Recent developments in adult attachment research have enabled researchers to make the empirical connections necessary to investigate the influence of internal working models (IMW) on adult behaviors (e.g., Bartholomew & Horowitz, 1991; Dozier, Stovall, & Albus, 1999; Hazan & Shaver, 1987; Main & Goldwyn, 1984). One of the most researched areas has focused on close, intimate, and romantic relationships (Crowell, Fraley, & Shaver, 1999; Shaver & Mikulincer, 2002). Hazan and Shaver (Hazan & Shaver, 1987) proposed a model in which early internal representations manifest themselves in adulthood as patterns of interpersonal behavior, particularly in close relationships. This model was later expanded by Bartholomew (1990), who suggested that the degree to which an individual's sense of self worth is internalized corresponds inversely to the level of need for external validation (i.e., dependent behavior or anxiety related to abandonment). The degree to which the "other" is perceived as available and responsive corresponds inversely to the level of avoidance of emotional intimacy. Subsequent studies have supported this two dimensional construct underlying adult attachment styles (e.g., Bartholomew & Horowitz, 1991; Brennan, Clark, & Shaver, 1998; Fraley & Waller, 1998).
More recently, Shaver and Mikulincer (2002) suggested that activation of the attachment system involves three sequential processes; threat assessment and monitoring leading to the activation or deactivation of the attachment system (i.e., is there threat?); evaluation of availability of the external or internal attachment figure (i.e., can I rely on myself or others?); and the assessment of viability of proximity seeking behaviors (i.e., will the attachment figure respond or not respond?).
Shaver and Mikulincer (2002) argued that insecurely-attached individuals either hyperactivate or deactivate their attachment system. Individuals with a negative "self" model show high attachment anxiety, chronically activating (or hyperactivating) their attachment system in an attempt to minimize the distance from the attachment figure and elicit their support (Kobak & Sceery, 1988). Individuals with such high attachment anxiety focus on the attachment figure, and use clinging and controlling responses in order to attain proximity, support and affection (Bartholomew, 1990; Mikulincer, 1998). Conversely, individuals with a negative "other" model show high avoidance in intimate relationships and use deactivation strategies to maximize the distance from the attachment figure and avoid painful experiences (Kobak & Sceery, 1988). Deactivation of the attachment system inhibits proximity seeking behaviors, thereby reducing reliance on the attachment figure.
Although the variety of self-report measures used in an attempt to tap these two dimensions has led to some confusion among researchers (Brennan et al., 1998), several valid and reliable measures have been developed recently (see Crowell et al., 1999 for review). For example, Brennan et al. (1998) developed a 36-item scale based on factor analyses of numerous previous measures. This measure has shown good internal-consistency, test-retest reliability and high construct, predictive, and discriminate validity (Crowell et al., 1999).
Bartholomew (1990) also argued that an individual's positioning on the self and other dimension-scales can be used to categorize four prototypical "peer" or "close relationship" attachment patterns in adulthood. These include secure (positive self/positive other), preoccupied (negative self/positive other), fearful (negative self/negative other) and dismissive (positive self/negative other). These four categories correspond with the Ainsworth et al. (1978) infant categorizations (Brennan et al., 1998).
Indeed, similar individual differences in the expression of attachment behaviors were shown in infants and adults. For instance, adults were found to exhibit signs of distress when they perceived the attachment figure as inaccessible, to show desire for proximity to the attachment figure in times of stress, and to experience increased comfort in the presence of the attachment figure (Shaver, Hazan, & Bradshaw, 1988; Weiss, 1991). Further, research with adult attachment measures indicates that IWMs play an important role in psychopathology (Crowell et al., 1999; Williams & Riskind, 2004b). Negative attachment representations have been linked with vulnerability to depression (e.g., Ingram, 2003) and symptoms of depression (Carnelley, Pietromonaco, & Jaffe, 1994,), anxiety (Bartholomew, 1990; Riskind et al., 2004; Safford, Alloy, Crossfield, Morocco, & Wang, 2004; Williams & Riskind, 2004a), eating disorders (Friedberg & Lyddon, 1996), and low self-esteem (Brennan & Morris, 1997; Collins & Read, 1994; Feeney & Noller, 1990). Conversely, a large body of evidence has demonstrated that a secure adult attachment style may buffer the effects of other risk factors (see Crowell et al., 1999; Thompson, 1999). For instance, self-reports of attachment security are associated with increased perceptions of self efficacy, positive affect regulation strategies and reliance on problem-solving for coping with personal and interpersonal stressors (e.g., Birnbaum, Orr, Mikulincer, & Florian, 1997; Collins & Read, 1990; Lussier, Sabourin, & Turgeon, 1997; Mikulincer, 1998; Mikulincer & Florian, 1998). These findings support Ainsworth's suggestion that attachment relationships in adulthood provide feelings of security and belonging, which facilitate functioning in the wider network of social relations (Ainsworth & Bowlby, 1991).
Importantly, research has also linked attachment relationships with dysfunctional cognitive processes similar to those 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) and dysfunctional perfectionistic tendencies (e.g., Wei, Mallinckrodt, Russell, & Abraham, 2004). Thus, negative IWMs have been found to be linked with cognitive processes potentially associated with OCD.
Overall, internal representations of attachment relationships seem to have a significant bearing on adult behavior and related cognitions. Attachment representations may have a direct influence on the development of dysfunctional cognitions, but also an indirect effect through their influence on self construct and world view. The following sections will explore the role of IWMs of self and world-view as vulnerability factors in OCD.