Asian Journal of Psychiatry 4 (2011) 125–128

Contents lists available at ScienceDirect

Asian Journal of Psychiatry

journal homepage: www.e lsev ier .com/ locate /a jp

Predictors of Group Cognitive Behaviour Therapy outcomes for thetreatment of depression in Malaysia

Firdaus Mukhtar *

Department of Psychiatry, Faculty Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia

A R T I C L E I N F O

Article history:

Received 2 June 2010

Received in revised form 14 March 2011

Accepted 9 April 2011

Keywords:

Group CBT

Depression

Predictors

Treatment outcome

Malaysia

A B S T R A C T

The aim of this study was to identify predictors of response to treatment for depression in Malaysia,

using demographic and cognitive predictors. 113 patients, that were diagnosed with depression, were

randomly assigned to the Treatment-As-Usual (TAU) (n = 55), or TAU plus eight sessions of Group

Cognitive Behaviour Therapy (TAU + GCBT; n = 58). Pre-treatment using the Beck Hopelessness Scale

(BHS), the Automatic Thoughts Questionnaire-Malay (ATQ-Malay), the Dysfunctional Attitude Scale-

Malay (DAS-Malay), a quality of life scale, and demographic characteristics, were used in a series of

multiple regression models, as potential predictors of the Beck Depression Inventory-Malay (BDI-Malay)

post-assessment scores. Regression results revealed that age, the quality of life scale, and all three

cognitive measures were significant predictors of outcomes in the Group Cognitive Behaviour Therapy

(GCBT) group, showing that Beck’s cognitive model for depression could be applied in Malaysia.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

Our previous study demonstrated that scores on the variousmeasures of maladaptive mood and cognition showed significantdecreases after patients with depression were exposed to GroupCognitive Behaviour Therapy (GCBT) (Mukhtar et al., 2006).Specifically, the treatment group (i.e., those who received TreatmentAs Usual (TAU) plus eight sessions of GCBT) showed significantlydecreased scores on the Beck depression scale and cognitivemeasures from pre-treatment to post-treatment, when comparedto the TAU only group. These findings provide evidence for theeffectiveness of GCBT, in terms of modifying maladaptive schemas,dysfunctional attitudes, and symptoms of depression in Malaypatients. However, the factors associated with the successful use ofGCBT in Eastern countries remains unclear. Subsequently, it is vitalto understand the role of pre-treatment variables in predictingtreatment outcomes when using GCBT to treat depression.

In Western literature, a wealth of empirical studies hasidentified a range of factors that tend to be associated withpositive Cognitive Behaviour Therapy (CBT) treatment outcomes.These include pre-treatment scores on the Beck DepressionInventory (Hamilton and Dobson, 2002), demographic values(e.g., marital status (Jarrett et al., 1991)), historical features of theillness (Hamilton and Dobson, 2002), specific intra-personalfactors (e.g., automatic thoughts and dysfunctional attitudes

* Tel.: +60 3 8947 2543; fax: +60 3 8941 4629.

E-mail addresses: drfirdaus@medic.upm.edu.my, fir_cp@yahoo.com.

1876-2018/$ – see front matter � 2011 Elsevier B.V. All rights reserved.

doi:10.1016/j.ajp.2011.04.002

(Oei and Shuttlewood, 1996)), and non-specific external factors(e.g., the nature of the therapeutic alliance or group processes (Oeiand Browne, 2006)).

Two other potential predictor variables, affecting the effec-tiveness of CBT, that were not discussed in previous studies, are asense of hopelessness and quality of life. Hopelessness is a corefactor in the cognitive-behavioural explanation for the develop-ment and persistence of depression, but is often overlooked byresearchers (Henkel et al., 2002; Westra et al., 2002). In terms of‘quality of life’ as a predictor variable, numerous studies havefound evidence that there is a significant relationship betweenthis variable and depression (McAlinden and Oei, 2006; Ola et al.,2006; Ong et al., 2006). One study, by Gore-Felton et al. (2006),identified it as the most important predictor of depression amongpatients with a major depressive disorder. Therefore, besidesdemographic and cognitive variables, quality of life was alsoinvestigated in this study, as a potential predictor variable fortreatment outcome.

