Abstract

Objective
The study aimed to investigate anxiety and its relation with obsessive-compulsive symptoms during the long-term course of obsessive-compulsive disorder (OCD).

Methods
We used data from the Netherlands OCD Association (NOCDA) study, which included 419 participants with OCD (aged 18–۷۹ years). Severity of obsessive-compulsive symptoms and anxiety at baseline and after two, four, and six years were entered into three models, which were analyzed using structural equation modeling: 1) the cross-lagged model, which assumes that anxiety and obsessive-compulsive symptoms are two distinct groups of symptoms interacting directly on the long-term; 2) the stable traits model, which assumes that anxiety and obsessive-compulsive symptoms result from two distinct latent factors, which are stable over the time and interact with each other; and 3) the common factor model, which assumes that anxiety and obsessive-compulsive symptoms are presentations of the same latent factor.

Results
The cross-lagged model and the stable traits model both were valid models with a good model fit. The common factor model had a poor model fit and was rejected.

Limitations
The duration of OCD varied widely between the participants (0–۶۴ years). The majority experienced obsessive-compulsive symptoms since several years, which may have affected results on the course of anxiety and the interaction between anxiety and obsessive-compulsive symptoms.

Conclusions
Anxiety and obsessive-compulsive symptoms in OCD patients do not result from a shared underlying factor but are distinct, interacting symptom groups, probably interacting by distinct latent factors.

Introduction

Obsessive-compulsive disorder (OCD) is defined by recurrent, intrusive thoughts, urges or images (obsessions) and repetitive and often ritualistic behaviors (compulsions) (DSM-5, APA, 2013). According to the learning theory obsessions provoke distress, which often presents as anxiety. Subsequently, compulsions are performed to diminish the distress or anxiety. Several studies and theories address the short-term relation between obsessions leading to anxiety leading to compulsions (e.g., Hartmann et al., 2019; Starcevic et al., 2011; Salkovskis, 1985; Rachman and Hodgson, 1980), but anxiety in OCD also occurs unrelated to the immediate distress response (Citkowska-Kisielewska et al., 2019; van Schalkwyk et al., 2016).

In fact, anxiety has a significant role in the clinical picture of OCD (Citkowska-Kisielewska et al., 2019). More severe anxiety is related to chronicity (Nakajima et al., 2018; van Oudheusden et al., 2018; Ferrão et al., 2006), impaired quality of life (Remmerswaal et al., 2020; Velloso et al., 2018; Subramaniam et al., 2013) and more functional impairment (Velloso et al., 2018; Storch et al., 2009). It is also associated with more severe obsessive-compulsive symptoms cross-sectionally (Klein Breteler et al., 2021; Sulkowski et al., 2008). In treatment studies, no effect of anxiety on treatment outcome was observed (Kathmann et al., 2022; Knopp et al., 2013; Farrel and Boschen, 2011; Steketee et al., 2019), possibly because most therapeutic interventions for OCD, such as cognitive behavioral therapy and serotonergic antidepressants, also effectively reduce anxiety (Anand et al., 2011; Blair Simpson et al., 2008). Therefore, natural follow-up studies may be preferred to study the long-term relation between anxiety and obsessive-compulsive symptoms.

Results

۳٫۱٫ Clinical characteristics
At baseline, 419 participants completed the interviews and questionnaires. Due to drop-out, at follow-up after two years 311 patients (74.2 %) still participated, at follow-up after four years 295 patients (70.4 %), and at follow-up after six years 268 patients (64 %). Complete questionnaires of the baseline and all follow-up assessments of the Y-BOCS and BAI were available for 187 participants (44.6 %). At baseline, the group of completers did not differ significantly from the group with missing data regarding age, sex, current diagnosis of OCD, current OCD symptom dimensions, Y-BOCS and BAI scores. Characteristics of all participants at baseline are summarized in Table 1.