UC Berkeley

Obsessive Compulsive Disorder (OCD) is characterized by intrusive and unwanted thoughts followed by specific patterns of behavior or mental acts that function to ‘neutralize’ those thoughts. These behaviors and mental acts, referred to as compulsions, share qualities with habits, responses triggered by a cue and independent of outcome values. In recent years, a number of studies have linked OCD to habit learning and failures in goal-directed control. It has been suggested that OCD, and disorders of compulsivity in general, are characterized by a bias toward learning habits and/or over-relying on habitual control at the expense of goal-directed control over behavior. The aim of the present study was to address two remaining gaps in our understanding of the relationship between OCD and failures in goal-directed control: first, to distinguish habit-related response styles from non-habit-related deficits in goal-directed control by modifying a commonly used slips-of-action test, and second, to examine more specific relationships between task performance and core motivational dimensions of OCD proposed to drive compulsive behavior. To address these goals, a sample of U.S.-based adult participants with symptoms of OCD were recruited from a mixture of undergraduate and community sources. One hundred fourteen participants completed an instrumental learning task with a subsequent slips-of-action test, and 82 of these also completed clinical self-report measures and a diagnostic clinical interview. The results support an alternative possibility to the habit hypothesis: that OCD is more strongly associated with failures in execution of goal-directed action than habit-related insensitivity to outcome devaluation. Further, results show that a commonly used task-based measure of habit-driven failures in goal-directed control in fact reflects multiple learning-related processes unrelated to habitual control, and what can appear as an increase in habitual control as measured by insensitivity to outcome devaluation on a slips-of-action test is strongly influenced by failures of learning implicit and explicit representations of causal relationships between stimuli, actions, and outcomes. These results implicate a need to investigate the intermediate cognitive processes between initial instrumental learning and execution of goal-directed control as they relate to OCD and related disorders.