ORIG I NALRESEARCH

PAPERS

Cognitive-behavioral therapy versuscombined treatment with grouppsychoeducation and fluoxetine inbulimic outpatients

V. Ricca*, E. Mannucci**, B. Mezzani ***, M. Di Bernardo*, E. Barciulli*, S. Moretti*,P. L. Cabras*, and C. M. Rotella***Psychiatric Unit, Department of Neurologic and Psychiatric Sciences, **Section of Metabolic Diseases, Unitof Endocrinology, Department of Pathophysiology, University of Florence, ***Casa di Cura “Villa dei Pini”,Florence, Italy

ABSTRACT. A series of 51 female bulimic outpatients, aged 23.4±3.9, were assigned eitherto Cognitive-Behavioral Therapy (CBT) or combined Group Psychoeducation and Fluoxetine(GPF) treatment. The Eating Disorder Examination (EDE) was performed at the beginning otreatment and after 6 months, together with the administration of self reported questionnaires for depression (BDI) and Anxiety (STAI). A significant (p<0.001) reduction of the number of monthly binge episodes (from 25.0±12.9 to 6.2±3.8 and from 24.8±9.1 to 8.0±4.3) forCBT and GPF respectively were observed. Similar reductions were obtained in the number oepisodes of compensatory behaviors. Both treatments reduced depression and anxiety(p<0.001) while CBT only determined a significative improvement of EDE scores. The dataobtained suggest that GPF is as effective as CBT in reducing bulimic symptomatology but itslong-term efficacy should be evaluated in a follow-up study.(Eating Weight Disord. 2, 94-99, 1997). ©1997, Editrice Kurtis

INTRODUCTION

Since the clinical description of BulimiaNervosa was defined by Russell in 1979 as adistinct, common diagnostic entity (1), studieson the treatment have focused on both psy-chological and pharmacological interventions.

Cognitive-behavior therapy (CBT) is themost extensively studied psychotherapeuticintervention (2). Controlled trials haveshown that it significantly reduces vomitingand binge eating; it also improves disturbedattitudes toward weight and shape, andreduces dietary restraint. In contrast withantidepressant drugs and behavior therapy(3), the results are maintained for at least oneyear after treatment (4). On the other hand,CBT does not appear to be more effectivethan interpersonal psychotherapy in reduc-ing associated psychopathology such asdepression, social adjustment, or anxiety (3).It is worthy of note that CBT requires anadequate training for therapists, particularlyregarding cognitive techniques such asproblem solving and cognitive restructuringabout problematic thoughts. Although CBT

seems to be effective for a majority of individuals, a number of patients does norespond adequately. According to Coker eal, non responders often have a history opsychoactive substance abuse, borderlinepersonality disorder and low self esteem (5).

Different antidepressant medications suchas imipramine (6), desipramine (7)phenelzine (8), bupropion (9), and fluoxetine(10), have been found useful in the treatmenof bulimia nervosa, reducing the frequencyof binge eating and vomiting. TheFluoxetine Bulimia Nervosa CollaborativeStudy Group, comparing fluoxetine 60mg/day with 20 mg/day treatment, showedthat the higher dose was more effective indecreasing binge eating, although a considerable proportion of patients discontinuedtreatment (11). A time-limited fluoxetine 60mg treatment is unlikely to produce a lastingreduction of binge frequency and improvements of shape and weight concerns.

CBT, when compared to psychopharmacological treatment, appears to be more effective in reducing symptomatology andimproving outcome of bulimia nervosa (12

Key words: Bulimia nervosa, cognitive-behavior therapy, grouppsychoeducation, fluoxetine,mood depression. Correspondence: Prof. Carlo M. Rotella,Sezione di Malattie delMetabolismo e Diabetologia,Unità Operativa diEndocrinologia, VialePieraccini 6, 50134 Firenze,ItalyReceived: April 10, 1997Accepted: September 1, 1997

94

V. Ricca, E. Mannucci, B. Mezzani, et al.

13). In one controlled study of bulimicpatients, combination treatment with psy-chotherapy and fluoxetine unexpectedly didnot show better results than CBT alone (14).

