European Eating Disorders Review

Eur. Eat. Disorders Rev. 16, 329–333 (2008)

Viewpoint
Many Roads Lead to Rome:
Why Does Cognitive Behavioural
Therapy Remain Unsuccessful for
Many Eating Disorder Patients?
Johan Vanderlinden 1,2*
1
University Psychiatric Center K.U. Leuven, Campus Kortenberg, Belgium
2
Faculty of Psychology, Department of Psychodiagnostics and
Psychopathology, Catholic University of Leuven, Belgium

This paper makes the case that the widely used evidence-based
’CBT’ approaches remain unsuccessful in many eating disorder
patients. Six critical ’personal reflections’ are formulated on why
many patients remain totally resistant toward our therapeutic
endeavours. My reflections suggest that probably many roads
may lead to Rome. Copyright # 2008 John Wiley & Sons, Ltd
and Eating Disorders Association.

Keywords: eating disorders, cognitive behavioural therapy, cognitive schemata, outcome of treatment

INTRODUCTION Within the CBT model (see Fairburn, Cooper, &

Shafran, 2003) disturbing or unrealistic preoccupa-
A core symptom in all eating disorder patients is the tions with regard to the body and weight are
presence of unrealistic cognitions about one’s body considered as important eliciting and maintaining
and weight. This applies to the DSM-IV criteria factors in the development of eating disorders.
(APA, 1994) for eating disorders, mentioning for According to the cognitive behavioural theory
anorexia nervosa (AN) a ‘disturbance in the way in central to the maintenance of eating disorders is a
which one’s body weight or shape is experienced’ dysfunctional scheme of self-evaluation. People
and for bulimia nervosa (BN) ‘that self-evaluation is with eating disorders judge themselves largely or
unduly influenced by body shape and weight’. Over exclusively in terms of shape, weight and eating
the past 25 years, significant progress has been control (and often all three). Most of the other
reached in the development and evaluation of clinical features of eating disorders derive from this
evidence-based psychological treatments for eating ‘core psychopathology’ (Fairburn et al., 2003). For
disorders. The treatment of choice for almost all both the treatment of AN (Garner and Bemis, 1982)
eating disorders is now called ‘cognitive beha- and BN (Wilson, 1997, Fairburn et al., 2003), a CBT
vioural therapy’ (CBT) and is widely recommended model has been described that focuses very directly
by most evidence-based treatment guidelines (for on the disturbing and unrealistic thoughts with
instance NICE guidelines, 2004). regard to food, body and weight. Changing
unrealistic cognitions has become also an essential
part in the treatment of the binge eating disorder
* Correspondence to: Dr Johan Vanderlinden, University Psy-
chiatric Center K.U. Leuven, Campus Kortenberg, Leuvense
(BED) (Fairburn, Marcus, & Wilson, 1993).
Steenweg 517, B-3070 Kortenberg, Belgium. Hence, CBT therapists consider the critical ques-
E-mail: [email protected] tioning of these unrealistic cognitions and then the

Copyright # 2008 John Wiley & Sons, Ltd and Eating Disorders Association.
Published online 30 July 2008 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/erv.889
330 J. Vanderlinden

