The aim of this study was to further establish the relevance to OC phenomena of internal representations of attachment, particularly self perceptions and world assumptions. A further aim sought to examine whether the relationship between these constructs and OCD is specific (i.e., in severity or proportion) relative to two control groups (i.e., individuals diagnosed with other anxiety disorders or community participants).
Overall, the results of this study replicate previous findings (see chapters 4, 5 and 6 of this thesis), and thus indicate an association between these cognitive structures and OC symptoms and cognitions. More specifically, within the OCD group, negative attachment representations were correlated with overall OC related beliefs and OC symptoms. IWMs of self showed a stronger link with all OC belief domains. Also within the OCD group, sensitivity of self in the domain of morality was associated with more severe OC symptoms and cognitions. Consistent with previous findings (see chapter 5 of this thesis), significant associations were found between sensitivity in the domain of morality and all OC symptom dimensions (except impulses of harm) and all OC beliefs (except perfectionism).
In addition, within the OCD group, stronger beliefs that harm is avoidable in the world (i.e., controllability beliefs), perceptions of oneself as able to prevent harm in the world (i.e., self controllability) and negative perceptions of self deservingness were associated with OC symptoms and cognitions. Different OC symptoms dimensions were also found to have different patterns of correlations with world assumptions. Finally, an examination of group differences further suggested that the severity of negative IWMs of self stemming from attachment experiences, and sensitivity of self in the domains of morality and job competence, are somewhat specific to OCD in relation to other anxiety disorders or community participants.
The results of this study are largely consistent with previous research and theory implicating attachment related representations in the dynamic of OCD (Doron & Kyrios, 2005; Guidano & Liotti, 1983). Within the restricted range of symptoms and cognitions in the OCD group, the link between attachment representation and OC phenomena was maintained, suggesting that even in the high end of symptoms severity negative IWMs of self and others were associated with greater OC symptoms and dysfunctional cognitions.
Consistent with predictions, IWMs of self (i.e., attachment anxiety) showed a stronger relationship than IWMs of other (i.e., attachment avoidance) with OC related beliefs. Previous findings (see chapter 6) suggested that IWMs of other is related to OC beliefs only through its link with world view assumptions. IWMs of self was found to directly associate with OC beliefs. Unexpectedly, the link between attachment representation and OC symptoms was of a smaller magnitude than that with OC cognitions. This may reflect the restricted range of symptoms within the OCD group (i.e., very severe range) thereby attenuating the size of the correlations.
Individuals with OCD showed more negative IWMs of self than the two other control groups when controlling for mood and OC related cognitions. OCD diagnosis was associated with more severe IWMs of others than the community control group, but not when controlling for depression.
These results are consistent with previous theory and findings (Doron & Kyrios, 2005; also see chapters 6 of this thesis) suggesting that IWMs of self is associated with cognitive affective structures unaccounted for by OC related beliefs and mood symptoms (e.g., sensitive self structure). Specifically, IWMs of self may increase vulnerability to OCD via its influence on the sensitive self structures and OC cognitions. In contrast, although IWMs of others may be associated with OC phenomena via world-view assumptions over and above OC related beliefs (see chapter 6 of this thesis), this relationship may be masked by the stronger relationship between OCD and mood symptoms. Findings suggest that OCD and depression commonly co-occur with concurrent unipolar depression seen in up to a third of individuals with OCD (Angst et al., 2004; Karno, Golding, Sorenson., & Burnam, 1988; Rasmussen & Eisen, 1992). Thus, IWMs of others may effect OCD and depression via shared cognitive vulnerabilities such as perfectionism or disfuntional parenting styles (Alloy, 2001; Bhar & Kyrios, 1999; Ingram, 2003; OCCWG, 1997; Wei, Heppner, Russell, & Young, 2006).
Thus, the findings of this study support the hypothesized relationships between internal representations attachment relationships and OC phenomena.
Recent developments in cognitive theory have implicated self-structures in the etiology and maintenance of OC-phenomena (e.g. Clark & Purdon, 1993; Doron & Kyrios, 2005; Doron, Kyrios & Moulding, in press; Rachman, 1997, 1998; Salkovskis, 1985; also see chapter 1, 2, 5 and 6 of this thesis). In this study, OCD diagnosis was related to a higher proportion of individuals showing sensitivity in particular domains of self (i.e., morality and job competence). Individuals in the community control group showed a significantly lower proportion of such sensitivity, while individuals in the anxiety control group showed no statistically significant relationship with such sensitivities.
