Disrupted pathways from the frontal-parietal cortices to basal nuclei and the cerebellum are a feature of the obsessive-compulsive disorder spectrum and can be used to aid in early differential diagnosis.

A marker for distinguishing patients with obsessive-compulsive disorder (OCD) spectrum has not yet been identified. Whole-brain resting-state effective and functional connectivity (rsEC and rsFC, respectively) networks were constructed for 50 unmedicated OCD (U-OCD) patients, 45 OCD patients in clinical remission (COCD), 47 treatment-resistant OCD (T-OCD) patients, 42 chronic schizophrenia patients who exhibit OCD symptoms (SCHOCD), and 50 healthy controls (HCs). Multivariate pattern analysis (MVPA) was performed to investigate the accuracy of using connectivity alterations to distinguished among the aforementioned groups. Compared to HCs, rsEC connections were significantly disrupted in the U-OCD (n = 15), COCD (n = 8), and T-OCD (n = 19) groups. Additionally, 21 rsEC connections were significantly disrupted in the T-OCD group compared to the SCHOCD group. The disrupted rsEC networks were associated mainly with the frontal-parietal cortex, basal ganglia, limbic regions, and the cerebellum. Classification accuracies for distinguishing OCD patients from HCs and SCHOCD patients ranged from 66.6% to 98.0%. In conclusion, disrupted communication from the frontal-parietal cortices to subcortical basal nuclei and the cerebellum may represent a functional pathological feature of the OCD spectrum. MVPA based on both abnormal rsEC and rsFC patterns may aid in early differential diagnosis of OCD.