Decision making in the real world is complex and is modulated by numerous factors. We and others have found that specific aspects of decision making are abnormal in individuals with obsessive-compulsive disorder (OCD). We hypothesize that these abnormalities are related to both dysfunctions in neural cortico-basal ganglia circuits and the development of clinical symptomatology in such domains as indecisiveness, behavioral inflexibility, and compulsive repetition of actions. OCD affects about 1 person in 40; obsessions and compulsions are found in many more. Elucidating their relationship to underlying neurocognitive abnormalities is of fundamental importance and may ultimately lead to diagnostic clarification and new therapeutic interventions. This proposal seeks to employ the computational drift diffusion model (DDM) framework to refine and quantify the cortico-basal ganglia theory of OCD, which is widely accepted but remains qualitative. We employ the ?hold- your-horses? model of subthalamic nucleus (STN) function to characterize individual variation in responses to changes in task difficulty and task context, on both behavioral and neurobiological levels, and their relations to OCD symptomatology. The model suggests that STN activity and connectivity contribute to individual ability to adjust the process of decision formation to meet current task demands, by modulating how much evidence needs to be accumulated before a choice between alternatives is made. This contributes to behavioral flexibility; we expect this STN function to be abnormal in OCD. Consistent with this hypothesis, prior studies and our pilot data suggest that individuals with OCD require more evidence before they make a perceptual categorization decision, which may lead to indecisiveness; but they tend to accumulate less evidence before making an value-based decision (when asked to choose which of two options is preferred), which may lead to reduced response inhibition and poor quality decisions (such as we have previously documented). Differences in decision making across these two contexts have rarely been systematically investigated (and never in OCD). Our novel task allows direct comparison of perceptual and value-based decisions, while varying tasks difficulty, using the rigorous DDM- based analysis on both behavioral and neural levels. Our pilot data demonstrate the feasibility of our approach. This proposal combines a novel theoretical framework of neurocognitive abnormalities in OCD, a sophisticated computational approach, and a specific mechanistic hypothesis implicating the STN and associated networks in OCD pathophysiology. These will be tested using hierarchical Bayesian estimation of DDM parameters in conjunction with state-of-the-art fMRI analyses. Quantifying STN abnormalities with DDM will help us to disentangle OCD-related abnormalities in basic neurocognitive processes and to begin to place the cortico-basal ganglia model of OCD on a more quantitative footing. In future work these analyses can be applied transdiagnostically. Ultimately, we hope this work will produce biomarkers of pharmacological treatment response, and may contribute to novel therapeutics modulating the implicated circuitries.
Obsessions and compulsions (OCs) occur in about a third of the population; they are central to the DSM diagnosis of obsessive-compulsive disorder (OCD) but are seen across a range of other diagnoses, as well as in subclinical forms that may nevertheless cause distress. Abnormal evidence accumulation, poor decision making, and behavioral inflexibility are well established in OCD, as are abnormalities in regional brain activity (e.g. in the subthalamic nucleus and associated cortico-striatal connections); however, how they map onto the heterogeneity of OCD symptomology remains unclear. Computational modeling may be instrumental in linking behavior, symptomatology, and underlying neurobiology; advancing this line of research may lead further to development of biomarkers of treatment response, and to new ideas about how to manipulate these networks using not only deep brain stimulation (which was already shown to be successful in refractory cases) but also less invasive procedures, such as transcranial magnetic stimulation or neurofeedback.
