Application of graph theory across multiple frequency bands in drug-naïve obsessive-compulsive disorder with no comorbidity

Abstract

Recently, graph theoretical analysis based on resting-state functional magnetic resonance imaging has provided a means of investigating the complex brain connectome in obsessive-compulsive disorder (OCD) patients. However, these studies have been restricted to spontaneous blood oxygen level-dependent (BOLD) signals with frequency bands between 0.01 and 0.08 Hz, and the parameters from graph theory across multiple frequency bands have seldom been studied. Here, we calculated global metrics (small-worldness, global efficiency and modularity) and nodal metrics (degree centrality, betweenness centrality, nodal clustering coefficient and shortest path) at four different frequency bands (slow-2 (0.199–0.25 Hz), slow-3 (0.074–0.198 Hz), slow-4 (0.027–0.073 Hz) and slow-5 (0.01–0.027 Hz), from 0.01 to 0.25 Hz) in seventy-three OCD patients and ninety healthy controls. The analyses were also calculated in traditional low-frequency bands (0.01–0.08 Hz) for reference. For the global metrics, the OCD patients showed increased small-worldness and modularity only in the slow-3 band. For the local metrics, we observed a frequency-dependent characteristic, with the main significant differences in regions including the right precentral gyrus, occipital region, right anterior cingulum cortex and fusiform cortex. Our results suggested frequency-specific abnormalities of the brain connectome in OCD and the future studies may need to consider different frequency bands when measuring spontaneous activity in the brain.