Network analysis of depression, anxiety, and NSSI symptoms in adolescents: a clinical sample study Authors: Xiaoyu Jiang, Zhikun Zhang, Ziyi Wei, Guangde Shi, Miaoyu Yu, Junduan Wu Published: 2026 ABSTRACT AbstractObjectiveThis study aimed to construct a symptom network model of depression, anxiety, and NSSI using network analysis and to identify the most central and bridge symptoms within the depression-anxiety-NSSI symptom network.MethodsA total of 200 adolescent patients with depression were recruited from hospitals and assessed using the Patient Health Questionnaire-9 (PHQ-9), Generalized Anxiety Disorder Scale-7 (GAD-7), and the Modified version of the Adolescent Self-Harm Scale (MASHS). The structural characteristics of depression, anxiety, and NSSI symptoms were characterized using “strength” and “bridge strength” as centrality indices within the symptom network. Network stability was examined using a case-dropping bootstrap procedure.ResultsThe nodes GAD2 (uncontrollable worry), GAD3 (excessive worry), GAD5 (restlessness), and PHQ9 (suicidal ideation) were the most central symptoms in the network model. Furthermore, within this sample, the symptoms PHQ8 (psychomotor agitation/retardation) and GAD6 (irritability) served as bridge symptoms connecting the anxiety and depressive symptom clusters with the NSSI symptom cluster.ConclusionThe central symptoms (such as uncontrollable worry, excessive worry, restlessness, and suicidal ideation) and key bridge symptoms (such as psychomotor agitation/retardation and irritability) within the depression-anxiety-NSSI symptom network may serve as potential intervention targets for adolescent patients with depression.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12888-026-08123-9.Keywords: Adolescent depression, Anxiety, Network analysis, NSSI Introduction Non-suicidal self-injury (NSSI) is defined as the deliberate, repeated destruction of body tissue without suicidal intent, with common forms including cutting, burning, and excessive scratching of the skin [1]. Epidemiological data indicate that adolescence marks the peak period for the initial onset of NSSI, with the average age of onset being 12–13 years [2]. During this stage, neuroendocrine remodeling, asynchronous development of the prefrontal-limbic system, and a significant increase in socio-emotional sensitivity render adolescents highly vulnerable to neuro-behavioral susceptibility in the face of negative experiences [3]. In China, the highly competitive educational environment and “involution” culture [4, 5], expose adolescents to multiple chronic psychological stressors, contributing to a continuous rise in NSSI detection rates. A recent meta-analysis reported that the overall prevalence rates of NSSI among Chinese middle school and high school students are 25.3% and 32.8%, respectively [6], significantly higher than the global average (17.7%) [7]. Although NSSI is not performed with the intent to die, multiple longitudinal studies have consistently demonstrated that this behavior significantly predicts subsequent suicidal ideation, suicide attempts, and completed suicide [8, 9]. Therefore, elucidating the mechanisms underlying NSSI in adolescents holds significant clinical and public health value for the early identification of high-risk individuals and the implementation of targeted interventions. The functional theory of non-suicidal self-injury (NSSI) posits that self-injury is often used by individuals to regulate emotions, manage interpersonal relationships, or inflict self-punishment, with the emotion regulation function being the most prominent. Following self-injury, negative emotions decrease rapidly, and individuals experience a sense of “relief” [10]. The Experiential Avoidance Model (EAM) further suggests that such “relief” is essentially a form of negative reinforcement, whereby the avoidance or escape from undesirable emotional experiences perpetuates NSSI behavior [11]. Extensive research has confirmed a strong association between NSSI and anxiety and depression [12, 13]. Some studies indicate that depression and anxiety are the strongest predictors of NSSI among adolescents [14], with 34.2% of individuals with depression having engaged in NSSI [15]. Anxiety symptoms are also significantly positively correlated with NSSI [16, 17]. However, longitudinal studies suggest a reverse pathway, indicating that NSSI may also induce or exacerbate anxiety and depression [18, 19]. Previous research has shown that adolescents with a history of NSSI often perceive their physical scars as “marks of shame or disgrace” and develop negative cognitive patterns, which may further aggravate their emotional disorders [20, 21]. In summary, the relationship among NSSI, anxiety, and depression is not unidirectional but likely constitutes a complex feedback loop of mutual causation. Traditional variable-centered approaches are inadequate for elucidating these dynamic mechanisms, necessitating the application of novel research methods to clarify them. Network analysis is increasingly being applied in the field of psychopathology [22]. This approach