Connectome-based predictive modeling of compulsion in obsessive-compulsive disorder

Area of Science:

• Neuroscience
• Psychiatry
• Computational Biology

Background:

• Compulsion is a core symptom of obsessive-compulsive disorder (OCD).
• Previous research explored neural mechanisms of compulsion but lacked predictive brain-based measures.
• No study has utilized whole-brain functional connectivity to predict compulsion severity.

Purpose of the Study:

• To identify brain networks capable of predicting individual compulsion levels in OCD patients using connectome-based predictive modeling (CPM).
• To investigate the importance of specific brain areas for compulsion prediction via computational lesioning.
• To compare predictive network strength in OCD patients, their unaffected first-degree relatives (UFDR), and healthy controls.

Main Methods:

• Applied connectome-based predictive modeling (CPM) to whole-brain functional connectivity data from 57 OCD patients.
• Utilized a computational lesioning approach within CPM to assess the impact of specific brain regions.
• Compared predictive network strength across OCD patients, UFDR, and healthy controls.

Main Results:

• CPM successfully predicted individual compulsion levels in OCD patients.
• Identified positively correlated networks (striatum, hippocampus) and negatively correlated networks (frontoparietal regions) with compulsion.
• Lesioning the prefrontal cortex and cerebellum significantly reduced the prediction power of the negative model.
• Observed similar network strength patterns in the negative predictive network for OCD patients and their UFDR.

Conclusions:

• CPM is a viable method for predicting compulsion severity in OCD based on functional brain connectivity.
• The prefrontal cortex and cerebellum are crucial for predicting compulsion levels.
• The identified network patterns may serve as potential vulnerability markers for OCD and related psychopathology.