replacement of these unrealistic cognitions by more wrong or that the patient is in the phase of total
realistic and adequate beliefs, as one of the most denial of the problem. So my suggestion no. 1:
important tasks in the therapy. At the same time, the Always evaluate motivation of the patient and plan
patient is invited to practice all kinds of behavioural your therapeutic strategy accordingly.
experiments in order to challenge their unrealistic
beliefs.
However, important problems remain. Not all our Critical Reflection No. 2: Too Much Focus on
patients get better with a CBT approach. Not all our the Content of the Cognition
patients are successful in adapting and installing
The third generation of CBT approaches such as
new and more realistic beliefs about themselves and
‘mindfulness’ (Kabat Zinn, 1990) and ‘acceptance
their bodies. In the longer term CBT seems to be
and commitment therapy’ (ACT; Hayes, Strosahl, &
successful in only 45–50% of bulimic patients and
Wilson, 1999) criticise the old CBT approaches
few or no data are available in the case of AN
because they have been focusing too much on the
patients (Gowers et al., 2007). Hence, research data
content of the cognition and not on the over
show that only half of our patients recover from
involvement (too much rumination) with the
their eating disorder problems (Steinhausen, 2002).
cognition. Both in the ACT and mindfulness
In this paper, I want to share some critical
approach, the aim is not so much to change the
reflections on why it remains so difficult to change
content of the cognition but becoming aware,
these typical eating disorder cognitions such as: ‘My
accepting and eventually distancing oneself from
belly is too fat’, ‘I need to be perfect’, ‘Nobody will
the cognitions. So feeling instead of thinking or in
ever love me at this weight’. . ..!
CBT terms some form of emotional exposure is
needed. Although ACT and meditation are finding
their way in the CBT field (Segal, Williams, &
Teasdale, 2002), CBT approaches in the field of
CRITICAL REFLECTIONS
eating disorders still focus too much on the content
Critical Reflection No. 1: Some Basic of the cognition. So my suggestion no. 2 is that in
Conditions Need to be Fulfilled some patients we probably have to focus less on the
content of the cognition and more on the over
In the last decade, evidence-based treatments have
involvement with the cognition. However I have to
become the gold standard also within the psy-
remark that at this time clear guidelines on when to
chotherapy field. Evidence-based treatments
focus on the content or the over involvement of the
suggest that patients will become better if one uses
cognition are lacking.
the right therapeutic strategy. Hence, the efficacy of
psychotherapeutic interventions seems to be largely
based on using the indicated ‘evidence-based
Critical Reflection No 3: Too Little Attention
technique’ for the presented problem. However,
to Family and Peer Influences
other aspects—which I call here basic conditions for
therapeutic progress —probably interfere strongly As a hard-working clinician and family therapist, I
with the therapeutic outcome. I want to am often surprised how parents indirectly suggest
mention two particular aspects namely (1) the to their daughters that it is better to stay anorectic
quality of the therapeutic alliance and (2) the timing via subtle, indirect messages and hence reinforce
when the therapeutic strategy will be introduced. anorectic thinking and being. During family ses-
With the exception of some recent studies (see for sions I am regularly totally shocked by the parents’
instance Loeb et al., 2005) both aspects have questions directed to their anorectic daughter such
received too little attention in the ED literature. as:
Especially in eating disorder patients, a good ‘Well sweetheart, is your belly not too fat?’
therapeutic alliance is a ‘condition sine qua non’ ‘Do you still have to put on more weight?’
before any therapeutic progress can be achieved. ‘Are you happy with your school results?’
And almost nothing seems to be possible as long as ‘Are you not relaxing too much?
the patient is not aware that ‘her thinking about her ‘Do you find yourself attractive with this mini
body and eating’ is part of the problem. CBT skirt?’
techniques aiming at changing the unrealistic ‘Is your target weight not too high?’
cognitions are not indicated when the patient is ‘Why is my daughter’s target weight higher then
still not aware of the fact that her cognitions may be my wife’s weight? (question to the therapist).

Copyright # 2008 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 16, 329–333 (2008)
DOI: 10.1002/erv
Failure of Cognitive Behavioural Therapy 331

