Organization of self in BD

The oldest and most consistently researched approach to the self in BD is the notion of a “compartmentalized” or “modularized” self (Alatiq, Crane, Williams, & Goodwin, 2010b; Power, De Jong, & Lloyd, 2002; Taylor, Morley, & Barton, 2007). This idea is an extension of Showers’ (1992) model of the “compartmentalized self” which was developed for application to the non-clinical self-concept and the “self” in unipolar depression. Showers’ model considers the self-concept as represented in relational “self-aspects” (e.g., “me and boss,” “me and parents,” etc.). In people with a compartmentalized self-organization, each self-aspect is coded as solely positive or negative according to the individual’s overall conceptualization of the relationship defined by the self-aspect. Contrastingly, in those with an integrated self-organization, each self-aspect is coded as a mixture of positive and negative self-beliefs, according to the individual’s complex conceptualization of the relationship defined by the self-aspect. Showers proposes that in those with a compartmentalized self-organization, activation of one negative self-aspect triggers the activation of an overall negative self-concept because that self-aspect contains only negative self-beliefs (and vice versa for the activation of positive self-aspects). Notably, Showers also found that if positive self-aspects were perceived as important, compartmentalization was associated with high self-esteem and positive mood (or low depression), while if negative self-aspects were perceived as important, compartmentalization was associated with low self-esteem and negative mood (or high depression).

This “compartmentalized” or “modularized” model of self-organization is relevant to BD due to its potential to explain dramatic changes in mood, such as those observed in BD. The model suggests that in people with a compartmentalized self-organization, events which trigger a particular self-aspect that is positively valenced will activate an overwhelming sense of positive self-belief, potentially leading to hypomanic or manic behaviors. Conversely, in the same group of people, an adverse life event or environmental trigger activating a negatively valenced self-aspect will promote exclusively negative self-belief, potentially leading to dysphoric or depressive mood and behaviors. In comparison, for those with integrated self-organization, events which trigger a particular self-aspect will activate both positive and negative (balanced) self-belief because their self-aspects contain a mixture of positive and negative self-beliefs. In an initial study of the compartmentalized self in BD, Power et al. (2002) demonstrated that in BD patients in the euthymic phase, key self-aspects are compartmentalized as either completely positive or negative, leading to a modularized self-concept, whereas in non-bipolar controls (diabetic sample) the same self-aspects were integrated.

Another model of self-organization which has been explored in relation to BD is “self-complexity” (Linville, 1985, 1987; see Bhar and Kyrios, Chapter 2 in this volume, for an introduction to self complexity). Linville (1985) suggested that low self-complexity could be a factor associated with dysregulated mood because activation of a self-aspect with a particular valence could trigger other self-aspects described by the same traits, leading to the perpetuation of an excessively negative or positive sense of self-belief. Contrastingly, if a particular self-aspect is activated in an individual with high self-complexity, activation is less likely to spread to other self-aspects because they are described using different traits. Alatiq et al. (2010b) and Taylor et al. (2007) explored self-compartmentalization and self-complexity in remitted BD patients compared to remitted unipolar depression patients and healthy controls. They found that self-compartmentalization was higher in BD and unipolar depression groups than healthy controls. Alatiq et al. (2010b) found no differences in self-complexity between groups while Taylor et al. (2007) found that remitted BD patients showed greater self-complexity than healthy controls when illness-related self-aspects (e.g., depression and mania-related) were included. Akin to Taylor et al.’s (2007) self-complexity findings, Ashworth, Blackburn, and McPherson (1985) found that manic patients showed complex self-concepts relative to depressed patients.

Findings of high self-complexity in BD may have been due to the interaction between self-complexity and self-compartmentalization. If self-aspects are compartmentalized in BD, as found by both Alatiq et al. (2010b) and Taylor et al. (2007), then high complexity would allow activation to spread to more compartmentalized self-aspects, thus spiralling the activated state (either positive for mania or negative for depression) because those self-aspects do not contain mixed valence to moderate the activated state. This is consistent with the findings of Taylor et al. that high complexity was only found when highly compartmentalized illness-related concepts were included. Conversely, Linville’s (1987) suggestion that high complexity could buffer against affective extremities would be supported if self-aspects were integrated, allowing activation to spread to more integrated self-aspects whose mixed valence would promote moderation of the initially activated state.