Although a number of predictor variables for CBT treatment ofdepression have been identified for Western populations, it isunknown what predictor variables are useful for Easternpopulations; particularly for patients suffering from depressionin Malaysia. It is important to discover these predictor variables,as they can assist in the management of depression. This presentstudy was designed to investigate potential predictors of CBTtreatment outcomes, for the Malaysian population. As there wereno previous relevant Eastern studies, the hypotheses were guidedby existing Western literature. Consequently, it was hypothesisedthat negative automatic thoughts, dysfunctional attitudes,

Table 1Mean scores for the BDI at pre and post-treatment for both groups.

Time GCBT (mean/SD) Waitlist (mean/SD)

Pre-treatment 34.24 (5.47) 34.53 (5.022)

Post-treatment 4.33 (6.12) 36.64 (7.163)

F. Mukhtar / Asian Journal of Psychiatry 4 (2011) 125–128126

hopelessness, and quality of life, would significantly predict BeckDepression Inventory-Malay (BDI)-Malay scores during post-treatment.

2. Methods

2.1. Participants

One hundred and thirteen patients (51 males) with majordepression were randomly divided into group one (n = 58) whoreceived TAU + CBT and group two (n = 55) who were in TAU only.The patients were 20–59 years old, with an average age of 40.5years. In terms of level of education, nine patients completedprimary school, 86 patients (74.1%) completed secondary school,14 patients (12.1%) completed certificate/diploma courses, andseven patients (6.0%) completed undergraduate studies. Themajority of patients (90.5%) were taking anti-depressant medica-tion during the course of therapy.

2.2. Materials

The (BDI-Malay; Mukhtar and Oei, 2008), the AutomaticThoughts Questionnaire-Malay (ATQ-Malay; Oei and Mukhtar,2008), and the Dysfunctional Attitude Scale-Malay (DAS-Malay;Mukhtar and Oei, 2010) description and psychometric properties,have been described in detail in our earlier study.

The Beck Hopelessness Scale-Malay (BHS-Malay) is a translatedversion of the original BHS (Beck and Steer, 1988) with a 20-itemscale for measuring negative attitudes about the future. The scale’smanual, reports KR-20 coefficients (measures of the scale’s internalconsistency) ranging from 0.82 to 0.93 and a test–retest reliabilityof 0.69 (Beck and Steer, 1988). Three subscales derived from theBHS, are feelings about the future, loss of motivation, and futureexpectations (Beck et al., 1974).

The WHO Quality of Life (WHOQOL) Brief version in BahasaMalaysia (WHOQOL Malay) (Hasanah et al., 2003) has 26 items. Ithas been validated by a Malaysian population, indicating gooddiscriminant validity, construct validity, internal consistency(0.64–0.80), and test–retest-reliability (0.49–0.88). The scale is avalid and reliable assessment of quality of life; especially for thosewith illnesses. The four domains that have been derived from thisquality of life scale are physical health, psychological, social, andenvironmental.

2.3. Procedure

A total of 203 patients completed the initial intake assessment,90 of which were either excluded from the study according to thecriteria, or they declined to participate. The remaining 113 patientswere randomly allocated into two groups following the initialintake assessment. Initial diagnosis was made by psychiatrists,with appropriate training on the diagnostic interview, whichfollowed the Diagnostic and Statistical Manual of Mental Disorders– IV (DSM-IV), and further verification of diagnoses was done bythe first author, using the Structured Clinical Interview forDiagnostic and Statistical Manual of Mental Disorders (SCID).Participants were recruited from psychiatry clinics over a period of1 year. The inclusion criteria required that participating patientsshould be aged between 20 and 60 years old, Malay literate, neverbeen treated with CBT, and meet the DSM-IV criteria, for majordepressive disorder (either single episode or recurrent) andDysthymia. Exclusion criteria included a DSM-IV determineddiagnosis of bipolar mood disorder or other major psychiatricdisorder, organic brain disorder, abuse of drugs and/or alcohol, orhaving any other major physical illness. The study was conductedat psychiatric clinics, in both East and West Malaysia, to gain a

more representative sample of the whole population. The scaleswere distributed and collected by research assistants with anundergraduate degree in psychology and who had been trained touse the instruments. The raters were blind about the treatmentallocation throughout the study. Data was collected at baseline andafter treatment was completed.

