Similar to the rest of the world (Vos et al., 2017), anxiety disorders are one of the most prevalent types of mental disorders in China (Huang et al., 2019). They are one of the top ten causes of disabilities among children, adolescents, and young adults (Vos et al., 2016). Generalized anxiety disorder (GAD) is a common form of anxiety disorder characterized by excessive anxiety and worry about a number of events or activities (American Psychiatric Association, 2013). It can lead to a considerable reduction in work productivity, and thus bring upon a heavy labor and economic burden on patients, their families, and society (Hoffman et al., 2008). In addition, approximately 30% of patients with GAD do not respond to SSRIs, the first-line treatment of this disease (Baldwin et al., 2011). Therefore, there is an urgent need for a better understanding of the etiology of GAD in order to develop more efficient treatments.

The rapid development of neuroimaging techniques in recent decades has provided novel insights into the onset, development, and outcome of anxiety and other neuropsychiatric disorders. In these neuroimaging studies, patients with GAD consistently exhibit abnormalities in gray matter volume, neural activation, and functional and white matter connectivity centered in the amygdala, cingulate, and prefrontal cortex (Hilbert et al., 2014). In our previous studies, using a seed-based and a region-of-interest-based method to investigate the aberrant connectivity among patients with GAD, we found alterations in task-evoked and resting-state functional connectivity could help differentiate patients with GAD not only from a healthy population but also from individuals with panic disorder, another common kind of anxiety disorders (Cui et al., 2020, 2016; Li et al., 2016). The Seed-based or region-of-interest-based methods comprise the majority of research studies in the field of neuroimaging and focus on determining the most affected regions of the brain in mental disorders. Nevertheless, we also need to look at the brain as a whole, such as using graph theory, since human brains do not work in isolative regions or connectivity (Bullmore and Sporns, 2009).

In recent years, researchers have implemented graph theory in the field of neuroscience and found that the structural and functional systems of healthy brains display features of the complex networks, such as small-worldness (Bullmore and Sporns, 2012). A small-world structure is a cost-effective system for optimal information transfer, commonly found in biological, technological, and social networks, where the majority of the non-neighboring nodes can be connected via hubs that are several steps away (Watts and Strogatz, 1998). Disruption of such an efficient network organization is commonly observed in neuropsychiatric disorders, such as Alzheimer's disease and schizophrenia (Liu et al., 2008; Supekar et al., 2008).

With respect to anxiety disorders, most studies investigate topological alterations using patients with social anxiety disorders, which generated promising findings (Xing et al., 2017; Yang et al., 2019; Yun et al., 2017; Zhu et al., 2017). For GAD, alterations of topological properties were found in structural (Yang et al., 2020) and static functional (Makovac et al., 2018) and dynamic functional (Li et al., 2019) networks. Therefore, the topological properties of brain networks may be promising biomarkers for anxiety disorders including GAD. However, to what extent these abnormalities can facilitate the diagnosis of GAD is yet to be elucidated. In short, the aim of this study was to investigate whether the small-worldness and other topological properties altered in GAD can be used to distinguish patients with GAD from healthy individuals.