Behavioral Activation System (BAS) differences in bipolar I and II disorder

Abstract

Background

A growing body of evidence supports the Behavioral Activation System (BAS) dyresgulation model of bipolar disorder, however its application to bipolar II disorder is limited. The current study examines its potential relevance to bipolar I and II disorders. We specifically sought to determine whether bipolar sub-types would differ in terms of BAS sensitivity, and examined for differential prospective relationships between BAS sensitivity and bipolar I and II symptom expression.

Method

Participants were recruited from the Sydney-based Black Dog Institute. Diagnostic groups were derived on the basis of agreement between clinician and DSM-IV diagnoses from structured interviews. Baseline measures of BAS sensitivity, mood symptoms and anxiety were completed. Self-rated mood was assessed over a 6-month period. Clinician-rated mood status was re-assessed at follow-up to determine the predictive utility of BAS scores.

Results

The sample comprised 151 bipolar participants (69 bipolar I, 82 bipolar II). BAS-Drive and Reward Responsiveness scores were significantly higher in bipolar I disorder participants. BAS sub-scale scores were uniquely positively associated with mood variability in bipolar I and II disorder. BAS-Drive and Reward Responsiveness scores were positively associated with bipolar I hypo(mania), and with the former also positively associated with bipolar II depression. BAS scores did not predict bipolar I or II mood episode status at 6-month follow-up.

Limitations

BAS sensitivity was self-reported; inability to establish independence of BAS scores from residual symptoms; lack of controlling for medication effects; inability to determine the influence of life events; length of follow-up period may have not been sufficient to evaluate the predictive utility of BAS sensitivity for mood episodes or detect course of illness differences across bipolar sub-types.

Conclusions

Differences in BAS sensitivity and associations with mood variability were quantified in bipolar I and II disorder, suggesting the need for tailored treatments for these separate conditions. Further investigation of the role of the BAS in bipolar sub-types is warranted.

Introduction

A complex interplay between psychological factors, biological processes, and the environment is likely to contribute to bipolar illness onset, course and expression. Psychological models provide the context from which to understand these associations, however such models of bipolar disorder are still in their infancy. Refining and understanding psychological models ultimately provides the foundation from which to develop more effective psychological treatments for bipolar disorder.

A prominent psychological model of bipolar disorder is the Behavioral Activation (or Approach) System (BAS) dyresgulation model (Depue et al., 1987, Depue and Iacono, 1989, Fowles, 1988, Fowles, 1993, Gray, 1991, Urosevic et al., 2008), variably also termed the BAS ‘sensitivity’ model. This model expands on cognitive-behavioral conceptualisations of bipolar disorder, but also attempts to integrate biological vulnerability associated with the condition. Biological vulnerability is hypothesized to result from dysregulation of the BAS—a neurobiological motivational system operating via dopaminergic pathways that regulates goal-directed approach to potential reward. Those with bipolar disorder are thought to have a deficit in the serotonergic regulation of such pathways, rendering them more reactive to cues of reward (Depue and Zald, 1993). The BAS is thought to influence positive affect, energy and attention—with hypo(mania) reflecting extreme sensitivity of this system. BAS functions include a range of cognitive and affective processes associated with goal-directed behavior (Johnson et al., 2012). According to the model, life events involving goal attainment serve as cues for behavior, resulting in increased BAS activity. Once the BAS is engaged, a range of responses are produced including motor activity, increased energy and confidence, and heightened interest and pleasure in rewards [i.e. symptoms of hypo(mania)]. Sensitivity levels of the BAS (sometimes referred to as reward sensitivity) are hypothesized to predict the level of output—so, those with heightened sensitivity will manifest symptoms to a greater degree. The most commonly used measure of BAS sensitivity is Carver and White's (1994) BAS scale—a self-report measure based on how BAS sensitivity would be manifested in terms of overt behaviors and subjective experience. As reviewed by Johnson et al. (2012), BAS sub-scales are associated with affective and behavioral measures of BAS sensitivity; a propensity for setting more approach goals and placing higher value on such goals (Jones et al., 2007, Alloy et al., 2009); predict high arousal positive affect during goal pursuit (Heponiemi et al., 2003); and predict responses after receipt of a reward such as affect, confidence levels and neural activity (De Pascalis et al., 2010, Germans and Kring, 2000, Meyer et al., 2010, Beaver et al., 2006, Van den Berg et al., 2011).

There is a growing body of evidence supporting the BAS dysregulation model. BAS sensitivity is elevated in those with a bipolar condition, and is associated with increased severity of mania in bipolar disorder after controlling for baseline mania levels (Johnson et al., 2012). BAS sensitivity is also related to bipolar illness course. In bipolar I disorder samples, elevated BAS scores prospectively predicted increases in manic symptoms after controlling for baseline symptoms (Meyer et al., 2001). Using a different measure of BAS sensitivity (the Sensitivity to Punishment and Sensitivity to Reward Questionnaire), Salavert et al. (2007) reported that increased BAS sensitivity was related to greater risk of manic episodes relative to depressive episodes at 18-month follow-up. Two studies of non-clinical samples of undergraduate students diagnosed with bipolar spectrum disorders reported that BAS scores predicted shorter time to hypo(manic) episodes after controlling for baseline symptoms (Alloy et al., 2008, Alloy et al., 2009). However, these studies combined bipolar II and cyclothymia conditions in their analyses, and as noted by the authors, the non-clinical sample was unlikely to be representative of community or clinical samples of individuals with bipolar disorder.