constructs dynamic interactive symptom networks by representing observed variables as nodes and connecting their statistical relationships with edges. Traditional models of depression often view depressive symptoms (e.g., low mood, sleep disturbances, suicidal ideation) as a collection of manifestations driven by a common underlying etiology [23]. In contrast, network theory proposes that symptoms are not merely co-occurring but form a causal system of mutual activation and dynamic evolution, where the presence of certain individual symptoms may influence the development of others [24]. For example, low mood may be associated with insomnia, which in turn may contribute to fatigue and impaired cognitive function, forming a self-sustaining vicious cycle. Similarly, improvements in one symptom may be linked to reductions in the severity of others [25]. By quantifying conditional associations among symptoms and mapping the topological structure, network analysis can identify central symptoms that exert a prominent influence on others within the network [26, 27]. Furthermore, by identifying “bridge symptoms,” network models help elucidate comorbidity mechanisms between different mental disorders at the symptom level [28]. These central and bridge symptoms play a critical role in the onset and maintenance of mental disorders [29]. Theoretically, targeting these central nodes in interventions may more effectively disrupt symptom cascades, holding significant promise for clinical practice [26]. Currently, network analysis has been widely employed in studies of depressive and anxiety symptoms across diverse populations, including adolescents [30, 31], medical students [32], patients with chronic obstructive pulmonary disease [33], and older adults with disabilities [34]. Network analysis has also been extensively used to explore the relationships between NSSI, anxiety, and depression [35–37]. However, the symptom dynamics identified in these samples may not directly apply to clinically diagnosed adolescents. In general adolescent populations, NSSI is often associated with factors such as impulsivity and bullying [38, 39]; in clinically depressed adolescents, by contrast, NSSI is frequently closely intertwined with core depressive symptoms—such as anhedonia, self-criticism, and hopelessness [40, 41]. Identifying symptom networks specific to this high-risk population is therefore essential for developing targeted interventions. Despite this, few studies have applied this integrated approach specifically to clinically diagnosed adolescent depression populations. To address this gap, the present study employs network analysis to explore the network structure of depressive, anxiety, and NSSI symptoms in clinically diagnosed adolescents with depression, with a focus on identifying both the most central symptoms and the bridge symptoms that connect NSSI with depressive and anxiety symptoms. Methods Sample size estimation and participant recruitment A total of 200 adolescent patients with depression were included in this study. Using a convenience sampling method, participants were recruited from the inpatient and outpatient departments of the Second Affiliated Hospital of Guangxi Medical University between September 2024 and March 2025. A total of 223 questionnaires were collected, and after excluding invalid responses, 200 valid questionnaires were retained for analysis. The study protocol was approved by the Ethics Review Committee of the Second Affiliated Hospital of Guangxi Medical University (Approval No. KY-0632). All participants received detailed information about the study’s aims and procedures before enrollment, and each provided written informed consent. The investigation was conducted in accordance with the principles of the Declaration of Helsinki. The inclusion criteria for study participants were: (1) diagnosis of major depressive disorder, current episode, confirmed independently by two senior psychiatrists according to the DSM-5 criteria; (2) aged 12–18 years; and (3) capability to read, understand, and complete the questionnaires. Exclusion criteria were: (1) secondary depression due to organic diseases or medication; (2) presence of severe physical illnesses, serious infections, organic brain diseases, or a history of alcohol or other substance dependence or abuse; (3) presence of bipolar spectrum features and (4) during clinical interview, a questionnaire omission rate greater than 10%. The sample size was estimated based on requirements for network analysis [42]. The minimum sample size should exceed the total number of parameters, calculated as the sum of threshold parameters (number of nodes) and pairwise association parameters [number of nodes × (number of nodes − 1)/2]. This study included 9 nodes for depressive symptoms, 7 for anxiety symptoms, and 1 for the total NSSI score, resulting in 17 nodes and a minimum required sample size of 153. The final sample of 200 met this requirement. Measures The 9-item patient health questionnaire (PHQ-9) The Patient Health Questionnaire-9 (PHQ-9) was used to assess the severity of depressive symptoms in