These questions and many other observations been developed as an efficacious and efficient
suggest to me that disturbing cognitions may have treatment for post-traumatic stress disorder (PTSD).
their origins within the family context. Other Shapiro (2001) has developed an adaptive infor-
patients mention that negative remarks of their mation processing model, that posits that pathology
peers regarding their physical appearance have results when distressing experiences are processed
stimulated them in their decision to start a diet. inadequately and hypothesises that EMDR accel-
However, CBT approaches rarely give attention to erates information processing, resulting in the
family and peer influences. adaptive resolution of traumatic memories. EMDR
Hence my suggestion no. 3 is that more thera- has demonstrated effectiveness in changing mala-
peutic work is needed with the family system daptive schemata mostly in trauma patients. I
focusing on dysfunctional cognitions and messages believe that therapists can learn a lot about ‘how
within the family communications and interactions. to change disturbing cognitions’ by studying the
The same remark applies to influences of peers. different steps in the EMDR protocol. In the EMDR
treatment the patient is ‘exposed’ to the most vivid
visual image related to the memory (if available), a
Critical Reflection No. 4: Lack of Flexibility
negative belief about self, related emotions and
in Thinking is Probably Genetically Based
body sensations. Simultaneously the patient is
Notwithstanding all new therapeutic efforts (CBT, invited to move his/her eyes back and forth
mindfulness, etc. . .), we still are confronted with following the therapist’s fingers as they move
patients who are unable to change their anorectic across his/her field of vision for 20–30 second. This
thinking and beliefs. So I often ruminate ‘why are exposure results in activating all kind of emotional
these cognitions so resistant to change’? Tchanturia, (re)experiencing and the exposure goes on until the
Brecelj, Sanchez, Morris, Rabe-Hesketh, and Treas- moment that the emotional reactions start to
ure (2004) are doing neuropsychological research in decrease, preferably until the negative emotions
AN patients and may have found some under- are no longer present (desensitisation). Only when
standing of why these cognitions are so resistant to the emotional level has decreased sufficiently will
change. Tchanturia et al. (2004) discovered an the EMDR procedure start to focus on the installa-
important rigidity and lack of flexibility in the tion of new and more realistic cognitions. So it
thinking of anorectic patients, which is probably seems that—at least in trauma patients—the
genetically based. Neuropsychological studies have emotional reactions related to the trauma experi-
shown that AN patients of the restricting type have ence need to be targeted first, before new and
difficulties in cognitive flexibility. These laboratory- adequate cognitive schemata can be developed and
based findings have been used to develop a clinical installed. Here is my suggestion no. 5: Since many of
intervention, cognitive remediation therapy (CRT) the eating disorder patients are victims of abuse and
which aims to use cognitive exercises to strengthen hence may have developed ‘maladaptive schemata’
thinking skills. Tchanturia, Whitney, and Treasure as a consequence of these abusive experiences, first
(2006) and Tchanturia, Davies, and Campbell (2007) some ‘emotional work’ needs to be done before
are now experimenting with teaching the patients to cognitions can be targeted.
develop more flexibility in their thinking. Patients
have to practice every day and first results are
Critical Reflection No. 6: Too Much Focus on
promising. Hence my suggestion no. 4: In some
the Cognitive Level and not on the
patients—probably the chronic AN patients of the
Underlying Implicit Emotional Meanings
restricting type—changing maladaptive schemata
can be indirectly approached by clinical interven- My 6th reflection is based on recent developments
tions based on CRT. Again I have to remark that we in basic research in cognitive psychology and
do not know in which patients this approach is neuroscience and somehow further corroborates
indicated. my hypothesis in critical reflection no. 5. It suggests
that long-term effectiveness of CBT approaches may
be enhanced by going beyond symptoms at the
Critical Reflection No. 5: Too Much Focus on
cognitive level while expanding the therapeutic
Cognitions and Less on Emotional
focus to the underlying, implicit emotional mean-
Experiencing
ings. In a fascinating paper, Samoilov and Godfried
In recent years, eye movement desensitisation and (2000) state that the long-term effectiveness of CBT
reprocessing therapy (EMDR; Shapiro, 2001), has interventions may be enhanced by making more use

