Due 10/24/19

EMPIRICAL ARTICLE

Cognitive-Behavioral Therapy for Bulimia Nervosa and Atypical Bulimic Nervosa:

Effectiveness in Clinical Settings

Glenn Waller, DPhil1* Emma Gray, DClinPsych2

Hendrik Hinrichsen, DClinPsych3

Victoria Mountford, DClinPsy4,5

Rachel Lawson, MA6

Eloise Patient, BSc7

ABSTRACT Objective: The efficacy of cognitive- behavioral therapy (CBT) for bulimic dis- orders has been established in research trials. This study examined whether that efficacy can be translated into effective- ness in routine clinical practice.

Method: Seventy-eight adult women

with bulimic disorders (bulimia nervosa

and atypical bulimia nervosa) under-

took individual CBT, with few exclusion

criteria and a treatment protocol based

on evidence-based approaches, utilizing

individualized formulations. Patients

completed measures of eating behav-

iors, eating attitudes, and depression

pre- and post-treatment. Eight patients

dropped out. The mean number of ses-

sions attended was 19.2.

Results: No pretreatment features pre- dicted drop-out. Treatment outcome was similar whether using treatment com-

pleter or intent to treat analyses. Approxi- mately 50% of patients were in remission by the end of treatment. There were sig- nificant improvements in mood, eating attitudes, and eating behaviors. Reduc- tions in bingeing and vomiting were comparable to efficacy trials.

Discussion: The improvements in this “real-world” trial of CBT for adults with bulimic disorders mirrored those from large, funded research trials, though the conclusions that can be reached are inevitably limited by the nature of the trial (e.g., lack of control group and therapy validation). VC 2013 Wiley Periodicals, Inc.

Keywords: bulimia nervosa; atypical bulimic disorders; cognitive-behav- ioral therapy; effectiveness

(Int J Eat Disord 2014; 47:13–17)

Introduction

There is substantial evidence that cognitive- behavioral therapy (CBT) is efficacious in the treat- ment of adult women with bulimia nervosa and atypical bulimic disorders.

1–7 However, that evi-

dence has come from funded research studies. Such findings are not necessarily generalizable to the wider range of clinical settings, due to factors such as the exclusion of comorbidity or atypical cases, treatment being delivered under highly stringent conditions, and the need to adhere strictly to proto- cols. Thus, such evidence of efficacy in the research environment needs to be translated into evidence of effectiveness in less specialized clinical practice, in order to avoid clinicians ignoring the evidence as being irrelevant to their client group.8 This attitude might explain the common omission of core techni- ques when delivering CBT for adults with eating dis- orders9 and the fact that only a minority of clinicians report using manuals when working with bulimia nervosa.

10 There is evidence for the clinical

applicability of research-based CBT for anxiety and depression.11,12 However, that is not the case in the eating disorders. Therefore, this study considered whether the efficacy of CBT for bulimic disorders (as shown by existing research trials) can be trans- lated into clinical effectiveness in routine clinical settings, where none of the exclusion- and protocol- based constraints outlined above apply.

Accepted 27 July 2013

*Correspondence to: Glenn Waller, Clinical Psychology Unit,

Department of Psychology, University of Sheffield, Western Bank,

Sheffield S10 2TN, UK. E-mail: [email protected] 1 Clinical Psychology Unit, Department of Psychology, University

of Sheffield, Sheffield, Sheffield, England, United Kingdom 2 British CBT and Counselling Service, London, England, United

Kingdom 3 Sutton and Merton IAPT Service, South West London and St.

George’s Mental Health NHS Trust, Springfield University Hospital,

London, England, United Kingdom 4 Eating Disorders Section, Institute of Psychiatry, King’s College

London, London, England, United Kingdom 5 Eating Disorders Service, South London and Maudsley NHS

Foundation Trust, Denmark Hill, London, England, United

Kingdom 6 South Island Eating Disorders Service, Christchurch, New

Zealand 7 North Staffordshire Wellbeing Service, Newcastle-under-Lyme,

Staffordshire, England, United Kingdom

Published online 1 September 2013 in Wiley Online Library

(wileyonlinelibrary.com). DOI: 10.1002/eat.22181 VC 2013 Wiley Periodicals, Inc.

