“I Won’t Do What You Tell Me!”: Elevated mood and the assessment of advice-taking in euthymic bipolar I disorder

https://doi.org/10.1016/j.brat.2006.01.002Get rights and content

Abstract

We hypothesized that when individuals with bipolar disorder are engaged in goal-directed behaviour and experience elevated mood, their decision-making becomes less constrained by advice from other people. This process may allow their goal-directed behaviour to persist, contributing to manic symptoms. Three groups of 32 participants: euthymic-bipolar-I, remitted-unipolar and never-depressed healthy controls were allocated to either a high or low mood induction. Advice-taking was assessed before and after the induction by a computerized task incorporating facial images. The bipolar group significantly opposed the advice given in the task after the high mood induction. The effect was specific to bipolar disorder and remained when controlling for possible confounds. Future work could investigate the effects of this process on manic symptoms, its origins, and the wider theoretical and clinical implications.

Introduction

Cognitive approaches to psychological disorders have developed as clinical researchers identify the processes that may be maintaining current symptoms, such as biases in attention, memory, appraisal, thinking and behaviour (Clark & Fairburn, 1997; Harvey, Watkins, Mansell, & Shafran, 2004; Salkovskis, 2002). Studies of cognitive processes in bipolar disorder have identified processes that are shared with unipolar depression such as overgeneralized memory (Scott, Stanton, Garland, & Ferrier, 2000) and selective attention to negative information (Lyon, Startup, & Bentall, 1999). Other studies have shown clear changes in processing during different naturally occurring mood states. For example, individuals with bipolar disorder appear to show mood-dependent memory biases (Eich, Macaulay, & Lam, 1997; Weingartner, Miller, & Murphy, 1977) and reduced sensitivity to negative facial expressions during mania (Lembke & Ketter, 2002). Individuals with mania have also been compared to individuals with unipolar depression and healthy controls on the affective go/no-go task (Murphy et al., 1999) and the decision-making paradigm (Murphy et al., 2001).

Murphy et al. (1999) found that, compared to controls, currently manic individuals were faster to respond to happy target words than sad target words, whereas depressed individuals were faster to respond to sad than happy target words. They concluded that mania is associated with an attentional bias for positive relative to negative information. Murphy et al. (2001) presented a task that was an analogue of gambling involving the accumulation of points in a computerized task. Participants were required to decide which of two outcomes were most likely and then to place a ‘bet’ on this decision. Manic individuals, but not individuals with unipolar depression, were more likely than healthy controls to choose the less likely of two outcomes. This risk-taking tendency was correlated with current manic symptoms. Thus, while individuals with bipolar disorder may be similar to those with unipolar depression in some respects, the manic state may be characterized by specific cognitive biases, such as a bias towards processing positive information and a strategy of risk-taking during decision-making.

In the current study, we also used a computerized laboratory task that is an analogue of a real-life situation of decision-making during goal-directed activity. In this task, we assessed the role of advice-taking in moderating decision-making. We proposed that people with bipolar disorder engage in the process of mood-dependent advice-taking, and that this may be responsible for the escalation of manic symptoms. The rationale for investigating this escalating process is described below.

Mania and hypomania are highly social forms of psychopathology, characterized by extreme interpersonal behaviour. During the approach of an episode, people with bipolar disorder often seek out other people for stimulation, validation and coercion into their ambitious ideas (e.g. encouraging them to join them in a religious revival or risky business venture). They are often very irritable and critical towards those around them. Systematic studies of hypomania and mania support these views (Goodwin & Jamison, 1990; Jamison, Hamman, Gong-Guy, Padesky, & Gemer, 1990; Janowsky, Leff, & Epstein, 1970). For example, Jamison et al. (1990) found that individuals with bipolar disorder rated themselves as more sociable during hypomania compared to euthymia, yet also more uncooperative and ‘wild’. The behaviour of the manic person, for example expressing unrealistic grandiose ideas, being highly irritable or engaging in promiscuous sexual activity, is likely to provoke reactions such as criticism, anxiety, overprotectiveness and hostility (Lam, Jones, Hayward, & Bright, 1999).

