According to the transdiagnostic theory of eating disorders (EDs), a treatment which targets maintenance factors (such as an over evaluation of weight and shape, and dietary restraint) shared across all EDs will result in many patients responding well, regardless of their diagnosis (Fairburn, Cooper & Shafran, 2003).

Since research has found that most ED patients tend to fluctuate between diagnosis over time (Fairburn & Harrison, 2003), it makes sense that a treatment for all individuals with EDs, regardless of their diagnosis would be the best course of action. Following limitations with treatments specific to diagnoses (such a Bulimia Nervosa specific CBT; CBT-BN) and the above transdiagnostic theory, an enhanced CBT (CBT-E) was developed for the treatment of EDs.

Previous research and systematic reviews (Groff, 2015; de Jong, Schoorl & Hoek, 2018; Dahlenburg, Gleaves & Hutchinson, 2019) on CBT-E have found it to be efficacious across the spectrum of EDs. However, an updated systematic review to assess the effectiveness and efficacy of CBT-E is required due to limitations of the previous reviews (such as inconsistencies in the reporting of effect sizes/ED outcomes, alongside statistical methods overestimating the effectiveness of the treatment) and a need for up-to-date research to be included.

Therefore, the current systematic review (Atwood & Friedman, 2019) aimed to summarise and critically evaluate the current evidence base and extend upon previous reviews by:

1. Updating the evidence since 2017
2. Reporting between-groups effect sizes
3. Reporting effect sizes for attrition, remission, ED behaviour, BMI outcomes and core ED psychopathology

Methods

A systematic review was conducted to identify all papers reporting on the effectiveness or efficacy of enhanced CBT (CBT-E) in adults or older adolescents with an eating disorder (ED), according to the Preferred Reporting Items for Systematic Review and Meta-analysis guidelines (PRISMA; Moher et al., 2009).

The two authors of this review each independently searched for papers via two databases in June 2019, with the references of relevant papers also manually searched.

Inclusion criteria

• Studies which used a quantitative design
• Participants who met the diagnostic criteria for an ED as defined in the DSM-5 or DSM-IV-TR
• Recruited adults or older adolescents (≥16 years old)
• Used an uncontrolled design examining the effectiveness of CBT-E from pre- to post-treatment; or a randomised controlled trial design (RCT) comparing CBT-E to a delayed treatment control, current standard of care or other psychotherapy (including comparing CBT-E with and without an additional treatment component)
• Reported on an outcome of interest, including treatment attrition, remission rates, ED behaviours, BMI, and/or core ED psychopathology
• A minimum of two assessment points at pre- and post-treatment were included

Exclusion criteria

• Case studies or case series

Quality assessment

Each author independently reviewed each article according to a modified version of the 27-item Downs and Black checklist (Trac et al., 2015) to assess quality. A quality index (QI) score was assigned to each trial.

The degree of agreement between the two authors was measured using an intraclass correlation coefficient (ICC) statistic, with any disagreements resolved by consensus.

Effect size reporting

The efficacy of CBT-E was compared using Cohen’s d effect size. Where trials did not report an effect size, these were calculated by the authors.

Results

Of the 1,445 articles identified through the literature searches, 20 papers were included in this systematic review (10 RCTs and 10 uncontrolled trials). Results for outcomes of interest are reported below:

Attrition

• For randomised controlled trials (RCTs):
  • Attrition rates ranged from 20-54%, except for one between-groups study (Zipfel et al., 2014) which found significantly higher attrition in a treatment as usual (TAU) group compared to CBT-E
  • Rates appeared higher in studies including anorexia nervosa (AN) participants, except for one study (Wade, Byrne & Allen, 2017) which found no differences between diagnoses
• For uncontrolled trials:
  • Attrition rates ranged from 15-50%, with most studies finding no difference in attrition by diagnosis
  • A significant positive association in one study (Byrne, Fursland, Allen, & Watson, 2011) was found between attrition and longer wait times for treatment

