Resistance to Treatment in Eating Disorders
Resistance to Treatment in Eating Disorders
We included in this overview 71 studies with a variable sample composition ranging from 14 to 748 participants. No reviews or meta-analysis on this topic were available. Three of the considered studies recruited clinicians instead of patients to investigate their countertransference. As regards the study design, we found debate articles, randomized controlled trial (RCT), qualitative interview, test validations, longitudinal cohort, cross-sectional, prospective cohort, and retrospective cohort studies (see Table 1).
As regards ED diagnosis, according to Diagnostic and Statistical Manual criteria (DSM-IV-TR), 32 studies included only AN individuals, in one case recovered-AN; 9 papers considered bulimia nervosa (BN) and AN individuals, in one case compared to healthy controls (HC); 14 studies AN, BN and eating disorder not otherwise specified (EDNOS) participants. A study compared AN patients with subthreshold-AN individuals. In another work have been included patients with other psychiatric diagnoses and HC. The remaining studies were debate articles or did not consider affected individuals.
Denial versus Insight of Illness (7 Studies) Konstantakopoulos and Coworkers highlighted how the restricting AN subtype is strongly characterized by lack of insight of illness and how the latter correlates with cognitive flexibility as measured with the trail making test. A couple of studies included in this area underscored that poor insight can assume delusional features, defining then a specific subgroup of AN patients even more treatment-resistant.
Moreover, two studies highlighted how insight of illness can be a helpful element in overcoming scarce compliance with therapy. Schoen and Coworkers in a recent paper demonstrated a correlation between insight of illness and the seeking of professional treatments whilst other authors supported indeed that duration of illness correlates with greater insight rather than with a more severe disorder.
Other three studies toned down the role of insight in overcoming poor compliance to treatments: it could be not only the lack of insight to maintain the disorder but also the combination of mature and immature defense mechanisms. In fact, the majority of treatment-resistant patients show a clear denial of illness rather than scarce insight. Moreover, it has been recently shown by the same authors – not in line with previous studies – that insight is not related to duration of illness and that poor insight can be maintained also in a chronic phase of the ED. Finally, Couturier and Lock demonstrated how denial of illness does not impact significantly the outcome of family therapy.
Motivation to Change (33 Studies) A few studies evaluated psychological determinants of motivation to change and only one paper investigated the correlation between motivation to change and quality of life as perceived by patients but - given its cross-sectional design - it is not clear whether motivation can be influenced by quality of life or vice versa.
As regards cognitive factors, only one paper found neurocognitive traits, in particular decision making impairment, as possibly involved in determining a less favorable outcome after Cognitive Behavioral Therapy (CBT), as well as lower weight gain and poorer motivation to change. AN patients would be cognitively oriented to choose short-term rather than long-term rewards with these features entailing poorer compliance to treatments and a less positive outcome. Recently, Danner and Coworkers highlighted a significant correlation between set-shifting and central coherence raising the hypothesis of a relevant link – as regards prognosis – between these traits and treatment outcome. However, the authors underscored also that such deficits are shared by both ill and recovered-AN individuals so they cannot be predictors of motivation to change.
Eleven studies included in this thematic area investigated quality and content of motivation. Three studies conducted by Nordbø and Coworkers demonstrated that a treatment independent willingness to recover is a fundamental requirement to readiness to change and another work showed that patients' willingness to change, as expressed at the beginning of treatment, is a relevant prognostic factor at six-month follow-up. In line with these findings, an improved motivation to change during therapy represents a crucial factor in overcoming relapses. All these studies highlight how patients' attitudes towards illness should be accounted for while defining motivation to change and investigated at the beginning of treatment.
It is noteworthy that those motivations verbally expressed by patients often do not correspond to an authentic intention to modify their eating disordered behaviors since ED patients can be strongly ambivalent about changing. The ambivalence issue has been confirmed also by another study suggesting the use of the Pros and Cons of Eating Disorders Scale as useful tool to evaluate patients' perspective of illness. Moreover, a longer duration of illness – index of poor motivation to change – is a negative prognostic factor mostly in AN.
