Disruptions in small-world cortical functional connectivity network during an auditory oddball paradigm task in patients with schizophrenia

Abstract

P300 deficits in patients with schizophrenia have previously been investigated using EEGs recorded during auditory oddball tasks. However, small-world cortical functional networks during auditory oddball tasks and their relationships with symptom severity scores in schizophrenia have not yet been investigated. In this study, the small-world characteristics of source-level functional connectivity networks of EEG responses elicited by an auditory oddball paradigm were evaluated using two representative graph-theoretical measures, clustering coefficient and path length. EEG signals from 34 patients with schizophrenia and 34 healthy controls were recorded while each subject was asked to attend to oddball tones. The results showed reduced clustering coefficients and increased path lengths in patients with schizophrenia, suggesting that the small-world functional network is disrupted in patients with schizophrenia. In addition, the negative and cognitive symptom components of positive and negative symptom scales were negatively correlated with the clustering coefficient and positively correlated with path length, demonstrating that both indices are indicators of symptom severity in patients with schizophrenia. Our study results suggest that disrupted small-world characteristics are potential biomarkers for patients with schizophrenia.

Introduction

Each local brain region has its own functions and is connected with other brain regions both structurally and functionally, which facilitates information transfer among distant brain regions (Dosenbach et al., 2007, Friston, 2011). Information processing through the complex cortical functional network has been known to be impaired in patients with schizophrenia, which therefore has been suggested to be a disrupted cortical and subcortical network disorder. Among neuroimaging modalities, electroencephalography (EEG) has proved to be one of the most useful tools to investigate brain information processing. Previous EEG studies reported altered event-related potential (ERP) waveforms, disrupted functional connectivity, and reduced source activity for specific ERP components in patients with schizophrenia (Pascual-Marqui et al., 1999, Winterer et al., 2003b, Kayser et al., 2010). For example, reduced P300 amplitude and prolonged P300 latency have been consistently found in these patients (Potts et al., 1998, Mathalon, 2000, van der Stelt et al., 2005). Some studies also reported disrupted functional connectivity, particularly in the temporo-patietal junction and fronto-temporal connection (Friston and F.C., 1995, Schall et al., 1999, Lawrie et al., 2002, Wolf et al., 2007). EEG source imaging studies have shown reduced source activities at the left insular, left postcentral gyrus, and left temporal areas in patients with schizophrenia (Pae et al., 2003, Kawasaki et al., 2007, Wang et al., 2010).

Recently, an increased number of researchers have focused on changes in the cortical functional connectivity network, because alterations in the cortical connectivity network might provide clues to reveal the underlying neural mechanisms of schizophrenia. Many of these studies adopted graph theory to quantify global and local changes in the cortical functional connectivity network (Stam and Reijneveld, 2007, Bullmore and Sporns, 2009, Rubinov and Sporns, 2010). In particular, the small-world network has been regarded as one of the most suitable models to elucidate information transfer in the human brain (Li et al., 2007, Bolanos et al., 2013). The small-world network is the middle ground between random network and regular network. The small-world network is characterized by a higher clustering coefficient than random networks and a shorter path length than regular networks, where the clustering coefficient and the path length reflect the amount of segregation of highly inter-connected units and the amount of integration of the whole network, respectively (Watts and Strogatz, 1998). Therefore, the small-world characteristics of the brain allow for more efficient information transfer among distant brain regions.

In previous EEG studies, patients with schizophrenia consistently showed disrupted small-world networks characterized by decreased clustering coefficients and prolonged path lengths in the resting state (Micheloyannis et al., 2006, Rubinov et al., 2009, Jalili and Knyazeva, 2011) and during the working memory task (Schinkel et al., 2011), compared to healthy control subjects. However, most EEG network analysis studies, including these examples, were based on sensor-level connectivity analysis. Therefore, these studies failed to report specific cortical regions that contribute to the disruption of the small-world cortical functional network in schizophrenia. This shortcoming was also noted by De Vico Fallani et al. (2010), who performed the first network analysis (node degree and network density) of EEG source-level functional connectivity in patients with schizophrenia during the 2-back working memory task. EEG topographies cannot be directly attributed to the underlying cortical regions, since sensors may contain information from multiple brain sources, some of which might overlap, and topographic maps are sometimes smeared out due to inhomogeneous conductivity distributions in the human head. This so-called “volume-conduction effect” can cause spurious connectivity between scalp EEG channels (Haufe et al., 2013), eventually leading to failure in identifying the region-specific changes in cortical functional connectivity networks. In addition, most previous studies also used binary (unweighted) functional networks for the estimation of small-worldness. The use of binary networks might lead to losses of information about interaction strength that might be useful to characterize small-world characteristics in patients with schizophrenia, because arbitrary threshold values are required to convert the original functional connectivity network into a binary network.

In the present study, the small-worldness of brain networks was evaluated using a source-level weighted functional connectivity network analysis. The use of source-level weighted network analysis allows for observing the alternation of small-world networks in specific local cortical regions as well as in the global brain network pattern. Moreover, the use of weighted networks is not only free from ambiguity in determining threshold values, but can also preserve the unique traits of the original network without distortion. Although a series of P300 EEG studies demonstrated that patients with schizophrenia generally show significant decreases in task performance and brain activity (Kügler et al., 1993, Polich and Kok, 1995, Pae et al., 2003, Kawasaki et al., 2007, Wang et al., 2010), to the best of our knowledge, auditory oddball tasks have not been used to investigate changes in small-world cortical functional networks in patients with schizophrenia. In addition, no previous studies have investigated the relationships between small-worldness and symptom severity of patients with schizophrenia. We hypothesized that the small-worldness of cortical functional connectivity networks would be disrupted during P300 processing of an auditory oddball task in patients with schizophrenia, and that the disrupted small-world characteristics would be correlated with symptom severity.