adolescents over the past two weeks [43]. This scale consists of nine items corresponding to the core symptoms of depression, including anhedonia, sad mood, sleep disturbances, fatigue, appetite changes, guilty, concentration difficulties, psychomotor retardation/agitation, and suicidal ideation. Responses are rated on a 4-point Likert scale ranging from 0 (“not at all”) to 3 (“nearly every day”). Total scores range from 0 to 27, with higher scores indicating greater severity of depressive symptoms. In this study, the Cronbach’s alpha for the PHQ-9 was 0.907. The 7-item generalized anxiety disorder scale (GAD-7) The Generalized Anxiety Disorder Scale (GAD-7) is a self-report measure developed by Spitzer et al. to assess anxiety symptoms [44]. It comprises seven items rated on a 4-point frequency scale from 0 (“not at all”) to 3 (“nearly every day”). Total scores range from 0 to 21. This scale consists of seven items corresponding to the core symptoms of anxiety symptoms, including nervousness, uncontrollable worry, excessive worry, trouble relaxing, restlessness, irritability, and feeling afraid. The internal consistency of the GAD-7 in this sample was excellent, with Cronbach’s alpha of 0.929. The modified version of the adolescent self-harm scale (MASHS) Non-suicidal self-injury (NSSI) was assessed using the Modified Version of the Adolescent Self-Harm Scale (MASHS). The scale was originally developed by Zheng [45], and later revised by Feng [46]. It includes 18 closed items and one open-ended question. The frequency of each NSSI behavior is rated on a 4-point scale: 0 (never), 1 (once), 2 (2–4 times), and 3 (≥ 5 times). The physical severity of each behavior is rated on a 5-point scale from 0 (“none”) to 4 (“extreme”). A total NSSI score is calculated by summing the products of the frequency and severity ratings for all items, with higher scores indicating more severe NSSI behavior. In the present study, the scale demonstrated good internal consistency, with a Cronbach’s alpha of 0.879. Statistical analysis Data were entered and a database was established using EpiData 3.1 software. Descriptive analyses were performed using SPSS 25.0. Categorical data were described using frequencies and constituent ratios (n, %), while continuous data were presented as mean ± standard deviation (M ± SD). Network analysis was conducted using R software version 4.4.1. The analytical procedure involved the following steps: first, item scores from the PHQ-9 and GAD-7, as well as the total NSSI score, were standardized. Subsequently, a Gaussian Graphical Model (GGM) was estimated using the EBICglasso function from the qgraph package [42]. This method integrates the Graphical LASSO with the Extended Bayesian Information Criterion (EBIC) to automatically select the optimal regularization parameter, thereby yielding a stable network structure. In the resulting network, nodes represent individual items of the PHQ-9 and GAD-7, as well as the total NSSI score. Edges between nodes represent partial correlations [30], with blue lines indicating positive correlation and red lines indicating negative correlation. Thicker and darker lines represent stronger associations. Given their relative stability, the strength centrality was used to quantify the importance of each node within the network, while the bridge strength was used to identify key nodes connecting different symptom communities. Finally, the accuracy and stability of the estimated network were assessed using the bootnet package. The accuracy of edge weights was evaluated by computing 95% confidence intervals (CIs) via a non-parametric bootstrap with 1000 iterations. Narrower CIs indicate more reliable edge weight estimates [42]. The stability of the network structure was assessed using the correlation stability coefficient (CS-coefficient), with a value of CS≥0.25 considered acceptable [42]. All statistical tests were two-sided, and a P<0.05 at a significance level of α = 0.05 was considered statistically significant. Results Study sample A total of 200 adolescent patients with depression were included in this study. The cohort comprised 39 boys and 161 girls, with a mean age of 16.02 ± 1.74 years and a mean body mass index (BMI) of 20.00 ± 4.09. Detailed characteristics are presented in Table 1. Table 1Demographic Characteristics of Study Participants (n = 200)Variablen(%)/M ± SDGenderboys39(19.5)girls161(80.5)Age (years)16.02 ± 1.74BMI20.00 ± 4.09ResidenceUrban103(51.5)Rural97(48.5)EthnicityHan101(50.5)Ethnic Minority99(49.5)Only ChildYes36(18)No164(82)Family Economic StatusPoor44(22)Average135(67.5)Good21(10.5)Father’s EducationPrimary or below37(18.5)Junior high81(40.5)Senior high37(18.5)College or above45(22.5)Mother’s EducationPrimary or below48(24)Junior high79(39.5)Senior high35(17.5)College or above38(19) Demographic Characteristics of Study Participants (n = 200) Network structure The symptom network of depression, anxiety, and NSSI among adolescent patients with depression is depicted in Fig. 1. The network exhibited strong interconnections, forming a tightly integrated structure across the three symptom domains. Within