Psychoeducation is an important compo-nent of cognitive-behavioral therapy, aswell as other kinds of intervention for eat-ing disorders (15); in fact, some maladap-tive beliefs grow out of incorrect or insuffi-cient information (16). The efficacy of apurely psychoeducational intervention hasbeen examined in some studies (17, 18), butinformation on long-term effects is notavailable. Psycoeducational group treat-ment seems to be a viable treatment for asignificant proportion of individuals withbulimia nervosa but it appears that patientswith a severe clinical and psychopathologi-cal impairment will respond more favor-ably to an intensive individual treatment;however, psychoeducational groups couldbe more effective when combined withpharmacotherapy. Psychoeducation, whichis less time-consuming than CBT, could bea cheap first stage of treatment for eatingdisorders. Its efficacy, alone or combinedwith pharmacotherapy, needs to beexplored more extensively, to verify its ade-quacy as a low-cost alternative to CBT inbulimic outpatients.

The aim of the present study is the com-parison of CBT with combined group psy-choeducation and fluoxetine pharma-cotherapy in order to ascertain the efficacyof the two treatments in improving the spe-cific and general psychopathology ofbulimic patients.

MATERIALS AND METHODS

Patients The study was performed on a consecutiveseries of patients with bulimia nervosareferred to the Outpatient Clinic of theSection of Metabolic Diseases of theUniversity of Florence and to the EatingDisorders Unit of Casa di Cura Villa deiPini, Florence.

All bulimic patients who attended the twoClinics for the first time betweenSeptember 1st, 1994 and October 31st,1995, were invited to participate the study,provided that they met the following inclu-sion criteria:

1) age≥14 years; younger patients wereexcluded in consideration of the fact that

95

bulimia nervosa in pediatric age may needspecific therapeutic action;

2) diagnosis of bulimia nervosa accordingto the criteria of the 4th edition of theDiagnostic and Statistical Manual of MentalDisorders (DSM IV, 19); diagnosis wasobtained using an Italian translation (20) ofEating Disorder Examination 12.0D (21);

3) no concurrent medical condition thatwould preclude the use of antidepressants(e.g., heart, liver, kidney diseases);

4) no present or known history of alcoholor drug abuse;

5) no comorbity with major depression orpsychosys;

6) literacy; illiterate patients were exclud-ed because they would have not been ableto compile self-reporting questionnaires,and to perform daily self-monitoring of eat-ing behavior;

7) no use of psychotropic drugs in theprevious two months, except for low dosesof anxiolytic or hypnotic compounds.

Of the 57 bulimic outpatients observed, 3were excluded because they were affected bymajor depression, 2 for previous history ofdrug abuse, and 1 for current alcohol abuse.All the patients (or their legal tutors, forthose aged <18 years) gave their informedconsent prior to being enrolled in the study.

AssessmentBefore the beginning of treatment and 6months after the end of the treatment,patients were studied using the followinginstruments:

– Structured Clinical Interview for DSMIII R (SCID), to investigate psychiatriccomorbidity (22).

– Eating Disorder Examination (EDE12.0D) (21) was used in its validated Italiantranslation (20); this interview confirms thediagnosis of bulimia nervosa according toDSM IV criteria, and provides a dimension-al measure of eating disorder psy-chopathology.

– Beck rating scale for depression (BDI,23) and Spielberg’s State-Trait AnxietyInventory (STAI), for a further characteri-zation of the psychopathological features ofthe patients (24).

TreatmentTreatment was provided on an outpatientbasis. Patients enrolled in the study beforeMarch 15th, 1995 (Group A) were treatedwith a individual CBT program by one of

CBT vs. psychoeducation and fluoxetine in bulimia

96

the authors (V.R.), while those enrolledafterwards (Group B) were treated with acombination of fluoxetine and group psy-choeducation provided by another one ofthe authors (B.M.). The characteristics ofGroups A and B are summarized in Table 1.CBT, as described by Fairburn, Marcus andWilson (25) is articulated in 19 weekly, 45min-sessions, and it applies combinationsof behavioral and cognitive techniquesaddressed to changing maladaptive eatingbehavior and self-esteem.