International Journal of Eating Disorders 47:1 13–17 2014 13

Method

Participants

All patients were treated in a publically-funded outpa-

tient eating disorder service in the UK. The only exclu-

sion criteria were psychosis, learning difficulties, or an

inability to work in English. The participants in the trial

were a case series of those patients with bulimic disor-

ders who opted to undertake CBT when assessed and

when treatment options were discussed. A small number

of patients opted to undertake a psychodynamic therapy,

while another group attended for assessment but

declined or failed to attend for treatment. However, the

numbers in these groups were not recorded. Therefore,

this is a study of those who opted to begin CBT, rather

than all who attended the clinic or who had bulimic dis-

orders. None were excluded from the trial due to missing

data (see below).

The sample entering treatment were 78 adult women

with bulimic disorders: 55 with bulimia nervosa (52

purging subtype; three nonpurging subtype) and 23 with

EDNOS involving bulimic behaviors (nine with subthres-

hold bulimia nervosa, involving bingeing and purging at

least once per week over three months; 10 with binge

eating disorder; and four with purging in the absence of

bingeing). None were in the anorexic weight range. All

were assessed using a semistructured interview proto-

col, 13

and diagnosed using DSM-IV criteria. 14

The mean

age of the sample at assessment was 27.8 years (SD 5

7.11) and their mean body mass index (BMI) was 22.1

(SD 5 3.26).

A minority of the bulimic sample (N 5 9) were receiv-

ing SSRI antidepressants (stabilized prior to treatment

and maintained throughout the treatment period) and a

small number had occasional dietetic reviews, but none

were receiving any other form of treatment in parallel

with CBT. A high proportion had some comorbidity

(major depressive disorder: 44% of cases; obsessive-

compulsive disorder: 26%; other anxiety disorders: 32%;

and substance misuse: 23%). The levels for anxiety and

depressive disorders are higher than in some efficacy

studies,3 but comparable for substance misuse.

Measures

Height and weight were measured objectively. Diaries

were used to assess frequency of bingeing and vomiting.

The women also completed measures of eating pathol-

ogy and depression at the beginning and end of

treatment.

Eating Disorders Inventory. The EDI15 is a self-report

measure of eating and related attitudes. Scores are

responsive to changes over treatment. Scores on the

eating-related scales (Drive for thinness; Bulimia; Body

dissatisfaction) were summed to provide an overall score

reflecting eating attitudes.

Beck Depression Inventory. The BDI16 is a self-report

measure of depressive symptoms, with good psychomet-

ric properties.

Treatment. This version of CBT13 is based on techni-

ques employed in evidence-based approaches to bulimia

nervosa (I,2,4). In keeping with those approaches, this

programme includes: individualized formulation, taking

into account different maintaining factors across cases

(e.g., nutritional and/or emotional drivers for binging);

agenda setting; homework; change in diet (particularly to

improve carbohydrate intake); diary-keeping; exposure;

behavioral experiments; cognitive restructuring; and sur-

veys. Comorbidity was usually addressed once the core

eating disorder symptoms were substantially reduced,

unless there was evidence that the patient was not

changing eating behaviors in the early part of CBT.

The clinicians were all clinical psychologists with at

least four years of experience in delivering CBT for eating

disorders, and were supervised routinely on these cases

(individually and in groups). The usual assumption was

that there would be around 20 one-hour CBT sessions.

However, this was reduced if the patient improved rap-

idly, and was increased if the patient had substantial

comorbidity (such increases were reviewed by the team

after each additional set of 10 sessions). Whatever the

duration, behavioral change was maintained as a focus,

along with changes in mood and cognitions. The mean

number of sessions delivered per patient was 19.2 (SD 5

12.4; range 5 7–80). Three follow-up sessions were

offered in addition.

Patients were regarded as in remission if they no lon-

ger had a diagnosis of any eating disorder by the end of

treatment (including being free of bulimic behaviors for

at least a month prior to the last session, and not having

pathological concerns about eating, weight, and shape).

This latter criterion was established through clinical

review. Drop-out was defined as the patient ending treat-

ment before the agreed termination point (defined by

patient and clinician), whether early or late in treatment.

