But why is there a dearth of studies on resistance? This apparent disinterest seems to be due to several reasons. Firstly, it is likely that many authors conceive resistance to treatment as a whole with the illness. According to this perspective, the investigation of treatment outcomes in AN would correspond to evaluating resistance too [
99]. Secondly, resistance to treatment is a widespread phenomenon. Hence, some of the papers we included in this overview [
14,
54,
100] pointed out how treatment-resistance in the AN field can be easy to notice but difficult to understand. In fact, it is a complex phenomenon that involves vulnerability and maintaining aspects with intertwined biological, psychopathological, and social features [
101] that cannot be easily disentangled. Thirdly, in the last years, research focused more on resistance-related biologic aspects [
102] rather than on clinical and relational ones [
98]. Also psychodynamic psychotherapies have been even less studied although the assessment of resistance represents the core of the psychoanalytic method [
103] since resistance can be enhanced by the relationship between patient and therapist. Finally, we could call into question clinicians' poor motivation to study their frustrated therapeutic attempts [
94]. Whatever the reason, it may be a mistake not to carefully consider this phenomenon since it is a hallmark of psychiatry [
104]. More in detail, since psychotherapy is an effective therapeutic instrument in the ED field, a better understanding of both resistance and strategies to address it should be carefully considered in treatments [
2]. From the research questions underpinning this article we sorted our results into four core areas with regard to AN treatment.
Denial versus insight of illness
This is the most immediate and straightforward correlate of resistance to treatments in AN and it was included even in DSM-IV-TR diagnostic criteria [
91]. Denial of illness is defined as the refusal to acknowledge and accept one’s own illness and it refers not only to psychodynamic therapy and defense mechanisms [
2] but also to a wider definition of disadaptive coping [
105]. Denial of illness is an intrinsic factor of the first phases of AN [
91] and it can last years [
86], given the ego-syntonic nature of the anorectic disorder [
106]. Accordingly, the Academy of EDs has clearly stated that EDs are severe mental illnesses that require a wide and multifaceted health care alert as like as other major psychiatric pathologies including bipolar disorder, schizophrenia, obsessive-compulsive disorder or depression [
3].
We included 7 studies converging on the fact that a large number of AN patients deny their disorder [
25‐
31]. This phenomenon can be particularly clear for those patients who need acute hospitalizations: a recent study conducted on a sample of 108 AN inpatients showed that the vast majority of them (63%) deny their illness [
107] with 20-30% of cases revealing a symptom-related psychotic status [
28,
30]. These features may be so widespread because many patients tend to carry out a deliberate denial [
27] or minimization [
25] of their illness trying to justify – with different degrees of awareness - treatment refusal. In this regard, the data in literature are debated and do not clarify whether denial of illness is a psychosis-like symptom or rather a rigid and disadaptive defense mechanism helping patients to protect themselves by anxiety and depression [
26,
31] and to avoid treatments [
89] or negative emotions [
22,
23]. It is likely that both these aspects are true; nevertheless, hypothesizing denial as a defense mechanism could be more in line with its time consistency [
86], independently of fluctuations in ED symptomatology [
28,
29].
Data are also controversial about the extent to what denial can impact prognosis since this element seems to be scarcely relevant as regards adolescents [
25]. One paper highlighted a linear correlation between insight and duration of illness [
31], although this finding has not been replicated [
28]. At now, it is not possible to distinguish whether denial is already expressed at the onset of illness or it is enhanced with time and potentially by inappropriate treatments. However, it should be considered as a central issue to be addressed in treatments rather than a prognostic element [
29].
The improvement of insight of illness could be an index of a good therapeutic alliance [
108]. In fact, the therapeutic relationship is now indeed considered as the most effective instrument to contrast denial of illness [
75,
109].
In sum, denial of illness, a shared factor with other severe mental disorders [
110], was not found to be a predictor of resistance to treatment but rather to correlate with the phenomenon of treatment resistance.
