Small-world networks in individuals at ultra-high risk for psychosis and first-episode schizophrenia during a working memory task

Abstract

Disturbances of functional interaction between different brain regions have been hypothesized to be the major pathophysiological mechanism underlying the cognitive deficits of schizophrenia. We investigated the small-world functional networks in individuals at ultra-high risk (UHR) for psychosis, first-episode schizophrenia (FESPR) patients, and healthy controls. All participants underwent the electroencephalogram during a control task and a working memory (WM) task. Small-world properties of the theta band were reduced in FESPR relative to controls during the WM task. Small-worldness of the UHR during the WM task exhibited intermediate value between that of controls and FESPR. These results imply that the suboptimal organization of the brain network may play a pivotal role in the schizophrenia pathophysiology.

Introduction

Schizophrenia is a chronic mental disorder with impairments in multiple cognitive domains including the working memory (WM) [18], [29]. Underlying these cognitive impairments, disturbances in functional interaction between different brain regions have been hypothesized [5], [6], [11]. A method of graph theoretical analysis investigating the so-called “small-world network” may offer valuable information in exploring this hypothesis. Characterized by a high clustering coefficient (an index of local clustering), and a short path length (an index of global integration), the small-world network reflects an optimal network topology for functional segregation and integration to maximize information processing efficiency between brain regions [1], [45], [46], [51]. Using the electroencephalography (EEG), small-world properties were found to be decreased in schizophrenia patients compared to healthy subjects during a WM task [38], [40]. It has also been demonstrated that decreased small-worldness is associated with lower cognitive function in healthy subjects [20]. Together, it has been suggested that the suboptimal small-world network may play an important role underlying the cognitive impairments of schizophrenia [38], [40].

Evidence suggests that cognitive deficits may be present not only in the first episode of schizophrenia but also prior to the emergence of frank psychotic symptoms. The ‘putative’ prodromal, ultra-high-risk for psychosis (UHR), subjects show cognitive impairments such as WM deficits [8], [15], [16], [30], [32]. In search of underlying neural dynamics, previous studies of small-world networks during cognitive tasks were limited in that subjects were chronic, multi-episode schizophrenia patients [38], [40]. The only small-world network study reported of first-episode patients explored the network properties during the resting state [41]. Furthermore, small-world network architecture in UHR subjects has not yet been investigated. Studying the small-world topology associated with cognitive function in first episode schizophrenia (FESPR) and those at UHR may provide an important insight at the ongoing pathophysiological alterations of network integration.

In the present study, we investigated the small-world network properties in UHR subjects, FESPR patients, and healthy controls (HC) during a WM task. Based on previous reports [38], [40], we hypothesized that the small-worldness will also be reduced in FESPR during the WM task. Since neurocognitive impairments are reported to be present before the frank onset of psychotic symptoms, we also hypothesized that small-world properties will be altered in UHR subjects, but to a lesser degree compared to FESPR patients.