Self-discrepancy in BD

Discrepancies between specific dimensions of self have been linked to particular mood states. Actual:ideal self-discrepancies have been associated with depression (L. Scott & O’Hara, 1993; Strauman, 1989), while actual:ought self-discrepancy has been associated with anxiety (L. Scott & O’Hara, 1993; Strauman & Higgins, 1988). In an examination of self-discrepancy theory in BD, Bentall, Kinderman, and Manson (2005) found that depressed BD patients had higher levels of actual:ideal and actual:ought self-discrepancies compared to BD patients in other phases and healthy controls, while manic BD patients showed extremely low levels of actual:ideal self-discrepancy compared to healthy controls. The authors interpreted their latter finding as consistent with the “manic defense” hypothesis (Neale, 1988), which contends that mania has the function of keeping depressive thoughts from entering consciousness and is triggered to block any stimuli that may promote depressive thinking. Aligning with this theory, the authors suggested that manic patients avoid distressing thoughts and negative affect about the self by overestimating success and underestimating weakness, thereby decreasing the discrepancy between selves.

Underpinned by evidence suggesting agitation is a common symptom of BD, Alatiq, Crane, Williams, and Goodwin (2010a) examined discrepancies between actual and feared “selves” in a student population. There were no group differences between BD and healthy controls. The authors suggest that this may drive motivation to avoid the feared self and consequently increase goal-directed behavior towards the ideal-self, thus potentially triggering elevated mood.

Self-referent processing in BD

Another broad stream of literature referring to the “self” in BD is associated with self-referent cognitive processing. This diverse group of studies is held together by the underlying idea that BD is associated with a dysfunction in in cognitive processing related to the self. Most of the empirical work conducted on self-referential processing in BD is an extension and replication of research conducted on unipolar depressed populations.

The primary theoretical basis for disturbed self-referent processing in unipolar depression is Beck’s cognitive theory of depression (Beck, 1967), which posits that depressive symptoms occur as a result of negative schemas of self, world, and the future, which colours the processing of external and internal stimuli. There is little theoretical basis underpinning empirical work examining self-referent processing in BD, with the implicit assumption being the application of Beck’s model to individuals in the depressive phase of BD and its mirror image, positing that manic symptoms occur as a result of overly positive schemas of self, world, and the future, applied to individuals in the manic phase of BD.

There has been a significant amount of empirical evidence suggesting a disturbance to self-referent processing in BD. These studies consistently demonstrate that in comparison to healthy controls, BD participants endorse and recall more negative than positive self-referent adjectives and attribute more negative than positive events to self, and this is often mediated by depressive symptomatology (e.g., Lyon, Bentall, & Startup, 1999; Molz Adams, Shapero, Pendergast, Alloy, & Abramson, 2014). Further, a pilot study of offspring of parents with BD found that there were no differences between high-risk children and low-risk children in the endorsement of positive and negative self-referent adjectives; however, high-risk children better recalled negative self-referent adjectives than low-risk children, providing preliminarily evidence that recall of self-referent material may be a vulnerability factor for BD (Gotlib, Traill, Montoya, Joormann, & Chang, 2005).

Aberrant self-referent processing in BD could also be understood as “damaged (or defensive)” self-esteem. Damaged self-esteem refers to high explicit self-reported self-esteem and negative implicit attitudes toward the self (also implied by the “manic defense hypothesis”). A number of studies have found such discrepancies in explicit and implicit self-associations (Jabben et al., 2014; Lyon et al., 1999). Damaged self-esteem has been linked to narcissistic behavior (Bosson, Swann Jr, & Pennebaker, 2000), high goal standards (Zeigler-Hill, 2006) and depressive attributional style (Creemers, Scholte, Engels, Prinstein, & Wiers, 2012).

Self-esteem in BD

Unlike other sections in this chapter, there has been a plethora of empirical evidence linking self-esteem to various domains of BD, thus here I will focus on empirical findings. There is significant evidence to suggest that BD is associated with low self-esteem during both euthymic and depressive phases and high self-esteem during manic/hypomanic episodes (Pavlickova et al., 2013, etc.). One study found that levels of self-esteem in BD patients were normal but more unstable in comparison to unipolar depressive patients and healthy controls (Knowles et al., 2007), a pattern which has also been found in children of parents with BD (Jones, Tai, Evershed, Knowles, & Bentall, 2006). Low self-esteem in BD has been linked to increased suicidality risk and worse prognosis (e.g., Halfon, Labelle, Cohen, Guilé, & Breton, 2013). Interestingly, studies which have examined vulnerability factors for BD have found that self-esteem does not predict vulnerability to BD (e.g., Pavlickova, Turnbull, & Bentall, 2014). Taken together, this evidence suggests that individuals’ perception of their self-worth or self-value may be a perpetuating factor or symptom associated with the onset of BD rather than a vulnerability factor preceding onset.