The role of the BAS in bipolar depression is less clear. Depue and Iacono (1989) hypothesized that hypo(manic) and depressive symptoms reflect opposite extremes of behavioral components of the BAS—excessively high BAS activity is linked to mania, while excessively low BAS activity is linked to depression. In contrast to hypo(mania), less research has focused on the role of the BAS in bipolar depression. As overviewed by Johnson et al. (2012), inconsistent correlations have been reported between BAS scores and depressive symptoms. The predictive utility of the BAS for bipolar depression has also produced inconsistent results. Alloy et al. (2008) reported a non-significant trend for BAS-Reward Responsiveness to predict a greater likelihood and shorter time to onset of depressive episodes after controlling for baseline symptoms—but their later study (Alloy et al., 2009) did not replicate these results. Other studies in bipolar I disorder samples (e.g. Meyer et al., 2001, Salavert et al., 2007) reported that BAS scales were unable to predict the onset of depression episodes over time, with the latter study reporting that those with a non-significant trend towards lower BAS sensitivity than controls experienced a depressive relapse. However, Meyer et al. (2001) reported a BAS and depression symptom correlation of 0.35, with BAS prospectively predicting depressive symptoms. They concluded that the trait characteristic of low BAS may predispose such individuals to experience less positive reinforcement from the environment, resulting in depressive episodes over time.

Whilst there is growing support for the role of the BAS in bipolar disorder, studies to date have focused on individuals at risk of developing bipolar disorder, or those with a bipolar I condition—thus little is known about BAS characteristics in bipolar II disorder. In a recent study conducted by Alloy et al. (2012), predictors of conversion to bipolar I or II diagnoses were examined in a sample of college-age individuals with ‘soft’ bipolar conditions. After controlling for family history of bipolar spectrum disorders, high BAS sensitivity (and in particular, the sub-scale BAS Fun-Seeking) predicted an increased likelihood of progression to bipolar II but not bipolar I disorder. When removing family history as a covariate, the converse results occurred. Such results were interpreted as providing partial support for the hypothesis that BAS sensitivity serves as a risk factor for a worsening course of bipolar spectrum disorder. More recently, cognitive style differences in a clinical sample of bipolar I and II disorder patients were examined, quantifying subtle differences between groups on one BAS sub-scale (Drive), and concluded that those with bipolar I disorder may show greater determination in their pursuit of rewards, leading to more severe mood dysregulation [i.e. mania, rather than hypo(mania)] than that observed in bipolar II disorder (Fletcher et al., 2013).

We have previously noted that whilst bipolar II disorder represents the most common phenotype of bipolar illness in the community, this group continues to be under-studied (Parker, 2008, Fletcher, 2012a, Fletcher et al., 2012b, Fletcher et al., 2013a, Fletcher, in press-a, submitted for publication, Fletcher, in press). Furthermore, treatments (both psychological and drug-based) for bipolar II disorder tend to be extrapolated from treatments used to manage bipolar I disorder. While psychological treatments most likely benefit both sub-types, it does not follow that optimal treatments for bipolar I disorder will necessarily be the same as those for bipolar II disorder (Yatham et al., 2013). Subtle characteristics may influence whether certain treatments are likely to be effective for certain bipolar sub-groups (Richardson, 2010).

Two key illness course features differentiate bipolar II from bipolar I conditions—the more pronounced depressive predominance and chronic course in bipolar II disorder (Vieta et al., 1997, Judd et al., 2005, Mantere et al., 2008, Baek et al., 2011); and the experience of hypomania rather than mania—yet the etiology of these differences is not well understood. We have conducted a series of studies examining psychological features of bipolar II disorder (Fletcher, 2012a, Fletcher et al., 2012b, Fletcher et al., 2013a, Fletcher, in press-a, submitted for publication, Fletcher, in press in an attempt to understand whether subtle differences between the bipolar sub-types in terms of these features may partially explain the differing illness course characteristics. The current study extends this line of research, focusing on the role of the BAS in bipolar I and II disorder. We sought to determine whether bipolar I and II disorder would differ on BAS sensitivity, and whether there would be differential prospective relationships between bipolar I and II symptom expression. In line with the two aforementioned illness course differences between bipolar I and II disorder, we were specifically interested in whether differences in BAS sensitivity would be more relevant for bipolar I hypo(mania) than bipolar II hypomania, and that such differences may be uniquely associated with bipolar II depression. We broadly hypothesized that BAS scores (i) would be higher in bipolar I disorder relative to bipolar II disorder participants; (ii) would be positively associated with bipolar I hypo(manic) symptoms but show fewer associations with bipolar II hypo(manic) symptoms; (iii) would be negatively associated with bipolar II depressive symptoms but show fewer associations with bipolar I depressive symptoms; and (iv) would predict hypomanic and manic episodes in bipolar I disorder and depressive episodes in bipolar II disorder at follow-up.