Risk of schizophrenia among people with obsessive-compulsive disorder: A nationwide population-based cohort study

Abstract

High comorbidity of obsessive-compulsive disorder (OCD) has been reported in patients with schizophrenia. The sequence of OCD and schizophrenia onset might clarify the underlying pathophysiological relationships between these two disorders, but available evidence is limited. In this study, we used a population-based cohort to investigate the risk of schizophrenia in people with newly diagnosed OCD. Patients who were first diagnosed with OCD from 2000 to 2013 were selected from the Longitudinal Health Insurance Research Database. The non-OCD group was randomly sampled, and matched with the OCD group by gender, age, urbanization level, and income. Cox regression analyses and competing risk model were used to estimate the risk of schizophrenia, adjusting for potential confounding factors. In total, 2009 patients with OCD and 8036 controls were identified. The crude incidences of schizophrenia in the OCD and non-OCD groups were 876.2 per 100,000 person-years and 28.7 per 100,000 person-years, respectively. After adjustment, a substantially higher risk of schizophrenia was observed in the OCD group (hazard ratio = 30.29, 95% confidence interval = 17.91–51.21). Male gender, age of OCD onset before 20 years, and antipsychotic prescription were associated with schizophrenia. Patients with comorbidity of autistic disorder have higher risk of schizophrenia (hazard ratio = 4.63, 95% confidence interval = 1.58–13.56). In conclusion, OCD diagnosis, male gender, age of OCD onset before 20 years, comorbidity of autistic disorder, and antipsychotic use were associated with higher risk of schizophrenia. It is essential for psychiatrists to note that OCD may be the initial presentation of schizophrenia.

Introduction

Schizophrenia and obsessive-compulsive disorder (OCD) are two chronic and debilitating mental disorders. The lifetime prevalence of schizophrenia is approximately 0.55% (McGrath et al., 2008), and that of OCD is 2%–3% (Rasmussen and Eisen, 1992; Ruscio et al., 2010). Meta-analyses revealed that OCD comorbidity rates in subjects with schizophrenia were approximately 12%–23% (Achim et al., 2011; Buckley et al., 2009; Swets et al., 2014). The prevalence of cormorbid OCD in patients with schizophrenia is significantly higher than those expected to occur independently. The greater-than-chance rate of comorbid OCD in patients with schizophrenia suggests a possible pathophysiological link between these two diseases (Berrios, 1989; Buckley et al., 2009; Hall, 2017; Poyurovsky et al., 2004).

Earlier studies have reported that the appearance of obsessive-compulsive symptoms (OCS) or OCD in patients with schizophrenia can prevent psychosis progression (Rosen, 1957). However, later studies generally have not supported this observation. Clinically, comorbid OCD in patients with schizophrenia is associated with earlier onset of psychosis, higher prevalence of comorbid depression, more severe psychopathology, poorer social function, and higher rates of hospitalization (Cunill et al., 2009; Cunill et al., 2013; Faragian et al., 2012; Grover et al., 2017).

OCD onset can occur at any time throughout the entire course of schizophrenia (Poyurovsky et al., 2012; Schirmbeck and Zink, 2013; Zink, 2014). The temporal relationships between OCD and schizophrenia might present in five different modes. Firstly, OCD occurs before schizophrenia as an independent diagnosis. Secondly, OCD occurs before psychosis as a part of the at-risk mental state. Third, OCD occurs simultaneously with schizophrenia onset. Fourth, OCD occurs after the first psychotic episode of schizophrenia. Fifth, OCD occurrence in patients with schizophrenia is the result of antipsychotic medication (Grillault Laroche and Gaillard, 2016).

The transition from OCD to schizophrenia has been described in the literature (Berrios, 1989; Cederlof et al., 2015; Devulapalli et al., 2008). A meta-analysis of four studies revealed that no difference exists in the mean age of onset between OCD and schizophrenia in patients comorbid with both disorders (Devulapalli et al., 2008). However, this meta-analysis is limited due to its small number and heterogeneity of studies (Devulapalli et al., 2008). Several sequential studies have revealed that there is a tendency of earlier onset of OCD than schizophrenia among patients comorbid with OCD schizophrenia (Faragian et al., 2012; Seedat et al., 2007; Sterk et al., 2011).

Recently two national cohort studies investigated the longitudinal risk of schizophrenia in patients first diagnosed with OCD. Analyzing the data from Swedish National Patient Register, Cederlof et al. (2015) concluded that patients with OCD diagnosis have a 2.7-folds higher risk of a subsequent diagnosis of schizophrenia compared with those without OCD. Using the data from longitudinal nationwide Danish registers, Meier et al. (2014) reported that patients with pre-existing OCD diagnosis have a higher risk of schizophrenia (incidence ratio = 6.90) and schizophrenia spectrum disorders (incidence ratio = 5.77).

The temporal relationship of OCD and schizophrenia is worth further studies. We used the nationwide population-based cohort study design to explore the sequential risk of schizophrenia in patients with an initial diagnosis of OCD.