Nicotine addiction is a global public health threat and a leading cause of death worldwide (Rasmussen et al., 2017). Cigarette smoking is closely related to cardiac, pulmonary, vascular diseases, and neuropsychiatric disorders (Glantz and Parmley, 1991). Previous studies have shown that chronic cigarette smokers exhibit poorer performance in some cognitive tasks than nonsmokers (Sabia et al., 2012; Wang et al., 2010; Weuve et al., 2012).

Chronic cigarette smoking may result in widespread structural and functional brain impairments (Fedota and Stein, 2015). Previous studies have shown reduced brain activities in regions such as the frontal cortex, cingulate gyrus, insula, thalamus, occipital and temporal lobes, and cerebellum among chronic cigarette smokers (Fedota et al., 2016; Stoeckel et al., 2016). Nicotine addiction is highly correlated with cerebral vascular impairment and inflammation, resulting in myelin degeneration and axonal injury (Hudkins et al., 2012; Savjani et al., 2014). White matter (WM) impairment is the core pathological deficit in the brain among nicotine addicts (Gons et al., 2011). Prior studies have revealed white matter (WM) morphological deficits or microstructural integrity, such as the corpus callosum, cingulum, and frontoparietal fiber tracts, in cases of nicotine addiction (Hudkins et al., 2012; Savjani et al., 2014; Wang et al., 2017). The degree of WM structural injury was closely associated with nicotine addiction severity or duration of habitual smoking (Huang et al., 2017). The level of WM structural deficits also reflects the degree of cognitive impairment in smokers, which is one of the core impairments in chronic smokers, and could be applied to predict nicotine cessation outcomes (Bi et al., 2017; Gons et al., 2011; Huang et al., 2017).

Given that white matter densely connects different regions of the gray matter and accounts for nearly half of the brain (Harris and Attwell, 2012), recent fMRI observations have demonstrated that brain WM also contains neural signal information rather than noise, as conventional studies usually assume (Ding et al., 2018, 2013, 2016; Ji et al., 2017; Li et al., 2020a; Marussich et al., 2017; Thompson et al., 2016). WM shares a similar BOLD pattern with GM signal changes correlated with hemodynamic responses to stimuli (Thompson et al., 2016), and the active mapping of WM function was comparable to task-related WM fiber pathways detected by DTI (Ding et al., 2016; Marussich et al., 2017). Researchers also found that the WM functional brain activity strength was closely correlated with participants’ cognitive task performance, while the WM structure remained unchanged, which indicates that the WM function may reflect brain function changes before structural alterations (Ding et al., 2016; Marussich et al., 2017). Researchers have also explored whether the WM function abnormalities could reflect the pathological alterations in neuropsychiatric disorders (Ji et al., 2019; Jiang et al., 2019a, 2019b; Makedonov et al., 2016). Resting fMRI showed that the WM functional activity was correlated with memory ability in Alzheimer's disease (Makedonov et al., 2016). Regional-level and network-level of WM function were both impaired in Parkinson's disease or in schizophrenia (Ji et al., 2019; Jiang et al., 2019a). The disturbance in WM functional network can reflect disease severity and cognitive impairment degree in a series of neuropsychiatric disorders, such as Alzheimer's disease, epilepsy, Parkinson's disease, major depression, and schizophrenia (Ji et al., 2019; Jiang et al., 2019a, 2019b; Li et al., 2020b; Makedonov et al., 2016). Moreover, in addition to the WM functional network showing similar abnormalities to the GM functional network, the WM functional network could reveal additional aberrations in the early stage of the disease in some cases (Jiang et al., 2022). Therefore, the WM functional features might provide new insights into the brain observation, and could be further used as clinical biomarkers in neuropsychiatric disorders.

The small-world network model is a widely accepted model that reflects the human brain network, comprising two fundamental principles of the brain network: information segregation and integration processing (Avena-Koenigsberger et al., 2017; Bassett and Bullmore, 2017). Abnormal small-world topological properties for the structural network or the functional network were widely presented in neuropsychiatric disorders (Borsboom et al., 2011; Hallquist and Hillary, 2019).The small-world topology refers to the network topological properties exhibit prominent small-world characteristics which demonstrate efficient information segregation and integration at low wiring and energy costs (Rubinov and Sporns, 2010; Li et al., 2020b). Alterations in small-world topology usually reflect the severity of the disease and have been considered promising clinical biomarkers (Borsboom et al., 2011; Hallquist and Hillary, 2019). Li et al. applied graph theory analysis to WM resting functional data from a large sample and identified that the whole-brain WM functional connectivity exhibited reliable and stable small-world architecture (Li et al., 2019a). Subsequent studies determined that the small-world topological properties of WM functional connectivity offer novel applicable neuromarkers for general fluid intelligence in healthy individuals (Li et al., 2020a). Abnormal small-world properties in WM functional connectivity were also detected in neuropsychiatric disorders such as depression or Parkinson's disease, and alterations in topological properties in WM functional connectivity could be applied as biomarkers for clinical diagnosis (Ji et al., 2019; Li et al., 2020b). Nicotine addiction has been considered an alteration of the whole brain network, which could be detected using the small-world model (Fedota and Stein, 2015). FMRI studies have shown that the GM functional network of chronic smokers revealed abnormal small-world topological features, which could be used to detect the severity of nicotine addiction, and DTI studies also showed similar structural small-world topology alterations of WM in chronic smokers (Lin et al., 2015; Tan et al., 2019; Zhang et al., 2018).

To the best of our knowledge, the small-world topological properties for WM function remains unknown in nicotine addiction patients. As mentioned above, white matter injury is a very common and critical deficit in nicotine addicts (Gons et al., 2011), WM functional networks play an important role in human function, and the small-world properties have been considered promising biomarkers in neuropsychiatric disorders (Ding et al., 2018, 2013, 2016; Gons et al., 2011; Ji et al., 2017; Marussich et al., 2017; Thompson et al., 2016). In addition, the WM function may demonstrate alterations before the WM structure has changed (Ding et al., 2016; Marussich et al., 2017), and the WM function may even provide additional abnormalities in disease compared with the GM function (Jiang et al., 2022). Therefore, in this study, we investigated the WM functional network alterations in nicotine addiction using small-world network analysis. We aimed to comprehensively detect the topological organization of WM function in nicotine addicts and to investigate whether the small-world topological properties of the WM functional network could be qualified as promising biomarkers for nicotine addiction. To be specific, we will investigate the following four points: (i) utilizing graph theoretical analysis to reveal the topological features of the WM functional network, (ii) identifying alterations in small-world properties in smokers, (iii) exploiting the clinical applications of small-world properties to differentiate smokers from healthy controls, and (iv) applying the altered small-world properties to nicotine addiction severity analysis.