Motivation to change
Papers on motivation to change in AN allowed a better understanding of denial of illness although one main difficulty encountered in this field of research is to understand patients’ authentic degree of motivation to recover since there is often a difference between what patients do and say [
100].
Studies on motivation to change (33 included in this clinical overview) suggested ways to measure this construct [
17,
20,
28,
30,
32,
34‐
52,
55‐
57,
60],[
61] and interventions to enhance it [
33,
58‐
60]. A lower motivation to change was found to correlate with lower BMI [
45] - and in general a more severe eating and general psychopathology [
35,
39,
54]-full diagnosis, purging behaviors [
33,
45,
46,
56], lower compliance to dietary recommendations and slow weight gain [
48,
50,
57], and worse quality of life [
51]. These data are in line with Kaye’s hypothesis of a vicious cycle of symptoms as maintaining factors in EDs [
102].
From the available body of literature emerged a correlation between motivation and psychopathology, even more peculiar in AN than BN [
37].
In spite of the number of conducted researches, to date it is still controversial whether motivational interventions can be a main road to improve resistance to treatment or not. There is a significant correlation between willingness to recover and good motivation to change [
36], but this association is variable and scarcely supported by other studies [
32,
34,
49,
60,
61]. Moreover, willingness to recover [
52,
100,
111] or the ability to recognize the negative effects of illness [
37] can be confused with motivation to change and this misinterpretation could lead not experienced therapists to enhance resistance through its underestimation. In fact, some studies highlight the need to train therapists to notice, understand and value those emotions that usually underpin AN [
97,
108,
112].
The role of neuropsychological aspects – promising and growing research area [
113] – is even less studied and the data in literature are contrasting [
38,
40]: in fact, cognitive rigidity and impaired decision making – stable traits also in recovered individuals - can only partially represent a hindrance to treatments.
Although several lines of evidence exist in support of motivational interventions [
58‐
61], a recent review conducted by Waller [
100] questions their effectiveness since the available studies are frequently biased by methodological flaws. In fact, motivational interventions are not stand-alone treatments; therefore, the psychotherapies (i.e. CBT) that they usually support could be responsible for the real effectiveness of these interventions. Hence, motivational interventions – as to date have been described in literature – can only scarcely impact motivation and failed to significantly improve outcomes with the only exception represented by Binge Eating Disorder patients [
100]. Other findings did not provide support for the effectiveness of such interventions with long-standing patients [
83].
Some hypotheses could be raised to bridge the gap as regards motivation to change: a) too much emphasis has been placed on words instead of facts [
100] without considering patients’ peculiar strive to please [
97] and need for approval [
108]; b) motivation has been conceived as a too linear concept whilst patients cannot switch directly from one stage to another; c) the assessment of motivational stages is not adequately considered during treatment planning: often motivational stage and phase of therapy do not match, having resistance to treatments enhanced as a result [
51,
58]; d) motivational models may be too simple and may not to consider patients’ ambivalence in an proper manner [
70,
114,
115]; e) an excessive use of verbal persuasion is usually made at the expense of patient autonomy [
100,
116].
Therefore, only a few studies tried to investigate the process of recovery in AN and to conceptualize more in detail ambivalence to change and its implications [
17].
Ambivalence and meaning of illness
In this sense, some papers [
17,
20,
43‐
45,
70] highlighted the relevance of both motivation and pervading ambivalence that paralyzing patients while making their decisions. Some authors pointed out the existence of a sort of “anorexic voice”, an inner entity disapproving patients and being sometimes overwhelming in respect to their sense of the self [
117,
118]. This voice is even more pervading when the illness gets worse, contributing to hamper treatments. It has been suggested that the link between patients and this AN voice could play a role in unravelling the issue of ambivalent attitudes towards change, typical of BN individuals too [
70]. Thus AN and BN patients would constantly struggle between facing resistance to treatment or valuing it.