The results of this study also support past research linking issues of morality to OCD (Rachman & Hodgson, 1980; Ferrier & Brewin, 2005; Bhar, 2004; see chapter 4 and 5 of this thesis). Sensitivity in this domain indicated higher severity of OC symptoms and OC cognitions within the OCD group. This effect was generally maintained when controlling for mood symptoms commonly seen in OCD, suggesting that certain individuals suffering from OCD may be particularly vulnerable in this domain and that this sensitivity is not accounted for by more general psychopathology.
In contrast to findings from a previous study (see chapter 5 of this thesis), sensitivity in the domain of job competence was not associated with more severe symptoms or cognitions within the OCD group. These results may reflect a qualitative difference in the OCD group such that individuals with OCD show sensitivity in this self domain because of the particular significance attributed to failure in their job. Indeed, OCD was suggested to be more common in 'oversocialized' individuals (Salkovskis, 1989, p.53-54) and associated with "a high degree of conscientiousness, marked by dedication to work and an acute sense of social obligation" (Salkovskis, Shafran, Rachman, & Freeston, 1999, p. 1060).
An alternative explanation to job competence not having a significant effect on OC severity within the OCD group may be that a third variable is associated with both sensitivity in job competence and OCD. The most likely variable to underlie such a relationship would be the high unemployment rate seen in individuals suffering from OCD (see Koran, Thienemann, & Davenport, 1996). However, while both clinical samples in this study (i.e., OCD and anxiety groups) show similar rates of unemployment , individuals in the OCD group showed increased sensitivity in this self domain, but individuals in the anxiety group did not. Thus, high unemployment rate does not account for the differences between the two clinical groups although the relationship between job competence and OCD may be diminished due to other factors. Despite this, the overall results of this study provide further evidence for the model of Doron and Kyrios (2005) implicating specific self domains in the dynamic of OCD.
Like the two previous studies (see chapter 4 and 6 of this thesis), the overall findings of this study suggest a link between assumptions about the world and OC symptoms and cognitions. However, the results indicate that the severity of adherence to these world assumptions (with the exception of negative self deservingness) does not differentiate between the OCD and the other control groups. Consistent with previous findings (see chapter 4), overall OC symptom severity was associated with stronger negative self-deservingness, but positive perceptions of one's own competence in avoiding harm (i.e., the self controllability). In additions, OC symptom severity was associated with the belief that people can avoid harm in the world (i.e., the controllability belief). These findings are consistent with Wroe, Salkovskis & Richards' (2000) suggestion that the presence of an intrusive thought signifying harm is perceived significant because it is interpreted as indicating personal agency in preventative action. Thus, individuals assuming misfortune in the world can be avoided may be more susceptible to act upon danger related intrusive thoughts, in particular when holding the belief that their own actions are useful in the prevention of harm.
Perception of oneself as undeserving was associated with an increase in OC symptom severity across all OC symptom dimensions and OC beliefs. Further, negative perceptions of one's own character differentiated between the OCD and the two other control groups when controlling for depressive symptoms. This is consistent with the recent emphasis on perceptions of self within OCD research (e.g., Rachman 1997, 1998; Rowa et al., 2005). Indeed, it has been previously argued (see chapter 4 of this thesis) that low perceptions of one's own moral constitution may increase the likelihood of being distressed by intrusive thoughts. That is, the degree of negative self evaluation may influence an individuals' sensitivity to intrusions by further challenging their self view, in the domain of morality in particular (see chapter 5 of this thesis). Alternatively, having OC symptoms and cognitions may result in increased negative self evaluations.
The relationship found between particular world assumptions and OC symptom dimensions was less clear. Consistent with the pattern of relationships previously found (see chapter 4), controllability based assumptions (i.e., controllability and self controllability) were associated with overt symptoms (i.e., checking and contamination symptoms) and covert symptoms showed an association with negative self deservingness. However, the justice assumption was primarily related to checking and OC-related cognitions. Perceiving the world as a place where "people get what they deserve" may tap into the orientation of checkers towards the prevention of future misfortune in others (Rachman, 2002). That is, believing in a just world increases sensitivity to intrusions that suggest future misfortune may occur to others (e.g., fires, burglaries, etc.). Finally, negative perceptions of the world were primarily linked with contamination symptoms. This finding is surprising considering that benevolence of the world has been found to relate to OC symptoms over and above OC related beliefs (see chapter 6 of this thesis). Nevertheless, this finding suggests contamination is related to negative perceptions of the social and physical world. Thus, overall, the results of this study are consistent with previous findings. Future research should clarify the more specific relationships between world assumptions and specific OC dimensions and beliefs.