Copyright # 2008 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 16, 329–333 (2008)
DOI: 10.1002/erv
332 J. Vanderlinden

of the qualitative distinction between intellectual More research is needed studying the potential role
and emotional modes of mental processing. These of in-session emotional processing and (re)exper-
core affective structures may involve different iencing, a topic largely neglected in the eating
neural processes and may be subject to different disorder field until today. Only one study by
change principles than those involving cognitions. Castonguay, Pincus, Agras, and Hines (1998)
Greenberg, Rice, and Elliott (1993) and Samoilov reported that patients’ increased emotional experi-
and Godfried (2000) make a distinction between encing during group cognitive-behavioural treat-
(1) the rational/intellectual cognitive schemata ment for BED predicted positive treatment outcome.
(cold cognitions) and (2) emotional schemata (hot And next, we urgently need more specific
cognitions). According to the experiential perspec- guidelines indicating in which patients we need
tive (Greenberg & Safran, 1984) it is emotional to focus on the content of the cognition and in which
schemes, and not cognitive schemata, that lie at the patients we would do better to introduce some
core of people’s personal meanings. Thus, the in- mindfulness approach aiming at changing the over
session emotional expression is used to activate involvement with the cognition. In younger patients
client’s emotional schemes in order to restructure in order to alter the cognitions we certainly have to
old meanings and to create new ones (Greenberg & collaborate with the family, especially the parents.
Paivio, 1997). Within this model the in-session In the absence of clear guidelines, what is my
emotional arousal may be viewed as essential in experience telling me? Let me share some personal
exerting pressures towards reorganisation of under- views with you on this topic. Some patients—I
lying emotional themes, assimilation of new think about trauma patients—probably first will
information and formation of new implicit meaning need some ‘emotional (re)experiencing’, before
structures (Samoilov & Godfried, 2000). Again it cognitions can be tackled. Here both the EMDR
suggests (suggestion no. 6)—as it is the case in the approach and meditation seem promising. EMDR
EMDR treatment of Shapiro (2001)—that first some and meditation can be easily integrated within a
therapeutic actions are needed which provoke some CBT approach: both can be considered as a form of
‘emotional arousal’ within the brain and that exposure technique resulting in the desensitisation
‘emotional (re)experiencing’ is needed before new of negative affect. Other patients—I think about the
cognitions can be installed. more chronic AN patients of the restricting type
who do not respond to any treatment may have
developed—more then other patients—a geneti-
cally based rigidity in their thinking that needs to be
CONCLUSION: WHAT HAS TO BE approached indirectly by, for instance, some
CHANGED?
computer games promoting problem-solving skills
This paper has made the case that the widely used and flexibility and/or CRT. If we want to alter
evidence-based ‘CBT’ approaches remain unsuc- cognitions in younger patients and adolescents, the
cessful in many patients. I hope most of my involvement of other family members—at least the
colleagues will agree on that statement. Many parents—in the treatment will be a condition sine
patients seem to be unable to change their thinking qua non. So, many interesting assumptions and
about their bodies and themselves. Equally I hope questions waiting—I hope—for rigorous research.
that most of my colleagues will also agree on the
assumption that the therapeutic work aiming at
helping ED patients to adopt more adequate REFERENCES
cognitions is crucial to the recovery process. It is
therefore important to reflect on how we can help American Psychiatric Association. (1994). Diagnostic and
statistical manual of mental disorders (4th ed., Text Revi-
more patients in giving up their ‘typical eating sion). Washington, DC: American Psychiatric Associ-
disorder thoughts’. I have formulated six critical ation.
‘personal reflections’ on why many patients remain Castonguay, L. G., Pincus, A. L., Agras, W. S., & Hines,
totally resistant towards our therapeutic endea- C. E. (1998). The role of emotion in group cognitive-
vours. I believe that my reflections suggest that behavioral therapy for binge eating disorder: When
probably many roads may lead to Rome. However, things have to feel worse before they get better. Psy-
chotherapy Research, 8, 225–238.
by now, we just do not know when we have to take Fairburn, C. G., Marcus, M. D., & Wilson, G. T. (1993).
or follow which road. Therefore, more studies are Cognitive behaviour therapy for binge eating and
needed to find out why some patients benefit from bulimia nervosa: A comprehensive treatment manual.
our ‘classical’ CBT approach and others do not. In C. G. Fairburn, & G.T. Wilson (Eds.), Binge eating:

Copyright # 2008 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 16, 329–333 (2008)
DOI: 10.1002/erv
Failure of Cognitive Behavioural Therapy 333

Nature, assessment, and treatment (pp. 361–404). New nervosa: A study of process and outcome. Journal of
York: Guilford Press. Consulting and Clinical Psychology, 73, 1097–1107.
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive NICE guidelines. (2004). Eating disorders guidelines: Core
behaviour therapy for eating disorders: a ‘‘transdiag- interventions in the treatment and management of anorexia
nostic’’ theory and treatment. Behaviour Research and nervosa, bulimia nervosa and related eating disorders.
Therapy, 41, 509–528. London: National Institute of Clinical Excellence.
Garner, D. M., & Bemis, K. M. (1982). A cognitive-beha- Tchanturia, K., Brecelj, M., Sanchez, P., Morris, R., Rabe-
vioral approach to anorexia nervosa. Cognitive Therapy Hesketh, S., & Treasure, J. (2004). An examination of
and Research, 6, 123–150. cognitive flexibility in eating disorders. Journal of Inter-
Gowers, S. G., Clark, A., Roberts, C., Griffiths, A., national Neuropsychological Society, 10, 513–520.
Edwards, V., Bryan, C., et al. (2007). Clinical effective- Tchanturia, K., Whitney, J., & Treasure, J. (2006). Can
ness of treatments for anorexia nervosa in adolescents: cognitive exercises help treat anorexia nervosa? Eating
Randomised controlled trial. British Journal of Psychia- and Weight Disorders, 11, 112–116.
try, 191, 427–435. Tchanturia, K., Davies, H., & Campbell, I. C. (2007).
Greenberg, L. S., & Safran, J. D. (1984). Integrating affect and Cognitive remediation therapy for patients with anor-
cognition: A perspective on the process of therapeutic exia nervosa: Preliminary findings. Annals of General
change. Cognitive Therapy and Research, 8, 559–578. Psychiatry, 5, 6–14.
Greenberg, L. S., & Paivio, S. (1997). Working with emotions: Samoilov, A., & Goldfried, M. R. (2000). Role of emotion in
Changing core schemes. New York: Guilford Press. cognitive-behavior therapy. Clinical Psychology: science
Greenberg, L. S., Rice, L. N., & Elliott, R. (1993). Facilitating and practice, 7, 373–385.
emotional change: The moment-by-moment process. New Segal, Z. V., Williams, J. M., & Teasdale, J. D. (2002).
York: Guilford Press. Mindfulness-based cognitive therapy for depression: A new
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). approach to preventing relapse. New York: Guilford Press.
Acceptance and commitment therapy an experiential Shapiro, Fr. (2001). Eye movement desensitization and repro-
approach to behavior change. NY: Guilford Press. cessing (EMDR): Basic principles, protocols, and procedures
Kabat Zinn, John. (1990). Full catastrophe living: Using the (2nd ed.). New York: Guilford Press.
wisdom of your body and mind to face stress, pain and Steinhausen, H. C. (2002). The outcome of anorexia ner-
illness. New York: Delta. vosa in the 20th century. American Journal of Psychiatry,
Loeb, K. L., Wilson, G. T., Labouvie, E., Pratt, E. M., Hayaki, 159, 1284–1293.
J., Walsh, B. T., et al. (2005). Therapeutic alliance and Wilson, G. T. (1997). Cognitive behavioral treatment of
treatment adherence in two interventions for bulimia bulimia nervosa. The Clinical Psychologist, 50, 10–12.

Copyright # 2008 John Wiley & Sons, Ltd and Eating Disorders Association. Eur. Eat. Disorders Rev. 16, 329–333 (2008)
DOI: 10.1002/erv