Dysfunctional assumptions in bipolar disorder

Abstract

Background: Despite the initial encouraging outcome in developing CBT for bipolar affective disorder [Arch. Gen. Psychiatry 2002 (in press); Psychol. Med. 31 (2001) 459–467], very little is known about whether there are any differences in dysfunctional attitudes between unipolar and bipolar patients. Both the behavioural activation system theory [J. Pers. Soc. Psychol. 67 (1994) 488–498; Major Theories of Personality Disorder, Guilford Press, New York, 1996; Psychol. Bull. 117 (1995) 434–449] and the cognitive model for bipolar affective disorder [Cognitive Therapy for Bipolar Disorder: A Therapist’s Guide to Concepts, Methods and Practise, Wiley, New York, 1999] postulate high goal striving as a risk factor for bipolar disorder. However, the existing subscales in the dysfunctional attitude scale (DAS) were derived from patients and relatives of patients suffering from unipolar depression, patients with a mixed psychiatric diagnosis or normal controls. None of the existing subscales reflects high goal striving beliefs. Using a sample of bipolar patients may yield different factors. Method: A total of 143 bipolar 1 patients filled in the short version of DAS 24. Principal component analysis was carried out to derive factors. The scores of these factors were compared with those of 109 unipolar patients to investigate if these factors distinguish bipolar patients from unipolar patients. Results: Three factors were derived: factor 1 ‘Goal-attainment’ accounted for 25.0% of the total variance. Factor 2 ‘Dependency’ accounted for 11.0% of the total variance. Factor 3 ‘Achievement’ accounted for 8.2% of the total variance. However, factor 1 appeared to consist of items that made a coherent theoretical construct. No significant differences were found when the validation sample was compared with 109 patients suffering from unipolar depression in any of the three factors. When subjects who were likely to be in a major depressive episode were excluded, the scores of bipolar patients (n=49) were significantly higher than euthymic unipolar patients (n=25) in factor 1 ‘Goal attainment’. Goal-attainment also correlated with the number of past hospitalisations due to manic episodes and to bipolar episodes as a whole. Conclusion: The Goal-attainment subscale captures the risky attitudes described by the behavioural activation system theory and the cognitive model for bipolar affective disorder. It is postulated that these beliefs may interact with the illness and predispose bipolar patients to have a more severe course of the illness.

Introduction

There has been recent interest in developing cognitive therapy for bipolar affective disorder as a combination treatment with medication (e.g. Lam et al., 2002, Scott et al., 2001) with encouraging results. However the issue of whether there are unique personality traits or attitudes, which can influence the onset or the course of bipolar affective disorder is important for the development of psychotherapy specific for the illness. Two different theoretical models have potential contributions in this area: the cognitive theory for mood disorder (Beck, 1967) and the Behavioural Activation System (BAS: Depue et al., 1994, Depue, 1996, Gray, 1994, Johnson and Roberts, 1995).

The BAS regulates a broad band of goal-seeking behaviour and characteristics in support of appetitive motivation including positive affectivity, energy, sleep and attention. Mania is hypothesised as the outcome of dysregulation in the BAS. The model suggests that positive affect, extraversion and achievement striving should be linked to increased levels of mania across time among bipolar individuals. Within the cognitive model framework, Lam et al. (1999) postulated that the extreme goal attainment attitudes in bipolar affective disorder might be the dysfunctional beliefs that lead to extreme striving behaviour and irregular daily routine. Hence both theoretical models seem to agree that extreme striving or extreme goal attainment attitudes are important and may influence the course of the illness in bipolar affective disorder.

It was found that euthymic bipolar patients showed significantly higher scores on the Dysfunctional Attitudes Scale (DAS) as well as the ‘Need for Approval’ and ‘Perfectionism’ subscales than healthy control (Scott et al., 2000). To understand the unique nature of the illness, one needs to understand the factors that distinguish bipolar patients from other mood disorders. Studies that compare bipolar with unipolar patients produced mixed results. Bipolar patients were significantly different from unipolar patients on the Openness to Experience subscale of the Five-Factor Model of Personality (Bagby et al., 1996). However, two studies have failed to identify differences in personality traits of bipolar patients when compared with unipolar patients (Clayton et al., 1994, Akiskal et al., 1995).

There is also evidence that both remitted unipolar and bipolar females were more critical than female controls (Rosenfarb et al., 1998). However, both depressed unipolar and bipolar women exhibited significantly higher scores on dependency needs than normal controls. These differences disappeared when the depression had remitted.

None of the current factor analysis studies produced subscales which reflect the high goal attainment attitudes hypothesised to be important for bipolar affective disorder. However, bipolar patients were not represented in these studies. For example, the conceptually similar Perfectionism (Weissman and Beck, 1978), Achievement (Power et al., 1994) and Performance Evaluation subscales (Cane et al., 1986) of the DAS consist of items, which reflect over-generalisation and excessive blame (e.g. If I fail at my work, then I am a failure as a person). Similarly the Need for Social Approval (Weissman and Beck, 1978), the Dependency subscale (Power et al., 1994) and the Approval by Others subscale (Cane et al., 1986) of the DAS reflect the need to be validated by another person (e.g. If others dislike you, you cannot be happy). The Self-control subscale derived from the Power et al. (1994) study contains high goal attainment attitudes (e.g. I ought to be able to solve my problems quickly and without a great deal of effort). However, it also consists of a mix of items that do not fit into the concept of extreme goal-attainment (e.g. “It is possible for a person to be scolded and not get upset” or “Whenever I take a chance or risk I am only looking for trouble”).

No studies have found any differences in dysfunctional attitudes between unipolar and bipolar patients. The DAS was found to correlate significantly with the BDI, r=0.65 (Weissman and Beck, 1978). Hence similar dysfunctional attitudes may exist between bipolar and unipolar patients when they are in a depressive episode. However, euthymic bipolar patients may exhibit certain characteristic attitudes that are distinct from remitted unipolar depressives.

This study set Out to:

• Investigate what factors can be derived from the DAS using a pure bipolar 1 sample

• Test whether euthymic bipolar patients are different from euthymic unipolar patients in these factors derived from a bipolar population

• Investigate whether any risky factors identified in the DAS in this study correlate with past bipolar episodes