The small-world networks have been proposed to be the optimal brain network with maximal efficiency in information processing at a minimal cost, and facilitating rapid information interactions. Some research suggested that patients with cognitive impairment have disrupted small-world networks. Leukoaraiosis (LA) is characterized by white matter hyperintensities (WMH) on magnetic resonance imaging (MRI), which is known to be associated with vascular cognitive impairment (VCI). In this study, we aimed to investigate the small-world properties of the brain functional network and their relationship with cognition impairment in LA patients.

Subjects were selected from communities using MRI, and given cognitive tests including Montreal Cognitive Assessment (MoCA) and Clinical Dementia Rating (CDR), and then divided into LA with normal cognition (LA-NC), LA with vascular mild cognitive impairment (LA-VaMCI) and the healthy control (HC) groups. Resting-state functional MRI data were collected and graph theoretical analysis was applied to evaluate the difference of the small-world networks among the three groups.

Both LA-NC and LA-VaMCI cases presented with modified small-world topological properties. Compared with HC group, LA-NC group had increased small-world properties, whereas LA-VaMCI group showed decreased small-world properties. Moreover, we also found the small-world network properties was negatively correlated with the score of MoCA in LA-VaMCI patients.

Our study indicated that disrupted small-world networks may be one mechanism of cognitive impairment in LA patients, and the degree of small-world network changes might be associated with the level of cognitive impairment. The small-world networks could be used as a potential marker to detect vascular cognitive impairment in LA patients.