Patients enrolled after March 15th (GroupB) were treated with fluoxetine (60 mg/day)and with group psychoeducational interven-tion. Two groups of patients were formed,with 12 and 13 patients respectively; the ses-sions of group 2 began two months aftergroup 1. For each group, 90 min-sessionswere held once a week for two months(phase 1), then twice a month for two months(phase 2), and finally once a month for twomonths (phase 3), for a total of 14 sessions.

During sessions 1-4, information on clini-cal features of eating disorders and theirmultidimensional nature, was provided:

– basic nutritional information;– dieting and its consequences;– medical complications of eating disor-

ders;– eating disorders and the sociocultural

context.Furthermore, daily self-monitoring of

eating behavior was explained and stronglyencouraged.

Sessions 5-8 were devoted to:– giving instructional information about

tecniques for stimulus control; – introducing a pattern of regular feed-

ing, starting with breakfast;– reintroducing “forbidden” foods and

monitoring emotional changes, appetite,hunger and satiety.

Session 9-12 were employed to verify,mainly through self-monitoring, changes ineating attitudes and behaviors and enhancebehavioral strategies.

Session 13-14 were used to repass themain topics of the program, and adaptingthem to the individual features of thepatients.

It must be pointed out that many therapeu-tical instruments originally designated as“psychoeducational” have been directlyimported into cognitive-behavioral treat-ments; on the other hand, stimulus-controltechniques and self-monitoring wereemployed in our treatment group. For thesereasons, the two kinds of interventions sharesome therapeutical tools. However, psychoe-ducation, when compared to CBT, puts morestress on the transmission of information,assuming that patients are often not awareof factors involved in the onset and main-tainance of eating disorders.

RESULTS

From among the patients of group A (CBT),6 (24%) discontinued the treatment, as wellas 5 (19.2%) patients of group B (psychoedu-cation plus fluoxetine). The dropouts werenot significantly different from those whocompleted the study in regard to age, BMI,education, marital status, working condition,psychiatric comorbidity, and EDE 12.0D,Stai 1-2 and BDI scores (data not shown).

In group B, none of the patients discon-tinued treatment due to side effects of flu-oxetine.

After 6 months from the beginning oftreatment, patients of group A showed a

TABLE 1Characteristics of the patients studied (mean ± SD).

Cognitive- Psycoeducational behavior therapy intervention and

fluoxetine

Number (sex) 25 (25 F) 26 (26 F)

Age (years) 23.4 ± 3.9 23.5 ± 3.8

BMI (kg/m2) 20.4 ± 1.7 20.6 ± 1.6

Duration of disease 4.0 ± 3.2 3.2 ±2.7

Marital statusunmarried 19 (76%) 22 (84.6%)married 4 (16%) 1(3.2%)separate/divorcee 2 (8%) 3 (11.5%)

Educationjunior high school 4 (16%) 2 (7.7%)high school 19 (76%) 21 (80.7%)degree 2 (8%) 3 (11.6%)

Working conditionunemployed 4 (16%) 6 (26.9%)employed 10 (40%) 10 (38.5%)student 11 (44%) 9 (34.6%)

Axis I diagnosisdysthimia 11 (57,9%) 10 (47.6%)adaptive disorder with depressed mood 1 (5,3%) 0

V. Ricca, E. Mannucci, B. Mezzani, et al.

97

significant reduction (p<0.001) of the month-ly number of episodes of binge eating (from25.0±12.9 to 6.7±3.8) and compensativebehavior (from 23.5±11.8 to 5.8±3.6). Ingroup B, a significant reduction of episodesof binge eating (from 24.8±9.1 to 8.0±4.3;p<0.001) was also observed. No significantdifference between the two groups in per-centual reduction of episodes of binge eat-

ing and compensatory behavior was found(Figure 1).

The EDE, BDI, and STAI scores arereported in Table 2. EDE total and subscalescores showed a significant reduction forpatients of Group A but not for those ofGroup B. BDI and STAI scores were signifi-cantly reduced in both groups.