The GCBT was conducted by the first author (a qualified clinicalpsychologist, who trained at an Australian university), where CBTwas deemed to be the most applicable mode of intervention inpractice. The GCBT treatment manual for depression is based onOei’s (2002) group CBT for depression workbook; Oei and Browne(2006), which was later translated and adapted into Malay languageby the first author. Both authors of these studies are qualified clinicalpsychologists, with extensive GCBT experience. The GCBT programconsisted of two sessions per week, for 4 weeks; where each sessionlasted for approximately 3 hours. The programs format includedmini-lectures, group exercises, guided readings, and homeworktasks (see Oei (2002) for the full program). In this study the criterionvariable was the post score of the BDI-Malay.

3. Results

3.1. Assumption testing

Before the main analyses, the data was checked for missingdata, the presence of outliers, multicollinearity and singularity,normality, linearity, and homoscedasticity. No multicollinear orsingular relationships were detected. An absence of univariateoutliers was evident from a series of box plots; screening ofMahalanobis distances revealed no significant multivariate out-liers. Table 1 describes the significant pattern of the treatmentoutcome between the two groups.

Pairwise comparison showed a significant pattern between preand post treatment of the BDI score (df = 4, 2.85; F = 130.36).Meanwhile, Table 2 shows the results of the correlational analysesbetween predictor variables and respective criterion variables(BDI-Malay post).

3.2. Predicting BDI-Malay post-treatment scores after controlling for

pre-treatment BDI-Malay

The predictor variables were demographic variables (i.e., age,gender, and level of education) and the pre-treatment total scores ofthe ATQ-Malay, the DAS-Malay, the BHS, and the WHOQOL-BREF. Allpre-treatment scores were mean-centred before the analyses wereconducted. In Step 1, demographic variables (i.e., age, gender, andeducation) and pre-treatment scores of the BDI-Malay were enteredinto the first regression model, and then Hierarchical MultipleRegression (HMR) was conducted. This was followed by pre-treatment scores on the ATQ-Malay and DAS-Malay in Step 2, andfinally, the pre-treatment scores on the BHS and WHOQOL-BREFwere entered. The results of the HMR are shown in Table 3, whichrevealed that at Step 1, age was significant (p < 0.01) and at Step 2,both the ATQ-Malay (p < 0.05) and the DAS-Malay (p < 0.05) weresignificant, with an R2 of 2%. At Step 3, both the BHS (p < 0.01) andthe WHOQOL-BREF (p < 01) were found to significantly predict BDI-Malay post-treatment scores, with an additional R2 change of 21%.All independent variables together explain 40% of the total variancein the BDI-Malay post-treatment scores.

Table 2Pearson correlations between pre-cognitive and demographic variables and the BDI-Malay post-treatment scores.

1 2 3 4 5 6 7 8 9

1. Age 0.243** �0.324** 0.052 �0.042 0.145 0.129 �0.191* 0.343**

2. Gender �0.114 0.085 �0.224* �0.146 0.186* �0.164 0.239*

3. Education �0.005 �0.074 �0.106 �0.106 0.070 �0.123

4. Pre BDI-Malay 0.158 0.112 0.012 �0.064 0.106

5. Pre ATQ-Malay 0.320** �0.135 0.061 �0.162

6. Pre DAS-Malay 0.133 �0.055 0.151

7. Pre BHS �0.429** 0.460**

8. Pre WHOQOL-BREF �0.486**

9. Post BDI Malay

ATQ = Automatic Thoughts Questionnaire, DAS = Dysfunctional Attitude Scale, BHS = Beck Hopelessness Scale, WHOQOL = The WHO Quality of Life and BDI = Beck Depression

Inventory.* p<0.05.** p<0.01.

Table 3Hierarchical multiple regression analyses incorporating predictor variables (demographic and pre-treatment total score) with the criterion variables as the BDI-Malay post-

treatment.