Data Analysis

Binary logistic regression was used to identify any pre-

treatment factors that predicted drop-out.17 Change was

measured in three ways—as the proportion of patients

ceasing individual and all bulimic behaviors (objective

binge-eating, vomiting, and laxative abuse) by the end of

treatment; as the proportion of patients who changed or

no longer met diagnostic criteria at the end of treatment

(remission); and as the dimensional differences in fre-

quencies of behaviors, BMI level, eating attitudes, and

depression. As the data were not sufficiently normal,

changes in symptom levels were tested using Wilcoxon

WALLER ET AL.

14 International Journal of Eating Disorders 47:1 13–17 2014

tests. This final analysis is done as both a treatment com-

pleter analysis and an intent-to-treat analysis (carrying

forward the most recent data to substitute for missing

data where a patient dropped out). At the end of treat-

ment, there were nine missing EDI scores and 12 missing

BDI scores. These were treated as absent for the com-

pleter analysis.

Results

Predictors of Attrition

Eight of the 78 patients dropped out over the course of treatment. This rate is similar to that reported in protocol-driven research studies.

3,4,18

Binary logistic regression was used to determine whether drop-out was related to pretreatment age, BMI, frequency of bingeing or vomiting, EDI scores, or BDI scores. The overall model was not significant (X2 5 11.5, df 5 6, p 5 0.075), and no individual variable approached significance (p > 0.16 in all cases). Therefore, no identified pretreat- ment variables predicted attrition from CBT.

Cessation of Bulimic Behaviors Following CBT

Among the completer group, 66 engaged in objec- tive binging at the beginning of treatment, and 28 at the end of treatment (abstinent 5 58%), 51 engaged in vomiting at the beginning and 25 at the end (abstinent 5 51%), and 17 engaged in laxative abuse at the beginning and three at the end (abstinent 5 82%). 56% were free of all binging and purging behaviors by the end of treatment (all patients had at least one of these behaviors at the start of treat- ment). These reductions are comparable with those reported across treatment in clinical trials.

3

Change in Diagnosis Following CBT

Table 1 shows shifts in diagnoses from beginning to end of CBT. Overall, 37 (52.9%) of the 70 patients who reached the end of treatment were diagnosis- free by that point (no bulimic behaviors, alongside

normalized eating attitudes). Assuming no such changes among the eight patients who had dropped out, this represents 47.4% of the 78 patients who started CBT. This remission rate is similar to that found in comparable research.

2–4

Diagnostic group at the outset of treatment was not broadly predictive of change in diagnosis. How- ever, those with purging disorder showed a mixture of positive and negative outcomes, suggesting that this form of CBT is more suitable for those who binge-eat.

Dimensional Change in Symptoms Following

CBT

Table 2 shows changes across therapy in body mass index, frequencies of bulimic behaviors, eat- ing attitudes, and depression. This is done sepa- rately for treatment completers (N 5 70) and as an intent to treat analysis (N 5 78). Regardless of the form of analysis, there were significant changes on all of these measures, with a small increase in BMI and larger reductions in objective binges, vomiting, eating attitudes, and depressed mood. The effect sizes (tau 5 Z/�[N]) for these changes varied between medium and large in both sets of analy- ses. The frequency of objective bingeing fell by 59% in the treatment completer analysis (intent to treat: 64%), and vomiting levels fell by 72% (intent to treat: 65%). These findings are similar to levels of change reported.7

Discussion

Research trials have demonstrated the efficacy of CBT for bulimic disorders. However, clinicians commonly regard such findings as irrelevant to their practice.8 Therefore, this study tested whether those findings can be translated into evidence of effectiveness in healthcare settings where there are few exclusion criteria and where the implementa- tion of the therapy is less intensively scrutinized. It

TABLE 1. Diagnostic outcomes at end of treatment, among those completing CBT (N 5 70)

Diagnostic Group at Beginning of Treatment

Diagnostic Group at End of Treatment

No Eating Disorder

Bulimia nervosa EDNOS

Purging Subtype

Nonpurging Subtype

Atypical Bulimia Nervosa

Binge Eating Disorder

Purging Disorder

Bulimia nervosa Purging subtype (N 5 46) 22 (47.8%) 22 (47.8%) 0 2 (4.3%) 0 0 Nonpurging subtype (N 5 1) 1 (100%) 0 0 0 0 0 EDNOS Atypical bulimia nervosa (N 5 9) 5 (55.6%) 0 0 4 (44.4%) 0 0 Binge eating disorder (N 5 10) 8 (80.0%) 0 0 0 2 (20.0%) 0 Purging disorder (N 5 4) 1 (25.0%) 0 1 (25.0%) 1 (25.0%) 0 1 (25.0%)

