The role of early maladaptive schemas in predicting exposure and response prevention outcome for obsessive-compulsive disorder

Abstract

This is the first study that explores whether early maladaptive schemas are related to treatment outcome for patients with obsessive-compulsive disorder (OCD). The sample consisted of 88 outpatients with a diagnosis of OCD who completed exposure and response prevention treatment. The Yale-Brown Obsessive-Compulsive Scale (Y-BOCS), the Beck Depression Inventory and Young Schema Questionnaire – Short Form were administered before and after treatment. Regression analyses using post-treatment Y-BOCS as the dependent variable indicated that higher scores on the abandonment schema at pre-treatment were related to poor outcome and explained 7% of the variance in symptoms at post-treatment. Higher scores on the self-sacrifice schema at pre-treatment were related to good outcome and explained 6% of the variance in obsessive-compulsive symptoms at post-treatment. During treatment, only changes in the failure schema were significantly related to good outcome and explained 18% of the variance in symptoms at post-treatment.

Introduction

One of the challenges facing cognitive behavioral therapy (CBT) today is developing effective therapies for “difficult-to-treat” patients. Although the efficacy of CBT for obsessive-compulsive disorder (OCD) has been well established in many studies (Abramowitz, 2006, Eddy et al., 2004), approximately 50% of patients still do not respond optimally to CBT, including many who drop out or relapse (Baer and Minichiello, 1998, Cottraux et al., 2005, Stanley and Turner, 1995).

Several researchers have proposed that many patients with underlying personality disorders and characterological issues do not fully respond to traditional CBT (Beck and Freeman, 1990, Fals-Stewart and Lucente, 1993; Sookman & Steekete, 2010). These patients often have difficulties in forming therapeutic alliances, are unable or unwilling to comply with treatment protocols, have difficulty complying with homework exercises and have limited access to cognitions and emotions. These factors contribute to reduced CBT response, particularly with short-term treatment protocols (Beck and Freeman, 1990, Young et al., 2003).

The prevailing view in contemporary clinical practice is that severe personality pathology has a negative impact on CBT outcome for OCD patients, but there is little empirical evidence supporting such a relationship. To date, most studies have focused on the influence of categorical Axis-II diagnoses on CBT response, and the findings from these studies have been inconsistent (Fricke et al., 2006, Keeley et al., 2008). The most consistent finding to date is that schizotypal personality traits predict poorer treatment outcome (Keeley et al., 2008). Some have argued that findings related to the relationship between personality pathology and CBT response are inconsistent because personality characteristics are camouflaged when utilizing broad, categorical Axis-II diagnoses. These researchers (Rees, Anderson, & Egan, 2005) advocate for the use of dimensional measures that may more effectively identify underlying personality factors related to treatment outcomes for OCD patients.

Maladaptive cognitive schemas are proposed as the core of personality disorders (Beck and Freeman, 1990, Young et al., 2003). These schemas are thought to influence how people view themselves, others and the world. Given that personality pathology is quite common in OCD (Reich & Simonsen, 2008), examining underlying schemas among people with OCD is important in order to gain a deeper understanding of the relationship between personality pathology and treatment outcome. Increased knowledge about underlying schemas, particularly among treatment non-responders, could inform more comprehensive and individually tailored OCD treatment programs.

As an expansion of traditional CBT, Young and Klosko (1993) developed schema therapy to treat patients with chronic psychological disorders when CBT has been unsuccessful. Although they have not studied OCD directly, their model may be relevant for the treatment of OCD patients who are resistant to standard CBT. Schema therapy integrates aspects of other theoretical schools (e.g., psychodynamic, gestalt, attachment theory) into a richer, unified conceptual treatment model that addresses the core themes that are typically present among patients with personality pathology. A central assumption of the schema model is that the origins of most severe personality pathology can be traced to unmet emotional needs in childhood. The model posits that early maladaptive schema (EMS) emerge from these unmet needs and that these trait-like vulnerabilities contribute to the development of maladaptive coping styles. Fifteen EMS are grouped into the following four domains: disconnection (i.e., the beliefs that one’s needs for safety, nurturance, empathy, acceptance and respect will not be met in a predictable manner), impaired autonomy (i.e., beliefs regarding one’s ability to separate and function independently from others), exaggerated standards (i.e., rigidity) and impaired limits (i.e., deficiencies in self-discipline and in setting emotional and interpersonal limits) (Hoffart et al., 2005). According Young and Klosko’s (1993) theory, EMS related to basic safety, abandonment and abuse, are particularly important. These EMS are most closely related to traumatic events in childhood, which may result in unsecure attachments and negatively affect the ability to form secure therapeutic relationships.