Remission

• The definition of remission varied between the studies included in this review, as it does in the literature
• Remission rates varied, typically falling between 30-50%, which is likely due to the complicated nature of defining remission in transdiagnostic samples, and the varying criteria for remission between (similar) definitions
• Whilst CBT-E was found to be superior to a delayed treatment control in RCT’s, no differences were found between CBT-E and comparison treatments, particularly in the longer term
• For uncontrolled trials, rates of remission for participants with AN were higher than for other diagnoses

Eating disorder behaviours

• For RCTs:
  • CBT-E appeared to result in moderate to large decreases in binge eating and small to moderate decreases in purging behaviours, except in those with BN where these effects were large
  • CBT-E did not seem to be superior to other treatments, except for psychoanalytic psychotherapy (PP)
• For uncontrolled trials:
  • Small to moderate effect sizes were reported in transdiagnostic samples, with reports of reductions in purging and binge eating (with a large effect size)
  • No significant decreases were found in AN samples at post-treatment or follow-up

Body mass index (BMI)

• Large significant increases in BMI were seen for those with AN after CBT-E in both RCTs and uncontrolled trials
• There was no evidence to suggest CBT-E to be superior to other treatments

ED psychopathology

• For RCTs:
  • There was strong support for CBT-E in reducing ED psychopathology across ED diagnoses, however, there was no evidence to suggest that CBT-E is superior to other treatments
• For uncontrolled trials:
  • Medium to large effect sizes were found across those with a transdiagnostic ED sample, with large effects found specifically for AN samples

Conclusions

• From the results of this systematic review, there is evidence to support CBT-E as an effective and efficacious treatment for EDs in adults and older adolescents, across the spectrum of ED diagnoses and for multiple outcomes of interest
• However, despite its effectiveness, the results of this systematic review do not show CBT-E to be superior to other treatments.

Strengths and limitations

This systematic review has many strengths:

• The authors calculated effect sizes for articles that did not report them. This allowed them to see the size of the effects in each individual study and allowed for the comparison of different articles included in the review
• The authors independently searched for articles and reviewed their quality, with their inter-rater reliability for their quality ratings found to be excellent (ICC=.95)
• This review included high-quality studies (as indicated by the QI scores), which used large and often transdiagnostic samples. Additionally, all but three used intention to treat (ITT) analyses (providing confidence in the effect sizes) and most studies reported clinically significant change alongside statistically significant change. This helps to ensure that the effect of the treatment is realistic and not overestimated.

However, it also had some limitations:

• Most of the samples comprised mainly White participants and >90% of the samples were female dominated, suggesting that there may have been a selection bias. This may well limit the generalisability of the review findings
• Most studies did not blind researchers in the assessment of outcomes
• The authors did not provide comment on the mechanisms of change relevant to the transdiagnostic model. Studies included in this review did measure things such as interpersonal difficulties, mood intolerance, perfectionism, and self-esteem, yet the mechanisms of change of these following CBT-E was not reported
• Only two databases were searched for this review, which also did not include grey literature. Even though reference lists were manually searched, key articles may still have been missed.

Implications for practice

One major limitation of eating disorders (ED) research in general is the homogenous samples used: mainly White females. Therefore, there is a dire need for ED research to include other neglected groups. This would include more research with males, who are believed to make up 25% of ED sufferers (Beat, 2017). Whilst males are less likely to seek help for EDs (Beat, 2017), research still appears to be actively excluding them, which means treatments will not be suited to their needs. Additionally, ED research is needed in minority populations (Marques et al., 2011). We currently do not know how effective CBT-E is for these groups, so research is required to ascertain its cultural sensitivity.

Furthermore, whilst the results from this review indicate that CBT-E is not more effective than other treatments, there was evidence to suggest that its effects were quicker. For example, when assessing improvements in core psychopathology, improvements from CBT-E may be seen after 5 months, compared to 2 years after PP (Poulson et al., 2014), along with CBT-E resulting in faster weight restoration (Zipfel et al., 2014). These things have various implications. Not only might it be more cost-effective to implement CBT-E over other ED treatments, but this may also help to improve access to treatment (as it might be cheaper to implement). Given the importance of reducing wait-times to treatment to improve prognosis, future research should explore how a more rapid response to treatment can be provided (MacDonald, McFarlane, Dionne, David & Oldmstead, 2017), which might be achieved by a faster acting and more cost-effective treatment such as CBT-E.

