the treatment of this disorder that continues to confound us with its complexity. However, we might be at this uncomfortable point of disagreement precisely because we are actually in agreement that this disorder presents in many different ‘shapes’ and ‘sizes’ and requires many treatments! Perhaps our collective energies will be better spent in doing more of what the UCAN, CRT and ANEXPR’s teams have done. That is, these novel treatments should inspire us to look for additional behavioral aspects that are unique to subsets of this patient population and better understand the underpinnings of this disorder, and develop appropriate novel interventions.

Until now, our large scale efforts, whether those involved CBT, supportive psychotherapies, focal psychodynamic therapy, and many other psychotherapies and pharmacotherapies, have delivered frustratingly few helpful treatment strategies for adult patients with AN. It is therefore conceivable that this new path, where we do not aim to cure all patients with one or two well-trodden therapies, but instead accept the more targeted alternative could collectively reach a larger cohort of patients with this most lethal of all psychiatric disorders.