(PS5-80) Using Item Response Theory to Develop a Shortened Version of the Obsessive-Compulsive Inventory-Revised for Eating Disorders

Eating Disorders (EDs) are serious mental illnesses characterized by disturbances in eating and body image and are associated with significant cognitive, emotional, and psychosocial impairment. This impairment is compounded by high rates of comorbidity with up to 85% of individuals with EDs meeting diagnostic criteria for another disorder. Comorbidity is associated with increased ED symptom severity, greater suicidality, and poor treatment response. EDs and obsessive-compulsive disorder (OCD) are highly comorbid, with 35-55% of individuals with AN also meeting criteria for OCD. Individuals with AN report significantly higher scores on dimensions of OCD than healthy control and OCD dimensions are associated with overall ED pathology. However, the relationship between EDs and OCD remains poorly understood, and a critical first step in advancing research is ensuring that OCD measures function appropriately in ED samples. The Obsessive-Compulsive Inventory-Revised (OCI-R) is an 18-item self-report instrument that measures the distress associated with obsessions and compulsions. The aim of the present study is to examine the psychometric properties of the Obsessive-Compulsive Inventory-Revised in an ED sample and to develop a shorter version of the OCI-R for ED samples based on item response theory (IRT) methods. The OCI-R was administered to 418 individuals with AN (n = 241; mean age (SD) = 30.65 (10.31) years) and AAN (n = 207; mean age (SD) = 29.61 (9.69) years). We used an exploratory factor analysis (EFA) to evaluate the dimensionality of the OCI-R and determine whether the data are appropriate for IRT analysis. We applied the Graded Response Model (GRM) to evaluate the item parameter estimates, item characteristic curves, and item information functions. Item parameter invariance between AN and AAN was inspected using IRT differential item function (DIF). Seven items with poor psychometric properties were removed, resulting in an 11-item short version of the OCI-R. The refined measure effectively differentiates between individuals with disparate levels of OCD symptom severity, provides maximal information at moderate to high levels of severity, and demonstrates equivalent psychometric properties across AN and AAN. The abbreviated OCI-R maintains a reasonable balance of content from the six original factors and will enhance the assessment of OCD in ED populations, offering a more efficient and psychometrically sounds tool for research and clinical use. Importantly, this shortened measure can be used to screen for OCD in ED populations.