Exploring latent clusters in pediatric OCD based on symptoms, severity, age, gender, and comorbidity

Given diverse symptom expression and high rates of comorbid conditions, the present study explored underlying commonalities among OCD-affected children and adolescents to better conceptualize disorder presentation and associated features. Data from 830 OCD-affected participants presenting to OCD specialty centers was aggregated. Dependent mixture modeling was used to examine latent clusters based on their age- and gender adjusted symptom severity (as measured by the Children's Yale-Brown Obsessive-Compulsive Scale; CY-BOCS), symptom type (as measured by factor scores calculated from the CY-BOCS symptom checklist), and comorbid diagnoses (as assessed via diagnostic interviews). Fit statistics favored a four-cluster model with groups distinguished primarily by symptom expression and comorbidity type. Fit indices for 3-7 cluster models were only marginally different and characteristics of the clusters remained largely stable between solutions with small clusters of distinct presentations added in more complex models. Rather than identifying a single classification system, the findings support the utility of integrating dimensional, developmental, and transdiagnostic information in the conceptualization of OCD-affected children and adolescents. Identified clusters point to the centrality of contamination concerns to OCD, relationships between broader symptom expression and higher levels of comorbidity, and the potential for complex/neurodevelopmental presentations.