There is indirect evidence that the interpersonal style described above may be evident (although to a lesser degree) when outside a bipolar episode. A study using the Depressive Experiences Questionnaire found that women with remitted bipolar depression reported less dependency needs than women with remitted unipolar depression and never-depressed controls (Rosenfarb, Becker, Khan, & Mintz, 1998). There was no difference between the bipolar and unipolar groups when currently depressed, which were higher than controls in both groups. These findings suggest that people with bipolar disorder report being less dependent on other people when they are not depressed (i.e. in a more positive mood). Lam, Wright, and Smith (2004) found that individuals with remitted bipolar disorder reported higher levels of ‘anti-dependency’ beliefs as assessed by two items on the Dysfunctional Attitudes Scale (Power et al., 1994). The items were “I don’t need the approval of others to be happy” and “A person doesn’t need to be liked to be happy”. While this study did not demonstrate whether these attitudes related to elevated mood states, a later study showed that dysfunctional attitudes were maintained during high mood in a remitted bipolar group, in contrast to non-clinical controls and a remitted unipolar group (Wright, Lam, & Newsom-Davis, 2005). Clinical observations indicate that these patterns of interpersonal behaviour are also evident during the days and weeks before an episode of hypomania or mania (Mansell & Lam, 2003; Molnar, Feeney, & Fava, 1988; Smith & Tarrier, 1992). These accounts indicate that people with bipolar disorder actively pursue personal goals when on an ‘up’ and while doing so, may ignore or dismiss other peoples’ attempts to influence their behaviour. They may avoid people who they perceive as trying to stop them doing what they want and seek out people who encourage them.

Following from the above evidence and clinical observations, we propose that bipolar individuals with bipolar disorder may be characterized by mood-dependent advice-taking that may escalate their manic symptoms. Specifically, individuals with bipolar disorder respond to elevated mood with a reduction in the extent to which they follow advice and feedback from other people that could moderate their increasing activity and increasingly grandiose beliefs. In other words, we propose that the better someone with bipolar disorder feels, the more they will try to prevent other people from affecting their thinking and behaviour. In contrast, it is proposed that when people without bipolar disorder experience an elevated mood, they may want to listen to other people and consider their advice. This can allow their goal-directed behaviour to be regulated within social norms. In contrast, the behaviour of the person with bipolar disorder will gradually deviate further from social norms, leading their behaviour to become more dysfunctional. It is important to note that it is often not possible to immediately gauge whether advice that one is given from somebody else will be helpful. We are therefore interested in the biases that individuals have to either follow or reject advice when the usefulness of that advice is perceived as ambiguous (regardless of whether the advice proves to be useful in the long-term or not).

In the current study, a computerized task was created to provide a controlled test of the hypothesis. The task assessed the use of social information in the form of ‘advice’ provided by images of faces, during goal-directed behaviour, when in either a low or high mood. In the computer-based study of goal-directed behaviour described earlier (Murphy et al., 2001), the ratio of different coloured shapes displayed on the screen represented explicit information concerning the chances of winning a reward. In the new task, information about the location of a reward on each trial was instead provided in the form of printed ‘advice’ from a person whose face was displayed on the screen. The face showed either a negative, neutral or positive emotional expression.

We predicted that after a high mood induction, a euthymic bipolar group would shift to use less information from the faces to modify their behaviour than non-clinical controls and individuals with remitted unipolar depression. We also planned to control for whether this effect might be explained by an increase in impulsivity after the high mood induction, because studies indicate that the manic phase is associated with increased impulsivity (e.g. Swann, Pazzaglia, Nicholls, Dougherty, & Moeller, 2003). If impulsivity was responsible for any reduction in advice-taking we would expect this to be reflected by shorter reaction times. We also planned to control for other possible confounding factors such as levels of depressive symptoms, past number of depressive episodes and the extent of the shift in mood following the mood induction. Finally, we explored whether facial expression (positive, neutral or negative) influenced the tendency to take advice, and whether this tendency might be related to the appraisal of different facial expressions regarding their levels of approval and trustworthiness. Specifically, we expected that individuals would be less likely to take advice from the images of people with negative facial expressions because they would be perceived as less approving and trustworthy. We speculated that this bias may be exaggerated in the bipolar group owing to the reduced processing of negative information during elevated mood states (cf. Murphy et al., 1999).

