Altered Network Topologies and Hub Organization in Adults with Autism: A Resting-State fMRI Study

Takashi Itahashi,Takashi Yamada,Daiki Jimbo,Nobumasa Kato,Seiji Shioda,Hiromi Watanabe,Kazuo Toriizuka,Ryuichiro Hashimoto,Motoaki Nakamura,Satoru Hayasaka

doi:10.1371/journal.pone.0094115

Takashi Itahashi, Takashi Yamada + Show 8 more

Open Access

https://doi.org/10.1371/journal.pone.0094115

Copy DOI

Abstract

Recent functional magnetic resonance imaging (fMRI) studies on autism spectrum condition (ASC) have identified dysfunctions in specific brain networks involved in social and non-social cognition that persist into adulthood. Although increasing numbers of fMRI studies have revealed atypical functional connectivity in the adult ASC brain, such functional alterations at the network level have not yet been fully characterized within the recently developed graph-theoretical framework. Here, we applied a graph-theoretical analysis to resting-state fMRI data acquired from 46 adults with ASC and 46 age- and gender-matched controls, to investigate the topological properties and organization of autistic brain network. Analyses of global metrics revealed that, relative to the controls, participants with ASC exhibited significant decreases in clustering coefficient and characteristic path length, indicating a shift towards randomized organization. Furthermore, analyses of local metrics revealed a significantly altered organization of the hub nodes in ASC, as shown by analyses of hub disruption indices using multiple local metrics and by a loss of “hubness” in several nodes (e.g., the bilateral superior temporal sulcus, right dorsolateral prefrontal cortex, and precuneus) that are critical for social and non-social cognitive functions. In particular, local metrics of the anterior cingulate cortex consistently showed significant negative correlations with the Autism-Spectrum Quotient score. Our results demonstrate altered patterns of global and local topological properties that may underlie impaired social and non-social cognition in ASC.

Highlights

Autism spectrum condition (ASC) has been characterized by impairments in social and communication skills, combined with repetitive and restricted behavior
The area under the curve (AUC) analyses revealed that, while s was comparable between the groups, r, C, and L were significantly lower in participants with ASC than in normal controls (NCs) (r: p = 0.013; C: p = 0.014; L: p = 0.002) (Figure 2)
This study examined the topological properties of the restingstate functional brain network in participants with ASC using a graph-theoretical analysis

Summary

Introduction

Autism spectrum condition (ASC) has been characterized by impairments in social and communication skills, combined with repetitive and restricted behavior. This spectrum comprises a range of heterogeneous populations with varying degrees of severity. In an effort to reveal functional brain alterations underlying the clinical symptoms of ASC, an increasing number of functional magnetic resonance imaging (fMRI) studies have recently examined intrinsic activity in the absence of externally imposed tasks Such resting-state fMRI (rs-fMRI) studies have found atypical intrinsic connectivity in ASC (for overview, see [2]): for instance, under-connectivity within a default mode network [3,4], and overand under-connectivity between many pairs of brain regions [5,6]. These abnormal functional interactions between distributed brain areas lend support to the hypothesis that brain abnormalities in ASC can best be understood as a network disorder [7]

Methods

Results

Discussion

Conclusion