Another paper investigated preoccupation with weight and body and found it to be determinant as regards CBT outcome and the chance to overcome treatment resistance. The intensity of such a preoccupation can assume a delusional connotation with repercussions on resistance to treatment by lowering motivation and generating strong ambivalence.
Twelve studies highlighted the correlation between patients' clinical features and their motivation to change. In fact, BN patients are usually more motivated to seek treatment and change than both AN and subthreshold-AN individuals, mostly if chronic, and there is general consensus that binge-purging AN individuals show an unfavorable outcome. Another recent study suggested indeed a more positive prognosis for EDNOS individuals; they seemed to achieve a more rapid and stable remission and showed indexes of higher motivation when compared to individuals affected by a full diagnosis, providing further support to data already known in literature. Those patients with normal body mass index (BMI), showed a more rapid improvement in motivation to change than those with a low BMI and, more in general, baseline BMI was the most significant predictor of outcome in the whole ED diagnostic group. Moreover, poor motivation to change correlated also with laxatives abuse, depression, and body dissatisfaction, although some researchers could not find a correlation between clinical severity and poor motivation to treatment, even if more recent studies did not confirm this hypothesis. However, two studies indicated that the rapidity of weight restoration is the only significant prognostic factor over the short and medium-term with Lund and Coworkers highlighting indeed how this can indirectly point out an enhanced motivation to treatment. Finally, Schedenbach and Colleagues underscored that the best predictors of treatment outcome are the ability to choose a variety of foods, mostly with high caloric density.
Eleven studies highlighted how motivation to change can quantitatively vary and several papers on AN described different stages of change and their influence on both outcome and resistance. Some authors demonstrated that the extent of clinical improvements can vary also depending on the stage of motivation achieved by patients and that a mismatch between stage of motivation and phase of treatment can enhance resistance to treatment. Other studies correlated motivation-to-change levels to the need and duration of hospitalizations, finding that high motivation correlates with short duration of inpatient treatment and better outcome. Conversely, if the motivation level is low – i.e. pre-contemplation phase according to Prochaska's model – the need for hospitalizations resulted to be higher.
Motivation to change can be improved by sharing treatment plans with patients and can be assessed with the Motivational Interview, the Readiness and Motivation Interview, or the Anorexia Nervosa Stages of Change Questionnaire. Some studies showed that the latter is a useful instrument to predict changes in eating symptomatology and outcome since motivation plays a mediator role between them.
Maintaining Factors and Treatment Outcome (22 Studies) Although early studies discouraged the search for ED-specific maintaining factors highlighting instead the need of long-term treatments, some recent papers called into question the need for their identification and reformulation.
Fairburn designed a transdiagnostic cognitive-behavioral therapy for EDs, aiming at addressing maintaining factors; he individuated as main element a scheme of dysfunctional self-evaluation by which patients attribute exaggerated relevance to eating, body shape, and weight. He considered both ED-specific factors (i.e. thoughts about eating, weight, body shape, hyperactivity) and non ED-specific factors like low self-esteem, interpersonal problems, emotional intolerance, and perfectionism. The latter interact with both individual's specific psychology and other maintaining factors.
Perfectionism has been considered also by other authors, demonstrating that some of its aspects could represent a transitional status associated to pathology and are no longer present in recovered AN individuals.
In addition, more authors individuated in body image an outcome predictor of hospitalization even more reliable than interpersonal problems and general psychopathology pointing out indeed how body perception instead of body dissatisfaction can be an indicator of treatment progression.
Other psychopathological factors with prognostic value were: inadequacy, high asceticism and maturity fear, impulsivity, and sexual problems. As regards intra-psychic elements instead, one study individuated as ED maintaining factors poor problem solving and relational skills.
Five studies considered personality traits as ED maintaining factors. Fassino and Coworkers pointed out how low novelty-seeking and high harm-avoidance – along with other psychopathological aspects – represent predictors of poor outcome in an ED multimodal treatment.
