A study of the psychometric properties of the Obsessive Beliefs Inventory and Interpretations of Intrusions Inventory on clinical Italian individuals

https://doi.org/10.1016/S0887-6185(03)00013-6Get rights and content

Abstract

Forty-three individuals with obsessive-compulsive disorder (OCD), 17 with generalized anxiety disorder (GAD) and 50 non-clinical controls (SC), completed the Italian versions of the Obsessive Beliefs Questionnaire (OBQ) and the Interpretations of Intrusions Inventory (III), along with measures of obsessive-compulsive (OC) symptoms, depression, anxiety and worry. OBQ and III showed an excellent reliability and temporal stability. The six OBQ subscales were reasonably distinct from each other, whereas the three subscales of the III were highly interrelated. The OBQ, and in less measure the III, discriminated between OCD, GAD, and normal controls. In the OCD cohort, the two instruments correlated moderately with a measure of obsessive-compulsive symptoms but also with a measure of worry. However, a linear regression analysis evidenced a specific relationship between the OBQ and obsessive-compulsive symptoms over and above worry. Overall, at least three cognitive domains (intolerance of uncertainty, excessive concern about the importance of controlling one’s thoughts and perfectionism) seemed specific to OCD, whereas overimportance of thoughts and inflated responsibility barely discriminate clinically anxious individuals from non-clinical ones. It is concluded that OBQ and III are useful measures in Italian individuals and that more research is warranted to possibly refine these two instruments.

Introduction

The Obsessive Beliefs Questionnaire (OBQ) and the Interpretations of Intrusions Inventory (III) are instruments developed by an international group (Obsessive Compulsive Cognitions Working Group—OCCWG) to assess cognitions thought to be relevant in the etiology and maintenance of obsessions and compulsions (OCCWG, 1997). The OBQ contains six scales measuring dysfunctional beliefs: inflated responsibility, overimportance of thoughts, excessive concern about the importance of controlling one’s thoughts, overestimation of threat, intolerance of uncertainty, and perfectionism. Aim of the III is to assess three appraisal domains: overimportance of thoughts, excessive concern about the importance of controlling one’s thoughts and inflated responsibility. According to OCCWG (1997), appraisals are considered ways in which meaning is given to a specific event such as the occurrence of an intrusions. Thus, they may take the form of expectations, interpretations, or other types of judgments. On the other hand, beliefs are relatively enduring assumptions that are held by an individual and that are not specific to a particular event (e.g., dysfunctional attitudes or irrational beliefs). To date, OBQ and III have been studied in both clinical and non-clinical individuals drawn from English-language populations, whereas for non-English-language populations data are available for non-clinical subjects only (OCCWG, 2001, OCCWG, in press; Sica et al., 2002, Taylor et al., 2002).

A recent study by the OCCWG (in press) examined 248 individuals with DSM-IV obsessive-compulsive disorder (OCD), 105 individuals with other anxiety disorders except OCD (mainly with panic disorder), 87 community controls, and 291 students. The results showed that OBQ and III scales had generally a good internal consistency (Cronbach alpha coefficients equal or above 0.80), an adequate test–retest reliability and were highly correlated with one another. OBQ and III scales also showed criterion-related validity since individuals with OCD scored higher than students and community controls on each scale. In addition, three OBQ scales (Control of thoughts, Importance of thoughts, Responsibility) and all the III scales distinguished individuals with OCD from anxiety controls. On the other hand, anxious patients scored higher than both non-clinical control groups on all subscales of both measures. Thus, three OBQ domains (tolerance of uncertainty, overestimation of threat and perfectionism) appeared to be OCD-relevant but not OCD-specific. Correlations of the OBQ and III subscales with measures of OCD symptoms, mood and worry, showed that both scales were as highly correlated with non-OCD symptom measures (anxiety, depression and worry) as they were with OCD ones. Likewise, other analyses showed that the OBQ and III had a non-unique relevance to OCD. Authors concluded that while the reliability, criterion-related validity, and convergent validity of the OBQ and III were satisfactory, the discriminant validity was not fully supported from the data.

Sica et al. (2002) examined data collected from Greek, Italian and US students. OBQ and III scales completed by these cohorts showed comparable values of internal consistency and item-remainder correlations. On the contrary, correlation patterns with measures of obsessive-compulsive symptoms were different across the three cohorts. Such differences were attributed to well-known cultural traits. That is, in US individuals, beliefs about personal control and perfectionism were more correlated with OC symptoms compared to the other two groups. In Italian individuals, importance of thoughts and responsibility were weakly related to OC symptoms. In Greek individuals measures of contamination and checking were not related at all to cognitive measures. It was concluded that OC cognitive domains seemed not to have the same relevance in non-clinical students belonging to different countries and that more research was necessary to evaluate if such findings were reliable and applicable to clinical populations.

Sica, Novara, and Sanavio (2002a) evaluated superstitious and non-superstitious individuals on OBQ, III, and well-established measures of obsessive-compulsive symptoms, depression, anxiety, and worry. After controlling for anxiety and depression, high superstitious subjects scored higher than low-superstitious ones on OBQ Overestimation of threat. It was concluded that superstitiousness may be a predisposing factor for general rather than specific psychopathology, because overestimation of threat appears to be a feature of almost all anxiety disorders (see also Borkovec & Roemer, 1995, Stoeber & Joormann, 2001).

