2 - (IOP 41) Comparison of Inference-based and Cognitive-behavioral Models of Obsessive-compulsive Disorder in a Clinical Sample

The traditional cognitive-behavioral (CBT) model of obsessive-compulsive disorder (OCD) posits that the misappraisal of intrusive thoughts as meaningful results in distress and leads to compulsive rituals. Yet this model does not entirely account for OC symptoms, so additional constructs that might add explanatory value should be considered. One prospect is the inference-based approach (IBA), which postulates that obsessions result from an inductive reasoning process termed "inferential confusion" (IC), which leads one to confuse a possibility with reality, and that vulnerable self-themes (e.g., fear of self) explain the heterogeneity of obsessions between individuals. The IBA has yet to undergo independent empirical evaluation with a clinical sample. The present study examined the CBT and IBA models as predictors of OC symptom dimensions in a sample of patients with OCD.

Participants were 131 individuals with an OCD diagnosis enrolled in an intensive treatment program who completed the following questionnaires at admission: the Dimensional Obsessive-Compulsive Scale (DOCS), Obsessive Beliefs Questionnaire-20 (OBQ), Inferential Confusion Questionnaire-Expanded Version (ICQ-EV), Fear of Self Questionnaire (FSQ), and Quick Inventory of Depression Symptomatology (QIDS). Hierarchical linear regressions predicted each DOCS subscale: Contamination (DOCS-C), Responsibility for Harm (DOCS-R), Unacceptable Thoughts (DOCS-UT), and Symmetry (DOCS-S). QIDS scores were entered in Block 1, OBQ scores in Block 2, and ICQ-EV and FSQ scores in Block 3. To reduce multicollinearity, only OBQ subscales that predicted a DOCS dimension in preliminary analyses were included in the model for that dimension. Variables exhibiting suppression effects were also excluded.

The model predicting DOCS-C scores was not significant (R2 = .08, p = .200) and had no significant individual predictors. The model predicting DOCS-R scores was significant (R2 = .35, p < .001), with QIDS and OBQ Threat Overestimation scores emerging as the only significant predictors. ICQ and FSQ scores did not significantly increase the variance accounted for in the model (∆R2 = .01, p = .545). The model predicting DOCS-UT scores was significant (R2 = .30, p < .001), with OBQ Responsibility scores emerging as the only significant predictor. ICQ and FSQ scores did not significantly increase the variance accounted for in the model (∆R2 = .00, p = .776). Lastly, the model predicting DOCS-S scores was significant (R2 = .19, p < .001). OBQ Perfectionism/Intolerance of Uncertainty scores were the only significant predictor. ICQ and FSQ scores did not significantly increase the variance accounted for in the model (∆R2 = .01, p = .361).

Contrary to previous findings, IBA constructs did not predict any individual OC symptom dimensions. Obsessive belief domains predicted Responsibility for Harm, Unacceptable Thoughts, and Symmetry, but not Contamination symptoms. The substantial overlap between the CBT and IBA models in our results raises doubts about IBA’s unique contribution. However, limitations of cross-sectional designs prevent causal inferences. Future research should continue comparing these two models, as well as others, to optimize etiological models of OCD.