Self-stigma in BD

Self-stigma refers to the process of internalizing negative external perceptions of mental illness leading to negative feelings about the self and behaving in ways consistent with the stigmatizing beliefs of the public (e.g., not pursuing a job because of discrimination) or anticipation of negative social reactions to mental illness (Ritsher, Otilingam, & Grajales, 2003). BD is an illness which is particularly prone to stigmatization (e.g., Mileva, Vázquez, & Milev, 2012), and recent literature has shown high levels of self-stigma in people with BD (see Ellison, Mason, & Scior, 2013).

Self-stigma in BD has been associated with poor psychosocial functioning (Vazquez et al., 2011), including low perceived social support (Cerit, Filizer, Tural, & Tufan, 2012), impaired social functioning (Perlick et al., 2001), low self-esteem, and high social anxiety (Hayward, Wong, Bright, & Lam, 2002). Evidence for the relationship between self-stigma and BD has been mixed with some studies finding associations with increased depressive (Cerit et al., 2012; Vazquez et al., 2011) and manic symptomatology (Vazquez et al., 2011), while others have found no association with depressive (Hayward et al., 2002) or manic symptomology (Cerit et al., 2012; Hayward et al., 2002).

Embodied self in BD

Some of the more contemporary conceptualizations of “self” in BD emphasize the phenomenological dimension of the concept. A developing phenomenological literature suggests that disturbances of the “lived/embodied/minimal” self, defined in greater detail below, may be intimately involved in psychopathology (Cermolacce, Naudin, & Parnas, 2007; Fuchs & Schlimme, 2009). This conceptualization of self is not purely cognitive but embodied, embedded and dependent on successful interactions in the world. Correspondingly, these phenomenological conceptualizations of self are either explicitly or implicitly associated with embodiment theory (Cermolacce et al., 2007).

Fuchs and Schlimme (2009) describe embodiment as the embedding of cognitive processes in the brain and the origin of these processes in an organism’s sensory–motor experience in relation to its environment. Thus, for phenomenologists, the self is intertwined with the conscious stream of subjective experiences with a world which affords certain actions and prevents others according to ones’ sensorimotor capabilities (Cermolacce et al., 2007). BD may be associated with diminished subjective experience in the world due to the body standing in the way of environmental interaction instead of providing access to the world; correspondingly, the awareness of self becomes disturbed.

For example, Fuchs (2005) describes melancholic depression as a corporealization of the body as it loses fluidity, becomes heavier and more rigid and eventually the loss of bodily mediation of emotional experience leads to a sense of detachment from emotions and a diminished sense of self, sometimes to a point to which the sense of existence in the world may be denied. This description of disturbed self in melancholic depression is relevant for BD because depressive episodes in BD have a more severe, melancholic flavor than those in unipolar depression (Mitchell et al., 2001).

Conversely, mania and hypomania in BD may be associated with a diminished subjective experience in the world due to the body and behavior becoming separated from an individual’s sense of self; correspondingly, the self becomes distorted. Qualitative and quantitative empirical research has suggested that manic and hypomanic phases of BD are associated with a lack of self-control and a feeling of being “out of control” (Crowe et al., 2012; Russell & Moss, 2013). These feelings are reflected in a sense of fast motion perception and thinking, feeling overwhelmed, loss of autonomy, and overall a lack of self-control of behavior (Crowe et al., 2012; Russell & Moss, 2013). Thus, the body and its behaviors phenomenally become separated from consciousness, and immersed in the world without a sense of self. Such issues have led to a call for interventions which encourage and foster self-control (e.g., Jones & Burrell-Hodgson, 2008).

The only study to date examining the embodied self in BD was conducted by Haug et al. (2014), who examined disturbances of self-awareness or “sense of self,” defined as being immersed in the world, continuous, and coherent, in individuals with schizophrenia and psychotic BD. The authors found high levels of disturbed sense of self, which was associated with poorer social functioning in both individuals with schizophrenia and psychotic BD.