DISCUSSION

Before the beginning of treatment the twogroups did not show any significant differ-ence in sociodemographic, clinical and psy-chopathological parameters. Consideringthat the assignment of individual patientsto different treatment groups was made onthe basis of an unbiased parameter (time offirst referral), the present study allows acomparison between the two treatments.

Both CBT and combined fluoxetine-psy-choeducation therapy, significantly improvedthe specific symptomatology of bulimia ner-vosa (i.e., binge eating and compensatorybehaviors). The two treatments appeared tohave a similar efficacy on those parameters.On the other hand, while CBT determined amarked improvement in EDE scores, com-bined fluoxetine-psychoeducational treat-ment did not. It may be speculated that CBTmight induce a reduction of bulimic symp-toms as a consequence of improvement ineating attitudes, which are measured by EDE.Combined fluoxetine-psychoeducationaltreatment seems to have little effect on specif-ic eating disorder psychopathology.

FIGURE 1Percentual variation (means ± SD) of the number of monthly episodesof binge eating (solid bar) and compensatory behaviors (grey bar)

after cognitive behavioral treatment (CBT) or combined psychoeduca-tional/fluoxetine treatment (PED).

TABLE 2Scores of EDE, BDI and STAI, before and after treatment (mean ± SD).

Cognitive-behavior therapy Psychoeducational intervention and fluoxetineBefore After Before After

EDE 4.9 ± 0.6 2.7 ± 0.4* 4.7 ± 0.3 4.5 ± 0.4restraint 4.4 ± 0.7 3.0 ± 0.4* 4.3 ± 0.6 4.2 ± 0.7eating concern 4.5 ± 0.5 2.5 ± 0.8* 4.3 ± 0.5 4.2 ± 0.4weight concern 5.0 ± 0.9 2.7 ± 0.4* 5.0 ± 0.7 5.0 ± 0.7shape concern 5.0 ± 0.9 2.8 ± 0.4* 5.1 ± 0.9 4.9 ± 0.8

BDI 18.6 ± 5.8 13.7 ± 3.5* 16.7 ± 4.9 9.0 ± 2.3*

STAI - 1 43.2 ± 9.2 35.7 ± 5.3* 42.0 ± 11.7 27.3 ± 2.2*

STAI - 2 43.2 ± 6.6 33.3 ± 5.4* 43.2 ± 7.5 27.6 ± 2.6*

*p<0,01 vs. before treatment

CBT vs. psychoeducation and fluoxetine in bulimia

The two treatments might exploit theireffects through different mechanisms.Although the use of antidepressant med-ication was initially prompted by theobservation of depressed mood in manybulimic patients, the “antibulimic” efficacyof fluoxetine could scarcely be related tothe effects on mood. One study reportedthat therapy with 60 mg daily of fluoxe-tine is more effective than a 20 mg. dailydose in reducing bulimic symptomatology,but not in improving mood depression(11). This observation has led to the specu-lation that mechanism of action of fluoxe-tine on bulimia nervosa could be differentfrom those involved in its antidepressantactivity.

The psychoeducational program is com-posed of several elements. Patients are edu-cated to consider bulimia nervosa as theresult of biological and sociocultural fac-tors, rather than personal weakness in fac-ing food and weight; self monitoring andbehavioral techniques increase awarenessand enhance a control over eating behav-iors, leading to an increase of self-esteemand markedly reducing helplessness. Thegroup setting provides support and under-standing, reducing shame and lonelinessand enabling the patients able to share dif-ficulties and improvements with others.

The design of the present study does notallow a differentiation between the therapeu-tical effects of fluoxetine and those of grouppsychoeducation. It is conceivable that thosetwo different treatment approches may havea sinergistic effect, but this should beassessed through investigations with a dif-ferent experimental design, (i.e., comparisonof fluoxetine, group psychoeducation, and acombination of the two).

Many studies have been performed toidentify the specific contributions of thecognitive and behavioral components ofCBT to the therapeutic effects of the proce-dure. The results are equivocal: most of thestudies suggest that both the behavioraland cognitive actions are needed to obtainthe maximal efficacy (26, 27), but othersreport that behavioral therapy alone is suf-ficient to achieve clinical benefits, and thatthe addiction of cognitive tecniques doesnot substantially improve outcome (28).