CBT FOR BULIMIC DISORDERS

International Journal of Eating Disorders 47:1 13–17 2014 15

also included atypical bulimic cases. The findings were broadly comparable to those found in research trials—the drop-out rate was low (10.3%), the remission rate was �50%, and there were sub- stantial reductions in levels of pathological eating attitudes and depression. In short, these findings demonstrate that this form of CBT for bulimia nervosa13 is effective in treating the eating disor- ders in “real-life” clinical settings. However, it is important to note that these results were achieved by clinicians within a specialist eating disorder clinic, who had relatively high levels of training, experience, and supervision. Its effectiveness in other settings or as delivered by less experienced clinicians remains to be demonstrated. It is also necessary to note that this was an uncontrolled trial, with no validation checks (beyond routine supervision) to confirm that the therapy delivered actually was CBT. While these features are inevita- ble consequences of delivering treatments in real life settings, they limit the strength of any conclu- sions that can be reached regarding the effective- ness of CBT.

Several forms of CBT for adults with eating disor- ders have been shown to be efficacious to a compa- rable degree in research settings.1–4 Each shares themes with the form used here—particularly the foci on individualized formulation, exposure, behav- ioral change, recording, and cognitive restructuring. Therefore, these findings suggest that other evidence-based forms of CBT for the bulimic disor- ders might have similar levels of effectiveness in purely clinical settings, although the lack of a con- trol group in studies of this sort makes it impossible to conclude definitively that it is elements of CBT that are responsible for the positive outcomes seen. Clinicians could be encouraged to use existing man- ualized forms of CBT for bulimia nervosa more than they currently do,

9,10 on the grounds that these rela-

tively structured forms of treatment for the eating

disorders can be as effective in everyday clinical set- tings as they are efficacious in research settings. Further work is needed to determine whether these effects are maintained in the long term, as they are in research trials. Such research would also benefit from a wider range of measures of eating pathology (e.g., body image, other purging behaviors), as the measures used here were relatively crude, and might have omitted key indices of change. It should also consider the potential role of factors such as dura- tion of disorder, previous treatment experiences, and socio-economic status, to allow for comparison with existing and future clinical trials. Finally, it will be important to determine whether the efficacies of other therapies for bulimic disorders (e.g., interper- sonal psychotherapy; dialectical behavior therapy) are matched by their effectiveness.

References

1. Bulik CM, Sullivan PF, Carter FA, McIntosh VV, Joyce PR. Predictors of rapid

and sustained response to cognitive-behavioral therapy for bulimia nervosa.

Int J Eat Disord 1999;26:137–144.

2. Fairburn CG, Norman PA, Welch SL, O’Connor ME, Doll HA, Peveler RC. A pro-

spective outcome study in bulimia nervosa and the long-term effects of

three psychological treatments. Arch Gen Psychiatry 1995;52:304–312.

3. Fairburn CG, Cooper Z, Doll HA, O’Connor ME, Bohn K, Hawker DM, et al.

Transdiagnostic cognitive-behavioral therapy for patients with eating disor-

ders: A two-site trial with 60-week follow-up. Am J Psychiatry 2009;166:311–

319.

4. Ghaderi A. Does individualization matter? A randomized trial of standar-

dized (focused) versus individualized (broad) cognitive behavior therapy for

bulimia nervosa. Behav Res Ther 2006;44:273–288.

5. National Institute for Clinical Excellence. Eating Disorders: Core Interventions

in the Treatment and Management of Anorexia Nervosa, Bulimia Nervosa

and Related Eating Disorders (Clinical Guideline 9). London, UK: National

Collaborating Centre for Mental Health, 2004.

6. Shapiro JR, Berkman ND, Brownley KA, Sedway JA, Lohr KN, Bulik CM. Buli-

mia nervosa treatment: A systematic review of randomized controlled trials.