Other studies demonstrated instead that narcissistic personality traits were related to strong resistance to weight gain in treated AN individuals and that depressive and psychotic traits entailed better or poorer prognosis, respectively. Another paper with a 9 and 14-year follow-up found that Axis I and II psychopathology could predict both poor outcome and numerous hospitalizations in the ED population. A couple of studies confirmed the negative role of general psychopathology on AN prognosis at the 12-year follow-up, whilst another paper found it significant mostly about BN.
Further authors underscored indeed how avoidant personality traits, coupled with a history of sexual abuse, can play a negative role on long-term prognosis after hospitalization. Assuming a categorical approach to study personality, Helverskov and Coworkers acknowledged the presence of a personality disorder as a negative prognostic factor shared by all EDs.
Castro and Colleagues investigated the prognostic meaning of parental bonding on the outcome of short-term therapy in AN. Even highlighting how parental bonding was not particularly different from healthy controls, the authors underscored that parental hyper control as well as having a rejecting father entail are both elements that strongly impact treatments, in addition to ED psychopathology. Recently, some authors supported the role of caregivers' expressed emotions and reinforcing behaviors as interpersonal maintaining factors. In their work it has been indicated that reducing caregivers' distress leads to improving patients' functioning and eating pathology.
Schmidt and Treasure considered as maintaining factors both intrapersonal and interpersonal factors, placing only scarce emphasis on biological elements and body weight. Also Treasure and Coworkers analyzed the interpersonal maintaining factors of EDs pointing out that over protection, coercive treatments, and isolation could be iatrogenic factors.
Finally, two papers investigated psychosocial factors as the lack of a partner, poor family support, and unemployment as relevant predictors of poor outcome at 21 and 12-year follow-up. Another study showed how the scarcity of friends is a negative prognostic factor in EDNOS patients.
Therapeutic Relationship, Countertransference and Management of Treatment Resistance (18 Studies) The studies considered in this thematic area – deepening poor compliance and scarce motivation to treatments – indirectly suggested different models to treat resistant patients.
Carter and Colleagues highlighted how improving and maintaining motivation to treatment during therapy can show a relevant impact on the long-run. Accordingly, it has been shown the need to specifically address motivation as much as social relationships and body image with tailored interventions to obtain an adequate weight restoration, even with acute AN patients.
A couple of studies illustrated rehabilitation – with a focus on psychosocial interventions or supportive therapy - constantly advocating the need for tailored-to-person treatments.
Five studies suggested interventions focused more directly on overcoming poor compliance to treatments with an approach aiming at improving motivation to change and treatment. Already Vitousek and Colleagues years ago underscored the importance of improving motivation to change. With their paper, they suggested some cognitive-behavioral strategies that can be applied also to other theoretical models to enhance emotively and cognitively the therapeutic alliance to overcome resistance. In fact, the authors highly recommend to emotionally validate patients by accepting their difficulties and by speaking their language, adopting a Socratic style in the exploration of both ambivalence and resistance to treatment.
Another effective intervention was the Motivational Interview, an approach based on the Socratic method, emphasizing patients' autonomy and discouraging direct persuasion. Geller and Coworkers demonstrated the effectiveness of Readiness and Motivation Therapy to lower ambivalence and improve change; also Motivational Enhanced Therapy was found to be effective to achieve this goal, even if in a less structured way.
A theoretical paper highlighted that clinical interventions should sometimes indulge patients' resistances; the discharge of poorly motivated patients can be necessary to maintain a therapeutic milieu focused on recovery and to avoid poor compliance.
Finally, other studies suggested interventions to overcome scarce compliance not directly focusing on patients but rather on therapists, highlighting the need to handle correctly their countertransference since it could play a negative role on treatment. Countertransference can be determined by both patients' and therapists' features and supervision of therapist's emotions is highly recommended.
Patients highly value psychotherapy and therapeutic relationship considering them as useful elements in treatment. Accordingly, it has been suggested to work with a particular focus on the patient-therapist interaction and on shared choices. Concluding, a comment article proposed indeed to address the impairment in mentalization skills of ED patients within the therapeutic relationship.