Statement of interests

None.

Links

Primary paper

Atwood, M. E., & Friedman, A. (2019). A systematic review of enhanced cognitive behaviour therapy (CBT-E) for eating disorders (PDF). International Journal of Eating Disorders, 53, 311-330.

Other references

Beat. (2017). Types of eating disorders.

Byrne, S. M., Fursland, A., Allen, K. L., & Watson, H. (2011). The effectiveness of cognitive behavioural therapy for eating disorders: An open trial. Behaviour Research and Therapy, 49, 219-226. [Science Direct Abstract]

Dahlenburg, S. C., Gleaves, D. H., & Hutchinson, A. D. (2019). Treatment outcome research of enhanced cognitive behaviour therapy for eating disorders: systematic review with narrative and meta-analysis synthesis (PDF). Eating Disorders, 27, 482-502.

de Jong, M., Schoorl, M., & Hoek, H. W. (2018). Enhanced cognitive behavioural therapy for patients with eating disorders: A systematic review (PDF). Current Opinion in Psychiatry, 31, 18-23.

Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive behavioural therapy for eating disorders: A transdiagnostic theory and treatment. Behaviour Research and Therapy, 41, 509-528.  [Science Direct Abstract]

Fairburn, C. G., & Harrison, P. J. (2003). Eating disorders. Lancet, 361, 407-416. [PubMed Abstract]

Groff, S. E. (2015). Is enhanced cognitive behavioural therapy an effective intervention in eating disorders? A review. Journal of Evidence-Informed Social Work, 12, 272-288. [Taylor and Francis Online Abstract]

MacDonald, D. E., McFarlane, T. L., Dionne, M. M., David, L., & Oldmstead, M. P. (2017). Rapid response to intensive treatment for bulimia nervosa and purging disorder: A randomized controlled trial of a CBT intervention to facilitate early behaviour change. Journal of Consulting and Clinical Psychology, 85, 896-908. [APA PsychNet Abstract]

Marques, L., Algeria, M., Becker, A. E., Chen, C., Fang, A., Chosak, A., & Diniz, J. B. (2011). Comparative prevalence, correlates of impairment, and service utilization for eating disorders across the U.E. ethnic groups: Implications for reducing ethnic disparities in healthcare access for eating disorders (PDF). International Journal of Eating Disorders, 44, 412-420.

Moher, D., Liberati, A., Tetzlaiff, J., Altman, D. G., & The PRISMA Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement (PDF). Annals of Internal Medicine, 151, 264-269.

Poulson, S., Lunn, S., Daniel, S. I. F., Folke, S., Mathiesen, B. B., Katznelson, H., & Fairburn, C. G. (2014). A randomized controlled trial of psychoanalytic psychotherapy or cognitive-behavioural therapy for bulimia nervosa (HTML). American Journal of Psychiatry, 171, 109-116.

Trac, M. H., McArthur, E., Jandoc, R., Dixon, S. N., Nash, D. M., Hackman, D. G., & Garg, A. X. (2015). Macrolide antibiotics and the risk of ventricular arrhythmia is older adults (PDF). Canadian Medical Association Journal, 188, 120-129.

Wade, S., Byrne, S., & Allen, K. (2017). Enhanced cognitive behavioural therapy for eating disorders adapted for a group setting (PDF). International Journal of Eating Disorders, 50, 863-872.

Zipfel, S., Wild, B., Grob., G., Friederich, H. -C., Teufel, M., Schellberg, D., … & ANTOP study group. (2014). Focal psychodynamic therapy, cognitive behaviour therapy, and optimised treatment as usual in outpatients with anorexia nervosa (ANTOP study): Randomised controlled trial (PDF). The Lancet, 383, 127-137.

Photo credits

• Photo by Sharon McCutcheon on Unsplash
• Photo by Hudson Hintze on Unsplash
• Photo by Clay Banks on Unsplash
• Photo by Arisa Chattasa on Unsplash
• Photo by Adam Winger on Unsplash
• Photo by Andrew Neel on Unsplash