From a therapeutic perspective, it could be useful to become familiar with the adaptive function of AN and its pros and cons, as experienced by the patients [
70,
115]. Therefore, it becomes possible not only to perform a cognitive restructuring, but also to mitigate their distress [
24] with two aims: a) to use the empathic approach to understand patients’ inner world [
89] and to dialogue on both diagnosis [
119] and recovery meanings [
17]; b) to move the focus of the intervention from pathological beliefs to therapeutic relationship [
120].
One way to start addressing resistance to treatment in ED patients is trying to understand the subjective meanings of the illness [
75] and patients’ environment [
121]. We found 5 articles that consider in detail patients’ meanings of the disorder and they all agree on considering the “positive” functions of AN as enhancing treatment resistance [
20]. In fact, the ED can be a way to feel safe, avoid threatening emotions, communicate with others, and feel strong, special and in control [
21].
Considering patients’ perspectives for treatments could have several positive implications. First, patients feel themselves as empathically understood [
83,
108,
122,
123]; second, it becomes possible to deepen and personalize the comprehension of those feelings that underpin the ED [
97] since they can be very different between patients [
20]. Third, it is possible to introduce mind and cognitions in therapy – in addition to eating behaviors and body distortions - to avoid resistance and relapses [
124]: in fact, if decades ago therapies tended to be excessively focused on family and intra-psychical aspects, now we could incur an opposite risk. Weight restoration should be the starting point of a treatment instead of its main goal. The eating pathology should be considered more as a disturbance of corporeality and as an impairment of embodiment [
125] rather than a neurological body image distortion [
126]. Fourth, talking about what patients think and feel about the meaning of their illness could provide a therapeutic framework, enabling them to experience their autonomy [
120] and supporting an empathic relationship. In fact, it is not possible to face the illness without “being with” the patient [
16].
However, the attention to the meanings that patients confer to their disorder is a necessary but not sufficient condition to understand the phenomenon of resistance to treatment. In fact, the therapist risks to become emotionally indulgent with iatrogenic and resistance-enhancing results [
94]. It is crucial for the therapist to achieve a “firm empathy” [
127] because an empathic understanding of the patient is not enough; firm boundaries are of vital importance to counterbalance empathy in the therapeutic relationship. Such elements are not only cognitive but have also a relational meaning and function. Therapists need to address patients’ need for boundaries, even if not verbally expressed. The treatment, with its implicit relational instruments [
128], enables patients to perceive their therapists as both holding [
97,
129] and handling [
89,
129] their self-harming attempts. This could be the result of the struggle for control to achieve a sense of identity that Bruch pointed out decades ago highlighting also that for many AN individuals “the experience of being listened to appeared to be of utmost importance … instead of having their feelings and the meaning of their communication interpreted” [
97]. This balance between the two elements – firmness and empathy - can promote changes in patient’s personality and coping through interiorization. Certain pathological behaviors should not be allowed or clearly prohibited. At the same time, it is useful to understand patients’ resistances, objections and even their need to feel alone, refused, and poorly understood.
Therapists may also incur the risk to consider only patients’ conscious meanings of illness and to underestimate those unconscious, deeper, and even more distressing. In this sense, conscious meaning can over the long-term hide useful elements in treatment and therefore contributing to treatment-resistance [
75].
In sum, the clinical effort performed on the meanings of illness – either conscious or not – could help overcome resistance to treatment. The understanding of patients’ inner world and attitudes towards the illness can become an operational tool to address the core of treatment-resistance within the therapeutic relationship.
Maintaining factors and treatment outcome
The complex available models to address resistance to treatment are mainly focused on AN maintaining factors and with this overview we found 22 articles investigating this research area showing interesting and well-organized models [
11,
18,
39,
56,
62‐
79]. The shared core is the attention to both “symptom treatment” eventually entailing a vicious cycle [
5,
18,
66‐
68] and negative effects of starvation on brain [
130]. Also body image distortions should be addressed in detail [
63,
65] and body dissatisfaction has been found to correlate with certain styles of attachment [
108]. There is robust evidence showing that mental health cannot be reached without recovery of weight, body perception, obsessive thinking on food and body and without regaining a good quality of life.