Results
Sample Composition
We included in this overview 71 studies with a variable sample composition ranging from 14 to 748 participants. No reviews or meta-analysis on this topic were available. Three of the considered studies recruited clinicians instead of patients to investigate their countertransference. As regards the study design, we found debate articles, randomized controlled trial (RCT), qualitative interview, test validations, longitudinal cohort, cross-sectional, prospective cohort, and retrospective cohort studies (see Table 1).
Diagnosis
As regards ED diagnosis, according to Diagnostic and Statistical Manual criteria (DSM-IV-TR), 32 studies included only AN individuals, in one case recovered-AN; 9 papers considered bulimia nervosa (BN) and AN individuals, in one case compared to healthy controls (HC); 14 studies AN, BN and eating disorder not otherwise specified (EDNOS) participants. A study compared AN patients with subthreshold-AN individuals. In another work have been included patients with other psychiatric diagnoses and HC. The remaining studies were debate articles or did not consider affected individuals.
Thematic Areas
Denial versus Insight of Illness (7 Studies) Konstantakopoulos and Coworkers highlighted how the restricting AN subtype is strongly characterized by lack of insight of illness and how the latter correlates with cognitive flexibility as measured with the trail making test. A couple of studies included in this area underscored that poor insight can assume delusional features, defining then a specific subgroup of AN patients even more treatment-resistant.
Moreover, two studies highlighted how insight of illness can be a helpful element in overcoming scarce compliance with therapy. Schoen and Coworkers in a recent paper demonstrated a correlation between insight of illness and the seeking of professional treatments whilst other authors supported indeed that duration of illness correlates with greater insight rather than with a more severe disorder.
Other three studies toned down the role of insight in overcoming poor compliance to treatments: it could be not only the lack of insight to maintain the disorder but also the combination of mature and immature defense mechanisms. In fact, the majority of treatment-resistant patients show a clear denial of illness rather than scarce insight. Moreover, it has been recently shown by the same authors – not in line with previous studies – that insight is not related to duration of illness and that poor insight can be maintained also in a chronic phase of the ED. Finally, Couturier and Lock demonstrated how denial of illness does not impact significantly the outcome of family therapy.
Motivation to Change (33 Studies) A few studies evaluated psychological determinants of motivation to change and only one paper investigated the correlation between motivation to change and quality of life as perceived by patients but - given its cross-sectional design - it is not clear whether motivation can be influenced by quality of life or vice versa.
As regards cognitive factors, only one paper found neurocognitive traits, in particular decision making impairment, as possibly involved in determining a less favorable outcome after Cognitive Behavioral Therapy (CBT), as well as lower weight gain and poorer motivation to change. AN patients would be cognitively oriented to choose short-term rather than long-term rewards with these features entailing poorer compliance to treatments and a less positive outcome. Recently, Danner and Coworkers highlighted a significant correlation between set-shifting and central coherence raising the hypothesis of a relevant link – as regards prognosis – between these traits and treatment outcome. However, the authors underscored also that such deficits are shared by both ill and recovered-AN individuals so they cannot be predictors of motivation to change.
Eleven studies included in this thematic area investigated quality and content of motivation. Three studies conducted by Nordbø and Coworkers demonstrated that a treatment independent willingness to recover is a fundamental requirement to readiness to change and another work showed that patients' willingness to change, as expressed at the beginning of treatment, is a relevant prognostic factor at six-month follow-up. In line with these findings, an improved motivation to change during therapy represents a crucial factor in overcoming relapses. All these studies highlight how patients' attitudes towards illness should be accounted for while defining motivation to change and investigated at the beginning of treatment.
It is noteworthy that those motivations verbally expressed by patients often do not correspond to an authentic intention to modify their eating disordered behaviors since ED patients can be strongly ambivalent about changing. The ambivalence issue has been confirmed also by another study suggesting the use of the Pros and Cons of Eating Disorders Scale as useful tool to evaluate patients' perspective of illness. Moreover, a longer duration of illness – index of poor motivation to change – is a negative prognostic factor mostly in AN.