Data obtained in the present study sug-gest that specific cognitive procedures(e.g., cognitive restructuring, developingproblem solving skills), addressed to the

98

correction of cognitive distortions aboutshape, weight and self-esteem, play a keyrole in the improvement of the whole spe-cific and general bulimic symptomatology.In particular, it is worthy of note that CBThas a therapeutic effect on mood depres-sion similar to that obtained with fluoxe-tine, suggesting that the improvement ofthe core psychopathology determined bythe cognitive intervention has marked con-sequences on the mood of bulimic patients.

It is noteworthy that the amount of treat-ment time was greater with group psy-choeducation than with individual CBT(1260 min versus 855 min). CBT seemstherefore to be less time-consuming forpatients than group psychoeducation; how-ever, this could depend on the differenttherapeutical setting (individual instead ofgroup). It should also be considered that, ingroup psychoeducation, one therapistcould treat simultaneously 12-13 patients,with an average time for patient of about100 min, which is less than one eighth ofthat needed for individual CBT.

In conclusion, the combination of psy-choeducation and fluoxetine is capable ofreducing specific and general symptoma-tology of bulimia nervosa, although thecognitive distorsions are not significantlymodified. This kind of intervention, whichcould be more cost-effective than CBT,may be beneficial in a significant number ofindividuals, but follow-up studies are need-ed to evaluate its long-term outcome.

REFERENCES

1. Russell G.: Bulimia nervosa: an ominousvariant of anorexia nervosa. Psychol. Med.,9, 429-448, 1979.

2. Garner D.M., Rockert W., Davis R., GarnerM.V., Olmsted M.P., Eagle M.: Comparisonof cognitive-behavioral and supportive-expressive therapy for bulimia nervosa. Am.J. Psychiatry, 150, 37-46, 1993.

3. Fairburn C.G., Jones R., Peveler R.C., CarrS.J., Solomon R.A., O’Connor M., Burton J.,Hope R.A.: Three psychological treatmentsfor bulimia nervosa. Arch. Gen. Psychiatry,48, 463-469, 1991.

4. Fairburn C.G., Norman P.A., Welch S.L.,O’Connor M., Doll H.A., Peveler R.C.: Aprospective study of outcome in bulimianervosa and the long-term effects of threepsychological treatments. Arch. Gen.Psychiatry, 52, 304-312, 1995.

V. Ricca, E. Mannucci, B. Mezzani, et al.

99

5. Coker S., Vize C., Wade T., Cooper P.J.:Patients with bulimia nervosa who fail toengage in cognitive-behavior therapy. Int. J.Eating Disord., 13, 35-40, 1993.

6. Pope H.G. jr, Hudson J.I., Jonas J.M.,Yurgelun-Todd D.: Bulimia treated withimipramine: a placebo-controlled, double-blindstudy. Am. J. Psychiatry, 140, 554-558, 1983.

7. Hughes P.L., Wells L.A., Cunningham C.J.,Ilstrup D.M.: Treating bulimia withdesipramine: a double-blind, placebo-con-trolled study. Arch. Gen. Psychiatry, 43, 182-186, 1986.

8. Walsh B.T., Gladis M., Roose S.P., StewartJ.W., Stetner F., Glassman A.H.: Phenelzinevs placebo in 50 patients with bulimia. Arch.Gen. Psychiatry, 45, 471-475, 1988.

9. Horne R.L., Ferguson J.M., Pope H.G. Jr,Hudson J.I., Lineberry C.G., Asher J., CatoA.: Treatment of bulimia with bupropion: amulticenter controlled trial. J. Clin.Psychiatry, 49, 262-266, 1988.

10. Fichter M.M., Leibl K., Rief W., Brunner E.,Schmidt-Auberger S., Engel R.R.: Fluoxetineversus placebo: a double blind study withbulimic inpatients undergoing intensive psy-cotherapy. Pharmacopsychiatry, 24, 1-7, 1991.

11. Fluoxetine Bulimia Nervosa CollaborativeStudy Group (FBNCSG): Fluoxetine in thetreatment of bulimia nervosa. Arch. Gen.Psychiatry, 49, 139-147, 1992.