To our knowledge, only two studies have directly examined the EMS’ of patients with OCD, but only at a descriptive level. Lochner et al. (2005) compared patients with OCD (n = 59) and trichotillomania (n = 26) using Young’s Schema Questionnaire (YSQ – Short Form) and found that patients with OCD reported more maladaptive schemas than patients with trichotillomania in five EMS. Unrelenting standards (i.e., the underlying belief that one must strive to meet very high internalized standards of behavior and performance, typically to avoid criticism) and self-sacrifice (i.e., excessive focus on voluntarily meeting the needs of others at the expense of one’s own gratification in daily situations) evidenced the highest scores in both the OCD and trichotillomania samples. However, exclusion of male participants in this study reduces the generalizability of these findings.

In a cross-sectional study conducted by Atalay, Atalay, Karahan, and Çaliskan (2008), the YSQ-total score in the OCD group (n = 45) was significantly higher compared to the healthy control group (n = 45). Compared to the control subjects, social isolation, vulnerability to harm and pessimism were the most activated schemas among OCD patients. To a lesser degree, patients with OCD also scored significantly higher on eight additional maladaptive schemas (emotional deprivation, defectiveness, failure, incompetence, subjugation, unrelenting standards, entitlement and approval seeking) compared to controls. A weakness in this study was the exclusion of Axis-I and Axis-II comorbidity, resulting in an atypical OCD sample because treatment-seeking OCD patients typically present with multiple co-occurring diagnoses.

Nevertheless, none of these studies provided information concerning whether EMS predict who might profit from OCD treatment. To our knowledge, there are no outcome studies involving standard CBT for OCD that address this issue. However, there is some empirical evidence that an extended, integrative CBT model focused on underlying schemas (Sookman & Pinard, 1999), can yield a positive outcome for patients who are resistant to standard CBT. In a recent book, Sookman and Steketee (2010) reported positive outcome data from 39 CBT-resistant OCD patients using a schema-based approach. The results were based on two studies (Sookman and Pinard, 1999, Sookman et al., 2003) that consisted of 7 and 32 patients, respectively. Results indicated that a total of 32 patients showed clinically significant improvement after once- or twice-weekly treatment sessions over a period of 10 months. The results also revealed that core beliefs/underlying schemas, measured with the Vulnerability Schemata Scale (Sookman, Pinard, & Beck, 2001), reliably improved for responders and did not change for non-responders.

The extended treatment length in the aforementioned studies of schema-based treatment of OCD (Sookman and Pinard, 1999, Sookman et al., 2003) likely played an important role in the positive outcomes observed. Personality trait variables, such as schemas, are by definition thought to be quite enduring and resistant to change, particularly in standard CBT for OCD. Although a few studies have found that standard CBT for OCD can lead to changes in personality variables [e.g., defense mechanisms (Albucher, Abelson, & Nesse, 1998) or personality disorders (McKay, Neziroglu, Todaro, & Yaryura-Tobias, 1996)], CBT alone is not likely to be of sufficient intensity or duration to impact enduring personality problems for most patients. Nevertheless, it remains unclear which maladaptive schemas do change as a result of standard CBT for OCD and for which patients these schema are likely to change.

The present study is the first to examine the role of EMS in the outcome of standard CBT for OCD. The main purpose of the present study was to explore whether EMS measured before treatment predicts changes in obsessive-compulsive symptoms after treatment. Based on Young and Klosko’s (1993) theory, we expected that patients with higher scores on the schemas related to basic safety (i.e., abandonment and mistrust/abuse) would have poorer outcome. We also expected that patients who failed to experience clinically significant changes in obsessive-compulsive symptoms would generally have more personality pathology at pre-treatment compared to patients who experienced clinically significant improvement. Furthermore, we investigated whether changes in EMS were related to post-treatment improvement in obsessive-compulsive symptoms.