Another paper investigated preoccupation with weight and body and found it to be determinant as regards CBT outcome and the chance to overcome treatment resistance. The intensity of such a preoccupation can assume a delusional connotation with repercussions on resistance to treatment by lowering motivation and generating strong ambivalence.
Twelve studies highlighted the correlation between patients' clinical features and their motivation to change. In fact, BN patients are usually more motivated to seek treatment and change than both AN and subthreshold-AN individuals, mostly if chronic, and there is general consensus that binge-purging AN individuals show an unfavorable outcome. Another recent study suggested indeed a more positive prognosis for EDNOS individuals; they seemed to achieve a more rapid and stable remission and showed indexes of higher motivation when compared to individuals affected by a full diagnosis, providing further support to data already known in literature. Those patients with normal body mass index (BMI), showed a more rapid improvement in motivation to change than those with a low BMI and, more in general, baseline BMI was the most significant predictor of outcome in the whole ED diagnostic group. Moreover, poor motivation to change correlated also with laxatives abuse, depression, and body dissatisfaction, although some researchers could not find a correlation between clinical severity and poor motivation to treatment, even if more recent studies did not confirm this hypothesis. However, two studies indicated that the rapidity of weight restoration is the only significant prognostic factor over the short and medium-term with Lund and Coworkers highlighting indeed how this can indirectly point out an enhanced motivation to treatment. Finally, Schedenbach and Colleagues underscored that the best predictors of treatment outcome are the ability to choose a variety of foods, mostly with high caloric density.
Eleven studies highlighted how motivation to change can quantitatively vary and several papers on AN described different stages of change and their influence on both outcome and resistance. Some authors demonstrated that the extent of clinical improvements can vary also depending on the stage of motivation achieved by patients and that a mismatch between stage of motivation and phase of treatment can enhance resistance to treatment. Other studies correlated motivation-to-change levels to the need and duration of hospitalizations, finding that high motivation correlates with short duration of inpatient treatment and better outcome. Conversely, if the motivation level is low – i.e. pre-contemplation phase according to Prochaska's model – the need for hospitalizations resulted to be higher.
Motivation to change can be improved by sharing treatment plans with patients and can be assessed with the Motivational Interview, the Readiness and Motivation Interview, or the Anorexia Nervosa Stages of Change Questionnaire. Some studies showed that the latter is a useful instrument to predict changes in eating symptomatology and outcome since motivation plays a mediator role between them.
Maintaining Factors and Treatment Outcome (22 Studies) Although early studies discouraged the search for ED-specific maintaining factors highlighting instead the need of long-term treatments, some recent papers called into question the need for their identification and reformulation.
Fairburn designed a transdiagnostic cognitive-behavioral therapy for EDs, aiming at addressing maintaining factors; he individuated as main element a scheme of dysfunctional self-evaluation by which patients attribute exaggerated relevance to eating, body shape, and weight. He considered both ED-specific factors (i.e. thoughts about eating, weight, body shape, hyperactivity) and non ED-specific factors like low self-esteem, interpersonal problems, emotional intolerance, and perfectionism. The latter interact with both individual's specific psychology and other maintaining factors.
Perfectionism has been considered also by other authors, demonstrating that some of its aspects could represent a transitional status associated to pathology and are no longer present in recovered AN individuals.
In addition, more authors individuated in body image an outcome predictor of hospitalization even more reliable than interpersonal problems and general psychopathology pointing out indeed how body perception instead of body dissatisfaction can be an indicator of treatment progression.
Other psychopathological factors with prognostic value were: inadequacy, high asceticism and maturity fear, impulsivity, and sexual problems. As regards intra-psychic elements instead, one study individuated as ED maintaining factors poor problem solving and relational skills.
Five studies considered personality traits as ED maintaining factors. Fassino and Coworkers pointed out how low novelty-seeking and high harm-avoidance – along with other psychopathological aspects – represent predictors of poor outcome in an ED multimodal treatment.