Criterion Step and predictor variable b SE(b) t p R2 DR2 F

BDI-Malay post-treatment score Pre-treatment total scores

Step 1

Age 0.463 0.149 3.10 0.002* 0.14 0.14 4.71

Gender 5.27 3.02 �1.74 0.085

Education �0.190 2.38 �0.080 0.937

Pre-treatment BDI-Malay 0.214 0.268 0.800

Step 2

Age 0.408 0.149 2.74 0.007 0.19 0.04 4.29

Gender �4.97 3.09 �1.61 0.110

Education �0.367 2.35 �0.156 0.877

ATQ-Malay minus mean �0.198 0.099 �2.00 0.047*

DAS-Malay minus mean 0.147 0.075 1.96 0.052*

Step 3

Age 0.326 0.131 2.48 0.014 40 0.20 8.69

Gender �2.40 2.73 �0.879 0.381

Education 0.128 2.05 0.062 0.951

Pre-treatment BDI-Malay 0.206 0.234 0.881 0.380

ATQ-Malay minus mean �0.137 0.087 �1.56 0.120

DAS-Malay minus mean 0.091 0.066 1.38 0.170

BHS minus mean 1.93 0.660 2.93 0.004*

WHOQOL-BREF minus mean �1.26 0.352 �3.59 0.001*

F. Mukhtar / Asian Journal of Psychiatry 4 (2011) 125–128 127

3.3. Predicting rate of change scores (between pre and

post-treatment BDI-Malay)

In order to discover to what extent the pre-treatment variablespredicted the rate of improvement, the BDI changes score (postminus pre-treatment scores) was used. The regression analysesresults are summarized in Table 3. As can be seen, the resultsrevealed that the demographic variables and the pre-treatmentindependent variables together explained 36% of the total variancein the change scores, which was highly significant (F (7,105) = 8.53,p < 0.0001). Individually, the pre-treatment scores on the ATQ-

Table 4Multiple regression analyses, predicting change scores, from demographic and predicto

Criterion Predictor b SE(b

Change score of BDI-Malay Age �0.317 0.13

Gender 1.28 2.84

Education 0.133 2.15

ATQ-Malay 0.182 0.09

DAS-Malay �0.075 0.06

BHS �1.98 0.69

WHOQOL-BREF 1.20 0.36

Malay (p < 0.05), the BHS (p < 0.01), the WHOQOL-BREF(p < 0.01), and age (p < 05) contributed independently to theprediction of the rate of improvement (see Table 4).

4. Discussion

These findings demonstrated that age, total pre-treatmentscores on the ATQ-Malay, the DAS-Malay, the BHS, and theWHOQOL-BREF significantly predicted the BDI-Malay post-treat-ment scores as well as the rate of improvement, i.e., change scores.

r variables.

) t p R2 AdjR2 F

7 �2.31 0.023* 0.36 0.32 8.53

0.452 0.652

0.062 0.951

0 2.01 0.046*

9 �1.08 0.281

2 �2.86 0.005*

9 3.26 0.002*

F. Mukhtar / Asian Journal of Psychiatry 4 (2011) 125–128128

The findings from this study have a series of importantimplications for both clinical and cultural arenas.

First, in terms of clinical implications, the group format therapyallowed younger patients to benefit from the treatment as theylearned cognitive and behavioural skills by encouraging, correct-ing, and motivating each other. This is particularly importantconsidering that group members tended to have negativethoughts, beliefs, or maladaptive behaviour regarding theirproblems. Moreover, the therapeutic alliance or collaboration thatformed between the therapists and group members whenpracticing the cognitive skills was more proactive, structured,problem-oriented, and focused; hence, assisting patients toidentify and modify their negative thoughts by sharing theirown real life experiences.

Second, this study found that all cognitive measures (ATQ-Malay, the DAS-Malay, and the BHS) and quality of life weresignificant predictors of treatment outcome. This finding suggeststhat mood and cognition are closely associated with depressionand that GCBT techniques seem applicable in a Malaysian context.This is important information and can be used to guide therapistsor clinicians when planning treatment for patients with depressionwho come from different backgrounds.

It is now well accepted that Beck’s cognitive model ofdepression is used for the CBT treatment of depression in theWestern world even though the validity of Beck’s cognitive modelin the East is still unclear. Our findings have provided empiricalevidence that Beck’s cognitive model is supported for thetreatment of depression in Malaysia. However, replication isneeded before the acceptance of Beck’s cognitive model for use inan Eastern culture can be made.

Future research must attempt to further isolate the factors thatcontribute to improving treatment outcomes in order to ensurethat the optimal treatment options for individual patients can begenerated. In summary, the results of this study provide supportfor specific factors or predictor variables that are important forGCBT treatment outcomes in Malaysia.