Int J Eat Disord 2007;40:321–336.

7. Vitousek KB. The current status of cognitive behavioural models of anorexia

nervosa and bulimia nervosa. In: Salkovskis PM, editor. Frontiers of Cogni-

tive Therapy. New York, NY: Guilford, 1996, pp. 383–418.

TABLE 2. Symptom change across treatment among those completing CBT (N 5 70) and using intent-to-treat analysis (N 5 78)

Type of Analysis Treatment Completers Intent-to-Treat

Point in Treatment

Beginning of Treatment

End of Treatment

Wilcoxon Test Effect

Size Tau

Beginning of Treatment

End of Treatment

Wilcoxon Test Effect

Size TauMean SD Mean SD Z P Mean SD Mean SD Z P

Symptom* Body mass index 22.8 (3.58) 23.3 (3.80) 2.98 .003 0.36 22.1 (3.26) 22.8 (3.67) 3.61 .001 0.41 Objective binges per week 4.45 (6.51) 1.84 (5.07) 3.25 .001 0.39 5.51 (9.65) 1.98 (4.86) 4.93 .001 0.56 Vomiting per week 5.31 (7.96) 1.49 (3.28) 2.93 .003 0.35 6.88 (10.1) 2.41 (5.29) 2.83 .005 0.32 Eating disorders inventory 46.0 (15.4) 26.3 (21.0) 2.39 .02 0.29 37.6 (21.8) 24.1 (22.5) 2.58 .01 0.29 Beck depression inventory 22.7 (9.95) 12.2 (11.0) 4.78 .001 0.57 21.9 (9.46) 12.1 (10.5) 4.59 .001 0.52

*No missing data for body mass index, objective binges, or vomiting episodes, and no missing start of treatment data for other variables. Missing data for eating disorders Inventory and beck depression inventory mean that completer N 5 61 and 58, respectively.

WALLER ET AL.

16 International Journal of Eating Disorders 47:1 13–17 2014

8. Tobin DL, Banker JD, Weisberg L, Bowers W. I know what you did last

summer (and it was not CBT): A factor analytic model of international psy-

chotherapeutic practice in the eating disorders. Int J Eat Disord 2007;40:

754–757.

9. Waller G, Stringer H, Meyer C. What cognitive-behavioral techniques do

therapists report using when delivering cognitive-behavioral therapy for the

eating disorders. J Consult Clin Psychol 2012;80:171–175.

10. Wallace LM, von Ranson KM. Treatment manuals: Use in the treatment of

bulimia nervosa. Behav Res Ther 2011;49:815–820.

11. Persons JB, Bostram A, Bertagnolli A. Results of randomized controlled trials

of cognitive therapy for depression generalize to private practice. Cogn Ther

Res 1999;23:535–548.

12. Persons JB, Roberts NA, Zalecki CA, Brechwald WAG. Naturalistic outcome of

case formulation-driven cognitive-behavior therapy for anxious depressed

outpatients. Behav Res Ther 2006;44:1041–1051.

13. Waller G, Corstorphine E, Cordery H, Hinrichsen H, Lawson R, Mountford V,

et al. Cognitive-Behavioral Therapy for the Eating Disorders: A Comprehen-

sive Treatment Guide. Cambridge, UK: Cambridge University Press, 2007.

14. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders, 4th ed. Washington, DC: American Psychiatric Association, 1994.

15. Garner DM. Eating Disorder Inventory-2. Odessa, FL: Psychological Assess-

ment Resources, 1991.

16. Beck AT, Steer RA. Beck Depression Inventory manual. San Antonio, TX: The

Psychological Corporation, 1987.

17. Carter O, Pannekoek L, Fursland A, Allen KL, Lampard AM, Byrne SM.

Increased wait-list time predicts dropout from outpatient enhanced cogni-

tive behaviour therapy (CBT-E) for eating disorders. Behav Res Ther 2012;50:

487–492.

18. Mitchell JE. A review of the controlled trials of psychotherapy for bulimia

nervosa. J Psychosom Res 1991;35:23–31.

CBT FOR BULIMIC DISORDERS

International Journal of Eating Disorders 47:1 13–17 2014 17

Copyright of International Journal of Eating Disorders is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.