12. Leitemberg H., Rosen J. Vara L.S., DetzerM., Srebnick D.: Comparison of cognitive-behavior therapy and desipramine in thetreatment of bulimia nervosa. Behav. Res.Ther., 32, 37-45, 1993.

13. Mitchell J.E., Pyle R.L., Eckert E.D.,Hatsukami D., Pomeroy C., Zimmermann R.:A comparison study of antidepressant andstructured group psychotherapy in thetreatment of bulimia nervosa. Arch. Gen.Psychiatry, 47, 149-157, 1990.

14. Golbloom D., Olmsted M., Davis R., ShareB.: A randomized controlled trial of fluoxe-tine and individual cognitive-behavioraltherapy for women with bulimia nervosa:short-term outcome. Paper presented at the6th International Conference of EatingDisorder, New York, 1994.

15. Davis R., Olmsted M.P.: Cognitive-Behavioralgroup treatment for bulimia nervosa:Integrating psychoeducation and psychother-apy. In: Harper Guiffre H., McKenzie K.R.(Eds.), Group Psychotherapy for EatingDisorders. Washington, American PsychiatricAssociation Press, 1992, pp. 71-103.

16. Garner D.M., Rockert W., Olmsted M.P.,Johnson C.: Psychoeducational principles in

the treatment of bulimia and anorexia ner-vosa. In: Garner D.M., Garfinkel P.E. (Eds.),Handbook of Psychotherapy for AnorexiaNervosa and Bulimia. New York, GuilfordPress, 1982, pp. 513-572.

17. Connors M.E., Johonson C.L., Stuckey M.K.:Treatment of bulimia with brief psychoedu-cation group therapy. Am. J. Psychiatry,141, 1512-1516, 1984.

18. Olmsted M.P., Davis R., Rockert W., IrvineM.J., Eagle M., Garner D.M.: Efficacy of abrief group psychoeducational interventionfor bulimia nervosa. Behav. Res. Ther., 29,71-83, 1991.

19. American Psychiatry Association,Diagnostic and Statistical Manual of MentalDisorders, 4th ed., Washington DC., A.P.A.Press, 1994.

20. Mannucci E., Ricca V., Di Bernardo M.,Rotella C.M.: Studio del comportamento ali-mentare con una intervista strutturata: laEating Disorder Examination. Il Diabete, 8,127-131, 1996.

21. Faiburn C.G., Cooper Z.: The EatingDisorder Examination. In: Fairburn C.G.,Wilson G.T. (Eds.), Binge Eating: natureassessment and treatment. 12th ed, NewYork, Guilford Press, 1993, pp. 317-348.

22. Spitzer R.L., Williams J.B.W., Gibbon M.,First M.B.: Structured Clinical Interview forDSM III R (SCID). Washington, AmericanPsychiatric Association Press, 1990.

23. Beck A.T.: Beck Depression Inventory.Philadelphia, Center for Cognitive Therapy,1978.

24. Spielberg C.D., Gorsuch R.L., Lushene R.E.:Manual for the State-Trait Anxiety Inventory(self evaluation questionnaire). Palo Alto,Consulting Psychologists Press, 1970.

25. Fairburn C.G., Marcus M.D., Wilson G.T.:Cognitive-behavioral therapy for binge eat-ing and bulimia nervosa: a comprehensivetreatment manual. In: Fairburn C.G., WilsonG.T. (Eds.). Binge Eating: Nature Assessmentand Treatment. New York, Guilford Press,1993, pp. 361-396.

26. Wilson G.T., Rossiter E., Kleifield E.I.,Lindholen L.: Cognitive-behavioral treat-ment of bulimia nervosa: a controlled evalu-ation. Behav. Res. Ther., 24, 277-288, 1986.

27. Thackwray D.E., Smith M.C., Bodfish J.W.,Meyers A.W.: Comparison of behavioraland cognitive-behavioral interventions forbulimia nervosa. J. Consult. Clin. Psychol.,61, 639-645, 1993.

28. Cooper P.J., Cooper Z., Hill C.: Behavioraltreatment of bulimia nervosa. Int. J. EatingDisord., 8, 87-92, 1989.