Acknowledgements

Special thanks go to all participants in this study, the Ministry ofHealth Malaysia, and the University of Queensland, Australia, fortheir significant support and contribution.

References

Beck, A.T., Steer, R.A., 1988. Beck Hopelessness Scale. Psychological Corp., SanAntiano, TX.

Beck, A.T., Weissman, A., Lester, D., Trexler, D., 1974. The measurement of pessi-mism: the hopelessness scale. Journal of Consulting and Clinical Psychology 42,861–865.

Gore-Felton, C., Koopman, C., Spiegel, D., Vosvick, M., Brondino, M., Winningham, A.,2006. Effects of quality of life and coping with depression among adults livingwith HIV/AIDS. Journal of Health Psychology 11, 711–729.

Hamilton, K.E., Dobson, K.S., 2002. Cognitive therapy of depression: pre-treatmentpatient predictors of outcome. Clinical Psychology Review 22, 875–893.

Hasanah, C.I., Naing, L., Rahman, A.R.A., 2003. World Health Organization quality oflife assessment: brief version in Bahasa Malaysia. Medical Journal of Malaysia58, 79–88.

Henkel, V., Bussfeld, P., Moller, H.-J., Hegerl, U., 2002. Cognitive-behavioural theo-ries of helplessness/hopelessness: valid models of depression? EuropeanArchive of Psychiatry and Clinical Neuroscience 252, 240–249.

Jarrett, R.B., Eaves, G.G., Brannerman, B.D., Rush, A.J., 1991. Clinical, cognitive, anddemographic predictors of response to cognitive therapy for depression: apreliminary report. Psychiatry Research 37, 245–260.

McAlinden, N.M., Oei, T.P.S., 2006. Validation of the quality of life inventory forpatients with anxiety and depression. Comprehensive Psychiatry 47, 307–314.

Mukhtar, F., Oei, T.P.S., 2008. Exploratory and confirmatory factor validation andpsychometric properties of Beck Depression Inventory for Malays (BDI-Malay)in Malaysia. Malaysian Journal of Psychiatry 17, 1.

Mukhtar, F., Oei, T.P.S., 2010. Exploratory and confirmatory factor validation of theDysfunctional Attitude Scale for Malays (DAS-Malay) in Malaysia. Asian Journalof Psychiatry 3 (3), 145–151.

Mukhtar, F., Oei, T.P.S., Yaacob, M.J., 2006. Effectiveness of group cognitive behav-iour therapy for mood disorders for Malays in Malaysia. In: Paper Presented at1st CBT Asia Conference, Hong Kong, May.

Oei, T.P.S., 2002. CBT Mood Disorder Day Treatment Program Work Book. ToowongPrivate Hospital, Toowong, Queensland.

Oei, T.P.S., Browne, A., 2006. Components of group processes: have they con-tributed to the outcome of mood and anxiety disorder patients in a groupcognitive-behaviour therapy program? American Journal of Psychotherapy60, 53–70.

Oei, T.P.S., Mukhtar, F., 2008. Exploratory and confirmatory factor analysis andpsychometric properties of Automatic Thoughts Questionnaire for Malays inMalaysia. Hong Kong Journal of Psychiatry 18, 92–101.

Oei, T.P.S., Shuttlewood, G.J., 1996. Specific and nonspecific factors in psychothera-py: a case of cognitive therapy for depression. Clinical Psychology Review 16,83–103.

Ola, B.A., Adewuya, A.O., Ajayi, O.E., Akintomide, A.O., Oginni, O.O., Olagun, Y.A.,2006. Relationship between depression and quality of life in Nigerian out-patients with heart failure. Journal of Psychosomatic Research 61, 797–800.

Ong, L., Irvine, J., Nolan, R., Cribbie, R., Harris, L., Newman, D., Mangat, I., Dorian, P.,2006. Gender differences and quality of life in atrial fibrillation: the mediatingrole of depression. Journal of Psychosomatic Research 61, 769–774.

Westra, H.A., Dozois, D.J.A., Boardman, C., 2002. Predictors of treatment change andengagement in cognitive-behavioural therapy for depression. Journal of Cogni-tive Psychotherapy 16 (2), 227–241.