Finally, Sica, Novara, and Sanavio (2002b) compared individuals with a different degree of religiousness by means of OBQ, III, and measures of obsessive-compulsive symptoms, depression and anxiety. OBQ Control of thoughts and Importance of thoughts were associated with OC symptoms only in religious subjects. It was concluded that religion might play a role in OCD phenomenology, particularly in individuals whose religion is a prominent part of their cultural values.

As noted above, studies on clinical samples have been carried out only with English-language individuals so far. The aim of the present study, therefore, was to investigate the psychometric properties of OBQ and III when administered to clinical Italian individuals. We were also interested in examining the criterion-related, convergent and discriminant validity of these measures. With respect to the first point it was hypothesized that OBQ and III scales would show reliabilities comparable to US counterparts, since such measures were developed with attention to cross-cultural differences. With regard to validity we expected to find results in line with US studies. Nevertheless, we did not assume that US and Italian would be equivalent data because the few studies on non-English population suggested an idiosyncratic association between OC cognitive domains and symptoms. In particular: (a) responsibility appeared to be less relevant than other OC domains in Italian individuals (Sica et al., 2002); (b) the construct of importance of thoughts seemed not specific to OCD, because in Italian individuals it was linked to religious beliefs (Sica et al., 2002b); and (c) overestimation of threat gave the impression of being a domain common to several anxiety disorders (Sica et al., 2002a).

Section snippets

Participants and procedure

Three groups of participants were enrolled in the present study: patients with DSM-IV diagnosed obsessive-compulsive disorder as their most severe problem, patients with DSM-IV diagnosed generalized anxiety disorder (GAD) as their most severe problem, and student controls (SC). OCs and GADs were excluded if they had a current or past psychotic disorder or a current substance use disorders. In addition, GADs were excluded if they had a current or past OCD.

The OCD and GAD groups were recruited

Descriptive statistics

Table 1 provides descriptive statistics on various demographic variables for the groups. The three groups differed in age but not in education or gender. About one-half of OCD and GAD patients were married or cohabiting, whereas almost all the students were single. About 60% of both clinical groups were employed; 6% of OCD and 14% of GAD individuals were unemployed. Table 1 also indicates levels of symptomatology across the three groups. The OCDs reported greater obsessionality on the PI than

Discussion

Our purpose was to test if the measures of obsessive-compulsive cognitions were reliable and valid when applied to Italian individuals with clinical symptoms. The reliability analyses revealed that in clinically anxious individuals, the OBQ and III scales measured obsessive-compulsive cognitions in a very consistent way. The only exception was the OBQ Tolerance of uncertainty scale that showed unreliable values in individuals with GAD. As showed above, such a result may depend by the scale

Acknowledgements

Authors are grateful to Gail Steketee and Randy Frost for the advice on revision of the manuscript.

References (28)

  • A.T Beck et al.

    An inventory for measuring clinical anxiety: psychometric properties

    Journal of Consulting and Clinical Psychology

    (1988)
  • A.T Beck et al.

    An inventory for measuring depression

    Archives of General Psychology

    (1961)
  • Centomo, C., & Sanavio, E. (1992). Padua Inventory e Beck Depression Inventoy: indagine su studenti di scuola media...
  • P.M.G Emmelkamp et al.

    Metacognition, specific obsessive-compulsive beliefs and obsessive-compulsive behavior

    Clinical Psychology and Psychotherapy

    (1999)
  • Cited by (59)

    • The relationship between cognitions and symptoms in obsessive-compulsive disorder

      2018, Journal of Affective Disorders
      Citation Excerpt :

      Numerous studies have investigated the specific association between the OBQ score and OCD symptoms. For example, several studies found specific associations between OC cognitions and OCD symptoms, while controlling for anxiety or depressive symptoms (Janeck et al., 2003; Morillo et al., 2007; Sica et al., 2004; Wheaton et al., 2010; Wu and Carter, 2008). Support for the cognitive model of OCD was also provided by experimental studies using laboratory paradigms, which suggested that cognitive biases are related to the pathogenesis of obsessions (See Calkins et al., 2013 for a review).

    • Confirmatory factor analysis of the Dutch Intolerance of Uncertainty Scale: Comparison of the full and short version

      2013, Journal of Behavior Therapy and Experimental Psychiatry
      Citation Excerpt :

      However, IU is found be related to other pathologies as well. Research including clinical (Sica et al., 2004; Steketee, Frost, & Cohen, 1998; Tolin, Abramowitz, Brigidi, & Foa, 2003) as well as non-clinical samples (Boelen & Reijntjes, 2009; Dugas, Gosselin, & Ladouceur, 2001; Holaway, Heimberg, & Coles, 2006) has shown that IU may also be involved in obsessive compulsive disorder (OCD). Steketee et al. (1998) demonstrated that IU was a strong predictor for the severity of OCD symptoms.

    View all citing articles on Scopus
    View full text