Moreover, the different models consider personality and interpersonal aspects as premorbid or maintaining factors. Several factors have been called into question as enhancing treatment resistance: 1) low self-esteem and 2) mood intolerance [
65]; 3) perfectionism [
65,
78]; 4) body experiences [
17,
125]; 5) general psychopathology [
39,
56,
64]; 6) personality [
66‐
68,
72,
73,
76,
79]; 7) interpersonal relationships [
65,
75]; 8) cognitive inflexibility and 9) avoidance of experience and emotions [
29,
131,
132]; 10) care givers’ expressed emotion [
14,
62,
71,
75]; 11) poor problem solving abilities [
69];12) scarce social support [
64,
74]; and 13) reduced relational abilities [
39,
69,
133] (for a review on widely used treatment models see: Hay and colleagues [
134]). Evidence on a relationship between dropout, treatment response and both character and temperament [
14,
67,
135] encourages to focus on considering in detail patients’ personality.
The extension of research to areas including not only eating symptomatology seems timely and promising: clinicians will probably obtain more instruments to understand their patients, individualize treatments, and handle resistance. However, to date there are no findings supporting such a more articulate approach to AN as regards treatment response. This model can be indeed too detailed for those patients with a less severe psychopathology [
136] and RCT are not currently available [
109]. Rather- although further studies are needed - an RCT [
137] did not show any significant effect in increasing treatment response.
For those patients with an enduring AN and consolidated maintaining factors, supportive or rehabilitative therapies have been suggested [
58,
87,
138,
139] since it is unlikely that certain severe patients will respond to treatments being aware of their resistance. Although these approaches are interesting and potentially useful, to the best of our knowledge there is still no clear consensus in literature on the criteria used to define the chronic course in the ED field [
87], making even more problematic to group those patients who would benefit from such supportive treatments.
Therapeutic relationship, countertransference, and management of treatment resistance
Although further research is needed, the aforementioned studies allowed a better understanding of the AN pathogenesis although they were not effective enough in improving prognosis. Sometimes we cannot see woods for trees and - focusing too much on specific aspects - we could miss the overall emotive exchanges constituting the therapeutic relationship [
24,
97] and its complex patterns of interaction [
53].
We retrieved 18 studies investigating the role of the therapeutic relationship on resistance to treatments in AN. It is of interest that these works on one hand refer to decades ago [
89], whilst on the other have been only recently conducted [
47]. These elements show well the gap in literature that now some authors are trying to fill although both Bruch [
97] and Garner [
140] underscored the issue of the iatrogenic effect of those therapists who are not able to manage their own emotions. All these papers agree on the importance of emotive aspects within the therapeutic relationship, in treating treatment-resistant AN people. Patients themselves – when describing their recovery process – individuate psychotherapy and relationships as fundamental tools to overcome resistance [
88]. In particular, psychotherapy has been described as a continuous and significant experience to achieve self-validation [
69]. Such opinions match some authors’ suggestions highlighting the relevance of sharing treatment plans with patients [
41].
In this regard, resistance to treatments should be considered within the therapist-patient interaction [
84] involving on one hand patients’ and illness features [
116] and on the other therapist-related factors [
98] and their interaction [
90]. In particular, the avoidant [
66‐
68,
79,
141] and narcissistic [
76] personality traits of AN patients, in addition to their disadaptive management of anger [
142,
143], make the therapeutic alliance difficult, sometimes enhancing the illness and patient’s relational isolation.
In fact, EDs – assuming an overarching psychodynamic perspective – are essentially disorders of the development of the self and personality, as Bruch [
97] originally conceived and as Skårderud [
24] and Stanghellini and Coworkers [
125] have recently suggested, placing emphasis also on insecure attachment [
108,
144] and mentalization impairments [
120,
145].
In particular, AN patients’ deep emotions are characterized by fear, emptiness, anger, and profound demoralization [
116,
142,
143,
146]. The illness is a desperate and self-harming attempt to control distress, on one hand avoiding emotions and on the other hand expressing them in an exaggerated way or developing an exasperated perfectionism [
95,
147]. In general, emotion avoidance and dysregulation are mostly related to treatment-resistance [
95].
