Assessment of worry and OCD: how are they related?

Abstract

This study investigates the overlap and differences between measures of worry and Obsessive Compulsive Disorder (OCD). It was expected that: (1) worry and obsessive compulsive symptoms are distinct concepts, yet that (2) worry and the cognitive components of OCD are more strongly related compared to the behaviour components of OCD. By means of confirmatory analysis it was found that all six components, i.e. worry, obsessive rumination, impulses, washing, checking and precision proved distinct components. Of the obsessive compulsive components, rumination was found to be most closely associated with worry. In further investigating the relative impact of obsessive compulsive symptoms on worry, it was found that obsessive rumination offered the largest unique contribution to the prediction of worry. When controlled for depressive mood, the overlap between rumination and worry dropped substantially. The results of this study clearly underline the differentiation between worry and obsessive compulsive symptoms.

Introduction

Both Generalised Anxiety Disorder (GAD) and Obsessive Compulsive Disorder (OCD) are characterised by excessive and/or uncontrollable cognitive processes associated with negative affect. Because of these common processes there might be considerable overlap between these two disorders. Worry, the basic characteristic of GAD (APA, 1994) and obsessions, a main feature of OCD (APA) share some similar characteristics. Turner, Beidel, and Stanley (1992, p. 265) summarise several similarities between worry and obsessions. The authors state that: (1) both phenomena occur in normal as well as patient populations; (2) the form and content of worry and obsessions appear to be similar in normal and clinical groups; (3) both occur with greater frequency and are associated with greater perceptions of uncontrollability in clinical populations than in normal groups; (4) both worry and obsessions are associated with adverse mood; and that (5) some type of vulnerability factor determines why individuals develop pathological worries or obsessions and others do not.

However, apart from these similarities, worry and obsessions appear to differ to a great extent on several dimensions. With respect to content, worries are defined as ego-syntonic, i.e. more realistic and concerning real-life problems, whereas obsessions are defined as ego-dystonic, i.e. inappropriate or not the kind of thought that one expects oneself to have (APA, 1994). GAD worries may be largely self-initiated and occur in response to specific common daily events and do not include themes of dirt and contamination, aggressive impulses, or horrific images so typical of obsessions in OCD (Turner et al., 1992). Although there is evidence that worry in GAD and normal worry can not be distinguished in terms of content differences (Craske, Rapee, Jackel, & Barlow, 1989), the results of the study of Dugas, Freeston, Ladouceur, Rhéaume, Provencher, and Boisvert (1998) indicated that high levels of worry about future events may differentiate GAD patients from other anxiety patients.

In terms of form differences worries typically appear in the form of thoughts whereas obsessions also consist of images and impulses (cf. Borkovec and Inz, 1990, Turner et al., 1992). Furthermore, worry does not appear to be resisted as strongly as obsessional thinking although both phenomena are experienced as uncontrollable (Craske et al., 1989, Turner et al., 1992). The behaviour response to worry consists of avoidance and reassurance seeking (Wells, 1997) and is not likely to comprise neutralisation behaviour in the form of rituals as in OCD. In terms of process characteristics, the study of Clark and Claybourn (1997) revealed that worrisome thoughts were considered more disturbing than obsessive-like intrusive thoughts in a non-clinical population. Moreover, the focus on the possible consequences of negative events was uniquely predictive of worry, whereas concern about personal meaning of the thought was a unique process dimension for obsessive intrusive thoughts.

There is evidence that on a symptom level GAD and OCD can, in general, reliably be differentiated by clinicians (Brown and Barlow, 1992, Brown et al., 1993). Moreover, the rate of co-occurrence of GAD and OCD seems to be rather low (Brown et al., 1993, Abramowitz and Foa, 1998). Nevertheless, the differentiation between GAD and OCD seems to be less clear when overt compulsions are absent (Brown et al., 1994, Freeston et al., 1994). This indicates that diagnostic differentiation between GAD and OCD is merely due to key features, such as compulsions, instead of clear diagnostic boundaries (cf. Brown et al., 1993). In an empirical study of Abramowitz and Foa, 20% of the patients with OCD also met DSM-III-R criteria (APA, 1987) for GAD. The severity of OCD symptoms, however, was not elevated by the presence of GAD. The authors conclude that the severity of obsessions is not related to additional worry present in those OCD patients with comorbid GAD, which provides further support that obsessions and worry are distinct phenomena.

In examining the relationship between GAD and OCD on the basis of self-report measures, it was found that worry was more systematically associated with measures of obsessions (i.e. checking and doubting) compared to measures of compulsions (i.e. washing and slowness; Tallis & De Silva, 1992). The authors therefore conclude that worry might represent a cognitive variant of checking. Rather than being functionally equivalent, Freeston et al. (1994) indicated on the basis of their study, that checking and worry are indirectly linked through stronger links with obsessional loss of mental control.

Since both worry and obsessive compulsive symptoms are associated with negative affect, the relationship between worry and obsessive compulsive symptoms can be mediated by depressive symptomatology. For example, in the study of Brown et al. (1993), GAD and OCD patients did not differ on depressive features as evidenced by their scores on self-report measures, clinician rating scales and patterns of diagnostic comorbidity. There is evidence that dysphoric mood increases severity of obsessions in OCD patients (Riccardi & McNally, 1995) although the pathways of this relationship are not clearly understood (Rachman, 1997). As is described by Rachman (1998, p. 398), the sequence in which depressive mood and obsessive compulsive symptoms affect each other varies: “Often the obsessions cause a lowering in mood but in other instances a lowering of mood is followed by an increase in obsessional activity”.

The aim of the present study was to investigate the relationship between obsessions and worry on the basis of two self-report questionnaires. The Padua Inventory-Revised (PI-R; van Oppen, Hoekstra, & Emmelkamp, 1995) and the Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) were used to measure obsessive compulsive symptoms and worry respectively. The dimensional structure of the PI-R was investigated by means of large clinical groups (i.e. OCD patients, panic disorder patients and social phobics). A five-factor solution emerged which could be interpreted as: (1) impulses; (2) washing; (3) checking; (4) rumination; and (5) precision. Looking at the phenomenology, the subscales of the PI-R adequately reflect the different symptoms of OCD. The reliability of these subscales proved to be satisfactory to excellent. With regard to discriminant validity, apart from the subscale impulses, the scales were able to differentiate between obsessive compulsives from patients with panic disorder, social phobics and normals. Finally, consistent evidence for the construct validity of the PI-R was found (van Oppen et al.). To date, the relationship between the PI-R and worry has not been investigated.

The PSWQ was developed to measure aspects of clinically significant worry, the generality of worry over time and situations, the excessiveness of the experience and the uncontrollability of the process (Molina & Borkovec, 1994). The PSWQ consists of one scale and had proven to possess high internal consistency (cf. Meyer et al., 1990, Molina and Borkovec, 1994, van Rijsoort et al., 1999).

It was expected that: (1) worry and obsessive compulsive symptoms are distinct components, yet that (2) worry and the cognitive components of OCD (i.e. subscales impulses and rumination) are more strongly related compared to the relationship between worry and the behaviour components of OCD (i.e. subscales washing, checking and precision). Further, given the relationship of both worry and obsessive compulsive symptoms to negative affect, the relationship between worry and obsessive compulsive symptoms will be investigated when controlled for depression. It was expected that: (3) the overlap between worry and obsessive compulsive behaviour will decrease when depression is taken into account.