The attachment system is hypothesized to be an evolutionally based biological mechanism that drives infants to seek proximity to the primary attachment figure (or caregiver) in cases of danger or need (Bowlby, 1969, 1973). Based on the child's interactions with their main caregivers, internal working models (IWMs) of self and others are formed (Bowlby, 1969, 1973). In particular, the accessibility and responsiveness of the attachment figure to the infantÂ´â”Â¢s emotional signals are considered pivotal for this process (Bowlby, 1969).
Research and theory suggest that IWMs influence the child's perceptions of the self and world by predisposing the child to select and interpret incoming information in a way consistent with his IWMs of self and others (Guidano and Liotti, 1983; Fraley, 2002; Sroufe et al., 1999; also see Chapter 2 for review). Thus, early experiences serve as an important influence on an individualÂ´â”Â¢s subsequent perceptions of self and the environment.
In adulthood, attachment representations manifest themselves as patterns of interpersonal behavior, particularly in close relationships (Hazan & Shaver, 1987; see Shaver & Mikulincer, 2002 for review). Bartholomew (1990) suggested that the adult IWMs consist of two-dimensions derived from early attachment: an internalized sense of self-worth, and beliefs about the responsiveness and trustworthiness of others . The two dimensions underlying adult attachment (i.e., self and other IWMs) have affective-motivational consequences (Brennan et al., 1998). Particularly, negative IWMs of self lead to increased anxiety related to abandonment. Negative IWMs of others result in high avoidance of intimacy/dependence due to fear of rejection (Bartholomew, 1990; Bartholomew & Horowitz, 1991). For instance, individuals with a negative IWMs of self, but relatively positive other IWMs demonstrate increased abandonment anxiety and reassurance seeking behaviors from others. Individuals with negative IWMs of self and of others show high abandonment anxiety and discomfort with dependence and intimacy (Brennan et al., 1998). Thus, attachment representations manifest themselves in belief-systems and styles of emotional responding, and motivate action.
IWMs have been found to influence the individual's later adult behaviors (Bartholomew, 1991; Dozier, Stovall, & Albus, 1999; Hazan & Shaver, 1987; Main & Goldwyn, 1984; also see Crowell, Fraley, & Shaver, 1999 for review) and particular IWMs are associated with adult psychopathology including depression (Carnelley, Pietromonaco, & Jaffe, 1994) and anxiety (Bartholomew, 1990; Riskind et al., 2004; Safford, Alloy, Crossfield, Morocco, & Wang, 2004; Williams & Riskind, 2004). Researchers have also examined mediators between attachment representations and distress and psychopathology (e.g., Fossati et al., 2005; Wei, Russell, & Zakalik, 2005). For instance, in a recent study using structural equation modeling (SEM) methods with a student sample, Wei, Mallinckrodt, Russell and Abraham (2004) found the influence of attachment anxiety (i.e., IWMs of self) on depressive symptoms was partially mediated and moderated by maladaptive perfectionism, a cognitive style associated with OCD. Furthermore, the influence of attachment avoidance (IMW of other) was fully mediated by maladaptive perfectionist cognitions.
Previous research has also linked attachment relationships with other dysfunctional cognitive domains identified in current cognitive models of OCD (OCCWG, 1997). For instance, negative IWMs of self (i.e., high attachment-anxiety) were found to be linked with increased appraisals of normal life circumstances as threatening regardless of the level of objective threat (e.g., Mikulincer, Birnbaum, Woddis, & Nachmias, 2000), difficulty in suppressing thoughts (Mikulincer, Dolev, & Shaver, 2004), and a tendency to devalue oneself in threatening situations (Mikulincer, Orbach, & Iavnieli, 1998). Furthermore, negative IWMs of others (i.e., high attachment-avoidance) were associated with success in emotional and cognitive distancing from the source of distress (Fraley & Shaver, 1997; also see Shaver, & Mikulincer, 2002 ), and with failure to suppress when the individual was experiencing high cognitive loads (e.g., Mikulnicer, Dolev & Shaver, 2004). Thus, negative IWMs have been found consistently to be linked with cognitive processes that are associated with OCD.