Section snippets

Participants

The three groups of participants were individuals with remitted bipolar disorder, individuals with remitted unipolar depression and normal controls. They were recruited from several sources. Local patients were contacted directly by letter after a confirmation from their consultant psychiatrist. Healthy controls and some participants were recruited through volunteer panels, the local employment centre, and circular emails and posters within Kings College and Kings College Hospital. The

Standardized scales

The differences between the three groups on the standardized scales were analysed using separate Mood Induction×Group repeated measures ANOVAs. For the BDI, it revealed no effects involving mood induction as expected, and a significant main effect of group, F(2,60)=6.55, p<0.01, η2=0.18. Paired samples t-test revealed that the bipolar group did not differ significantly in BDI from the unipolar group, t(31)=1.2, but the bipolar and unipolar groups had higher scores than the never depressed

Overview of results

The remitted bipolar group shifted to use less advice after the high mood induction in contrast to both the never depressed group and the remitted unipolar group. Moreover, the bipolar group followed the advice on significantly less trials than would be expected by chance. In other words, they were making decisions that opposed the advice. This indicated that they were not simply recognizing the lack of contingency between the advice and the location of the token, but they were imposing a

Acknowledgements

The authors would like to thank Kim Wright, Richard Brown and David Hemsley for their helpful comments on the initial research proposal for this study. We are grateful to Timothy Dalgleish, David Peck, Daniel Klein and the anonymous reviewers who provided further useful feedback on earlier versions of this manuscript. The authors are also grateful to the colleagues who helped in the recruitment of participants. The study was supported by the Psychiatry Research Trust and the Clinical Psychology

References (38)

  • D.M. Clark et al.

    Science and practice of cognitive behaviour therapy

    (1997)
  • R.A. Depue et al.

    A two-dimensional threshold model of seasonal bipolar affective disorder

  • E. Eich et al.

    Mania, depression, and mood dependent memory

    Cognition & Emotion

    (1997)
  • F.K. Goodwin et al.

    Manic-depressive illness

    (1990)
  • M. Hamilton

    A rating scale for depression

    Journal of Neurology and Psychiatry

    (1960)
  • D. Healy et al.

    Moods, misattributions and mania: an interaction of biological and psychological factors in the pathogenesis of mania

    Psychiatric Developments

    (1989)
  • A.G. Harvey et al.

    Cognitive behavioural processes across psychological disorders: a transdiagnostic approach to research and treatment

    (2004)
  • Jamison, K. R., Hamman, C., Gong-Guy, E., Padesky, C., & Gemer, R. H. (1990). Self-perceptions of interpersonal...
  • D.S. Janowsky et al.

    Playing the manic game: Interpersonal maneuvers of the acutely manic patient

    Archives of General Psychiatry

    (1970)
  • Cited by (52)

    • A self-report and behavioural assessment of impulsivity in bipolar disorder: Evidence of elevated trait but not behavioural facets

      2021, Journal of Affective Disorders Reports
      Citation Excerpt :

      Johnson et al. (2013) observed that those at risk for mania demonstrate elevated behavioural impulsivity when an elevated mood is elicited. Additionally, Mansell and Lam (2006) found that when a positive mood was induced, euthymic BD patients were more likely to behave impulsively. Considered together, these results support the possibility behavioural measures of impulsivity may be dependent on the current mood state.

    • State representation in mental illness

      2019, Current Opinion in Neurobiology
    View all citing articles on Scopus
    View full text