It should be noted that the results of a previous study (Chapter 4 of this thesis) suggested more complex relationships between self and world view assumptions (i.e., moderation effects) in the prediction of OC symptoms. These effects remained when controlling for OC related cognitions and mood symptoms. Theoretically, it is a constellation of world view assumptions that may differentiate between the OCD and the control groups. Power considerations did not allow such analysis to be undertaken in this sample. Future research may consider such an investigation.
Overall, the results of this study are consistent with suggestions that attachment relationships may play a role in the development and maintenance of OCD (e.g., Guidano and Liotti, 1983; Doron & Kyrios, 2005; also see chapter 1, 2 and 6 of this thesis) and with recent developments in cognitive theory implicating self concept and world view in the etiology and maintenance of OC-phenomena (e.g., Clark & Purdon, 1993; Rachman, 1997, 1998; Salkovskis, 1985). In particular, a diagnosis of OCD was found to be linked with more negative attachment representations than individuals in the community. The severity of negative IWMs of self may be greater in OCD than in other anxiety disorders and more strongly related to OC beliefs within the OCD group. Also, sensitivity in the domain of morality and job competence may be more frequent in the OCD group than in other anxiety disorders and the community. Sensitivity in the morality self domain may also be linked with increased severity on all OC symptoms dimensions (except checking) and cognitions (except perfectionism). Finally, the findings of this study suggest that severity of symptoms within the OCD group is linked with increased conviction in particular world assumptions. Thus, the severity of negative IWMs of self and sensitivity in particular self domains may be specific to OCD, providing some clues as to why certain individuals develop OCD, while others do not.
Theoretically, the findings of this study are consistent with recent suggestions that implicate perceptions of self, others and the world to stem from attachment experiences in OCD (Doron & Kyrios, 2005; Bhar & Kyrios, 2000, 2005; Guidano & Liotti, 1983; Frost et al., 2005; Kyrios, Hordern & Bhar, 2004). For instance, findings and theory suggest that negative IWMs of self are associated with the chronic overactivation of the attachment-system, resulting in anxiety, sensitivity to threats and attempts to control the environment through action (see Shaver & Mikulciner, 2002 for review). Individuals with OCD may constantly scan for internal/external stimuli that threaten their self perceptions, such as intrusions challenging sensitive self domains (e.g., an intrusion about hurting someone may endanger the individuals' perceptions of morality). In this way, such individuals become sensitized to intrusive thoughts, overestimate their negative consequences to the self or the environment and feel the need to act upon them. Individuals with OCD are more likely to perceive threat in the world, but are also more likely to act on such threats, due to dysfunctional attachment-derived models of the self and others. Thus, the findings that attachment representations and sensitive self structures are specific to OCD enable us to better understand why individuals experience OC-cognitions and why they feel the need to act on intrusive thoughts.
The present study also has implications for therapy. Specifically, it may be particularly important to address attachment, self and world-related issues when dealing with clients with OCD. For individuals that are refractory to current treatments attachment and self issues may represent a core-level belief underlying other OC-beliefs, such as perfectionism and overestimation of threat and negative perceptions of others and the world (see chapter 6 of this thesis). Examining the profile of sensitivity of self in domains such as morality and job competence may inform the content of psychotherapy(see chapter 5 of this thesis). Cognitive techniques, such as cognitive restructuring, can be used to identify and modify specific, as well as more general, distorted self perceptions by challenging perceptions of incompetence in valued domains of self and providing alternative interpretations to the triggers of such dysfunctional self perceptions (e.g., intrusive thoughts).
This study presents several methodological limitations. Firstly, several difficulties were evident due to the relatively small sample size of the clinical groups. In particular, a larger sample will allow for an investigation of lower magnitude and more complex relationships between OC phenomena and world view assumptions. Specific relationships between OC-symptom subtypes and these constructs could also be examined with larger sample sizes. Further, larger sample sizes will also enable the examination of the structural relationships between these cognitive affective variables, OC cognitions and symptoms for individuals suffering from OCD.
Secondly, the present study was cross-sectional, and while difficult, future research using longitudinal approaches would be valuable in providing stronger evidence for the causal relationship between attachment and these constructs. Finally, it would be beneficial to examine parenting patterns linked with the overvaluing of particular self domains and the perceived consequences of failure in these self domains.
In conclusion, this study has found some evidence for specificity of negative attachment representations and dysfunctional self perceptions in OCD. This will hopefully lead to the expansion of therapeutic models of OCD. Future studies may investigate whether the incorporation of notions of self and attachment into cognitive treatments of OCD enhance its efficacy and reduce relapse, particularly for individuals who are refractory to current OCD treatments.