Attachment related perceptions such as perceptions of self, others and the world have also been previously linked with OC-symptoms. For example, OCD has been associated with individuals overestimating how threatening the world is (OCCWG, 1997) and perceiving others as having unrealistic perfectionistic expectations of them (e.g., Bhar & Kyrios, 1999). Increased OC symptoms and/or cognitions have also been associated with an ambivalent sense of self (see Bhar & Kyrios, 2000), a view of oneself as dangerous to others (Ferrier & Brewin, 2005) and perceiving oneself as responsible for preventing harm (see Salkovskis & Forrester, 2002 for review).
Doron and Kyrios (2005) proposed that internalized attachment experiences may increase the likelihood of developing a limited number of valued self-domains (e.g., morality, job competence, social acceptability), in which the individual feels incompetent (i.e., sensitive domains). In addition, attachment experiences generalize and influence the development of the individualsÂ´â”Â¢ perceptions of the world. Doron and Kyrios (2005) suggest that sensitive self structures underlie vulnerability to intrusive thoughts resulting in the triggering of OC related cognitions and anxiety. Assumptions about the world (e.g., negative perceptions of the social and physical world) further increase feelings of vulnerability in the world. In additions, a more complex relationship of self perceptions and assumptions about the world may determine the susceptibility to specific types of intrusions and associated responses. For instance, assuming misfortune in the world can be avoided may increase the likelihood to act upon danger-related intrusive thoughts (e.g., being contaminated). This may be particularly true for individuals holding the belief that their own actions are useful in the prevention of harm (e.g., washing). This in turn, may increase the likelihood of them using such neutralizing behaviors.
Two studies have directly examined this model. Results of the first study (see chapter 4 of this thesis) indicated that assumptions regarding the benevolence of the world added to the prediction of OC symptom severity over and above OC related cognitions and mood symptom scores in a non-clinical cohort. Specific OC dimensions were also found to be linked with particular self perceptions and world view assumptions (e.g., the perception of oneself as active in the prevention of harm moderated by beliefs that negative events in the world can be avoided was associated with compulsions). Findings from the second study directly assessing this model (see chapter 5 of this thesis) suggested that individuals who value the domains of morality, job and scholastic competence, and social acceptability, but who feel incompetent in these domains (i.e., 'sensitive' domains of self), held a greater level of OC-related beliefs (e.g., overestimation of threat/responsibility, perfectionism/intolerance for uncertainty and importance/control of thoughts). Sensitivity in the domains of morality, job competence and social acceptability was also related to higher levels of OC-symptoms. Further, these findings were generally maintained when controlling for global self-worth. Based on these results, it was argued that sensitivity of self-concept may constitute a vulnerability to OC cognitions and phenomena.
While the above research suggests that attachment relationships and associated cognitive structures may be etiologically important in OCD, only one study has directly examined the relationship between adult attachment and OCD. In this study, 36 out-patients with OCD, 16 depressed out-patients and 26 controls were compared on three attachment related dimensions (i.e., comfort with closeness, capacity to depend on others, and fear of abandonment) using the Revised Adult Attachment Scale (RAAS; Collins & Read, 1990). Although the OCD and depressed groups were found to be more insecure on the fear of abandonment than non-clinical participants on this attachment scale, no differences were found between the two clinical groups . However, the RAAS has been criticized for having relatively low degree of measurement precision (see Fraley, Waller, & Brennan, 2000). Also, the RAAS does not directly tap into the currently accepted two-dimensional model of adult attachment (i.e., self and other models). Finally, the relationship between attachment dimensions, OC related cognitions and underlying cognitive-affective structures, such as an individual's perceptions of self, others and the world was not investigated. Such an investigation would assist in the conceptualization of OCD through contextualizing the current cognitive models of OCD within broader models of attachment literature.