Such emotions reverberate in line with therapists’ ones, mainly if they are young [
98] or lacking supervision [
81]. Particularly frustration and anger, but also despair, excessive worry, boredom, and feeling of being manipulated [
81,
85,
98] are common when investigating therapists’ countertransference features. Resistance to treatments itself is thought to be the most challenging aspect of the AN treatment, according to studies conducted on clinicians’ perspectives [
98]. In this regard, treatment-resistance could be linked to a contagious fear of aggressiveness and despair that could involve the therapist too.
Treatments could be influenced by an overemphasis on cognitive [
24] or explicit communication factors, whilst the issue could be related to the avoidance of an emotive confrontation on symptomatology and real-life experiences or, more simply, to the lack of an authentic relationship between patient and therapist [
16]. An effective therapy to overcome resistance to treatments could be indeed a cognitive-analytic therapy as Dare and Coworkers [
148] suggested and tested with RCT.
Psychotherapy – as well as those therapies focused on AN symptomatology – should then help patients to achieve a multi-dimensional understanding of themselves and to manage their feelings and relationships, gradually reshaping the adaptive function of the illness. This psychotherapeutic model was only sporadically tested in literature [
68,
135] and - although it may look outdated - it could be proposed again in the light of the studies on psychotherapy currently available.
In fact, psychodynamic psychotherapy is currently regaining its role in psychiatry [
149], mostly as integrative discipline and science of intimacy useful to achieve a developmental psychopathology and overcoming its traditional concept of science of interpretation [
128]. Neurosciences significantly highlighted how relational our mind is [
150] and that human beings are wired to be social [
151]. Initial findings showed that shared emotions can synchronize brain activity [
152]: from a meta-analytic studies of dynamic psychotherapies we now know that they are effective in several mental disorders [
153] and that the more they consider affective and emotional aspects the more effective they are [
154].
In psychodynamic psychotherapies, two aspects have been considered as key-elements: a) a secure, sensitive, and interactive therapeutic alliance; and b) encouraging patients to experience the previously avoided threatening feelings [
155]. It is the time to (re)introduce these therapeutic processes in the study of AN treatments. Moreover, the concept of resistance to treatments arose from psychoanalysis, as recently remarked [
156], on the basis of Freud’s statements [
157] clarifying that resistance is intertwined with treatment and that it represents a compromise between the strengths related to recovery and those opposing to it.
AN patients ask their clinicians acceptance, intensity, challenge and mostly competence [
53], confronting their knowledge, but even more their relational skills. Being able to provide an empathic understanding is fundamental to train the patients to recognize themselves, and restrain their distress with a good balance of implicit and explicit messages in the here and now of the therapeutic relationship through transference and countertransference [
158].
Psychotherapy cannot be manualized enough to avoid the unpredictability of the relationship [
128]. To improve the quality of the therapeutic relationship the therapist needs to be authentic, implicit and empathic [
128,
159]. If the attunement of the therapeutic relationship turns so profound and intense to become embodied simulation [
160], also through mirror neurons [
161], therapist and patient can start to share not only the distress but also the ability to limit it. Studies on personality and EDs confirm how often emotional coping can be impaired [
162]. Therapist’s coping skills can be a useful model to enhance the development of patient’s coping. In fact, affected individuals can implicitly feel and consider the change, starting to overcome their resistance and fear, as like as therapists handle their fears of being too frustrating or too sympathetic with their patients [
11].
In the therapeutic relationship, AN patients can experience new theories of others’ mind [
163] and more adaptive forms of reflective self-functioning [
164] and through this integration of psychic realities they will be again more aware of their own body [
24].
Unfortunately, to date there is still little in the way of addressing the treatment-resistance issue in AN. The studies on the importance of therapeutic relationship in facilitating emotive experiences are still sparse or provide only pilot data [
143,
165]. Although an RCT has been designed in this regard [
166], further studies are still warranted to bridge this gap.