Other studies demonstrated instead that narcissistic personality traits were related to strong resistance to weight gain in treated AN individuals and that depressive and psychotic traits entailed better or poorer prognosis, respectively. Another paper with a 9 and 14-year follow-up found that Axis I and II psychopathology could predict both poor outcome and numerous hospitalizations in the ED population. A couple of studies confirmed the negative role of general psychopathology on AN prognosis at the 12-year follow-up, whilst another paper found it significant mostly about BN.
Further authors underscored indeed how avoidant personality traits, coupled with a history of sexual abuse, can play a negative role on long-term prognosis after hospitalization. Assuming a categorical approach to study personality, Helverskov and Coworkers acknowledged the presence of a personality disorder as a negative prognostic factor shared by all EDs.
Castro and Colleagues investigated the prognostic meaning of parental bonding on the outcome of short-term therapy in AN. Even highlighting how parental bonding was not particularly different from healthy controls, the authors underscored that parental hyper control as well as having a rejecting father entail are both elements that strongly impact treatments, in addition to ED psychopathology. Recently, some authors supported the role of caregivers' expressed emotions and reinforcing behaviors as interpersonal maintaining factors. In their work it has been indicated that reducing caregivers' distress leads to improving patients' functioning and eating pathology.
Schmidt and Treasure considered as maintaining factors both intrapersonal and interpersonal factors, placing only scarce emphasis on biological elements and body weight. Also Treasure and Coworkers analyzed the interpersonal maintaining factors of EDs pointing out that over protection, coercive treatments, and isolation could be iatrogenic factors.
Finally, two papers investigated psychosocial factors as the lack of a partner, poor family support, and unemployment as relevant predictors of poor outcome at 21 and 12-year follow-up. Another study showed how the scarcity of friends is a negative prognostic factor in EDNOS patients.
Therapeutic Relationship, Countertransference and Management of Treatment Resistance (18 Studies) The studies considered in this thematic area – deepening poor compliance and scarce motivation to treatments – indirectly suggested different models to treat resistant patients.
Carter and Colleagues highlighted how improving and maintaining motivation to treatment during therapy can show a relevant impact on the long-run. Accordingly, it has been shown the need to specifically address motivation as much as social relationships and body image with tailored interventions to obtain an adequate weight restoration, even with acute AN patients.
A couple of studies illustrated rehabilitation – with a focus on psychosocial interventions or supportive therapy - constantly advocating the need for tailored-to-person treatments.
Five studies suggested interventions focused more directly on overcoming poor compliance to treatments with an approach aiming at improving motivation to change and treatment. Already Vitousek and Colleagues years ago underscored the importance of improving motivation to change. With their paper, they suggested some cognitive-behavioral strategies that can be applied also to other theoretical models to enhance emotively and cognitively the therapeutic alliance to overcome resistance. In fact, the authors highly recommend to emotionally validate patients by accepting their difficulties and by speaking their language, adopting a Socratic style in the exploration of both ambivalence and resistance to treatment.
Another effective intervention was the Motivational Interview, an approach based on the Socratic method, emphasizing patients' autonomy and discouraging direct persuasion. Geller and Coworkers demonstrated the effectiveness of Readiness and Motivation Therapy to lower ambivalence and improve change; also Motivational Enhanced Therapy was found to be effective to achieve this goal, even if in a less structured way.
A theoretical paper highlighted that clinical interventions should sometimes indulge patients' resistances; the discharge of poorly motivated patients can be necessary to maintain a therapeutic milieu focused on recovery and to avoid poor compliance.
Finally, other studies suggested interventions to overcome scarce compliance not directly focusing on patients but rather on therapists, highlighting the need to handle correctly their countertransference since it could play a negative role on treatment. Countertransference can be determined by both patients' and therapists' features and supervision of therapist's emotions is highly recommended.
Patients highly value psychotherapy and therapeutic relationship considering them as useful elements in treatment. Accordingly, it has been suggested to work with a particular focus on the patient-therapist interaction and on shared choices. Concluding, a comment article proposed indeed to address the impairment in mentalization skills of ED patients within the therapeutic relationship.
Source...