the depressive symptom community, the strongest edge weight was observed between PHQ6 (guilty) and PHQ9 (suicidal ideation) (r = 0.28, P<0.001). Among anxiety symptoms, the strongest connection was between GAD2 (uncontrollable worry) and GAD3 (excessive worry) (r = 0.46, P<0.001). The most robust connection between depressive and anxiety symptoms was between PHQ8 (psychomotor agitation/retardation) and GAD5 (restlessness) (r = 0.29, P<0.001). The strongest edges linking NSSI with depressive and anxiety symptoms were NSSI-PHQ9 (NSSI-suicidal ideation) and NSSI-GAD6 (NSSI-irritability), with edge weights of r = 0.12 (P<0.05) and r = 0.14 (P<0.05), respectively. Complete edge weight information is provided in Supplementary Material Table S1. Network Structure of Depression, Anxiety, and NSSI Symptoms in Adolescents with Depression Centrality and Bridge Centrality In network analysis, nodes with the highest strength values are typically considered the most influential [47]. As illustrated in Fig. 2a, GAD2, GAD3, GAD5, and PHQ9 emerged as the strongest nodes, forming a central cluster. Bootstrapped difference tests confirmed that these four nodes had significantly higher strength than peripheral nodes such as NSSI, PHQ1, PHQ3, and PHQ5 (see Supplementary Figure S1). However, no significant differences in strength were observed among the four central nodes themselves, and their strength was comparable to that of moderate-strength nodes like GAD4 and PHQ6. Network node strength and bridge strength. (a) Strength centrality of the depression, anxiety, and NSSI symptom network (nodes are ranked by strength; higher-strength nodes indicate greater importance in the network). (b) Bridge strength centrality of the depression, anxiety, and NSSI symptom network (nodes are ranked by bridge strength; higher bridge strength nodes serve as key connectors between symptom clusters) Bridge centrality identifies symptoms that may facilitate the activation spread between distinct symptom clusters [28]. For bridge strength, PHQ8 and GAD6 were identified as key bridge symptoms (Fig. 2b). The difference between them was not significant, but both exhibited significantly higher bridge strength than most other nodes (Supplementary Figure S2). Network stability and accuracy The stability of the network structure was evaluated using the correlation stability coefficient (CS-coefficient), with values ≥ 0.25 indicating acceptable stability [42]. The network demonstrated robust stability for both strength (CS = 0.52) and bridge strength (CS = 0.44) centrality, indicating that the relative order of node centrality remained highly consistent after randomly dropping up to 52% and 44% of cases, respectively (Fig. 3). Bootstrap validation of network structure stability using sampling without replacement However, as shown in Supplementary Figure S3, bootstrapped confidence intervals for edge weights were relatively wide for many connections, suggesting limited precision in estimating individual edge weights. Taken together, these findings indicate that while specific edge weights should be interpreted cautiously, the network exhibits adequate overall stability, with CS-coefficients exceeding the recommended threshold [42]. This supports the reliability of centrality-based interpretations. Discussion This study aimed to explore the network relationships among depression, anxiety, and NSSI symptoms in adolescents with depression. Based on a combination of node strength/bridge strength values and bootstrap-based difference tests, GAD2 (uncontrollable worry), GAD3 (excessive worry), GAD5 (restlessness), and PHQ9 (suicidal ideation) emerged as particularly central and were identified as the core nodes within the network. Additionally, PHQ8 (psychomotor agitation/retardation) and GAD6 (irritability) were identified as bridge symptoms connecting the anxiety and depressive symptom communities with the NSSI symptom community. “Uncontrollable worry” (GAD2) was identified as one of the core symptoms within the network of depressive, anxiety, and NSSI symptoms among adolescent patients with depression. Similar findings have been reported in previous studies involving different populations [32, 34]. According to the cognitive avoidance theory of generalized anxiety disorder, worrying may function as an emotion regulation strategy driven by an individual’s inability to tolerate negative emotional states—essentially using “persistent worrying” to avoid distressing emotional experiences [48]. In the context of adolescent depression, this maladaptive strategy may exacerbate feelings of hopelessness and sadness, potentially leading to the use of NSSI as an attempt to alleviate such uncontrollable anxiety [49]. Cognitive Behavioral Therapy (CBT) has been demonstrated to effectively improve emotion regulation capacity and help individuals better cope with a sense of loss of control. For instance, metacognitive therapy can be applied to shift adolescents’ focus from the content of their worries to their thoughts about worrying itself, thereby altering their relationship with persistent worry [50]. “Excessive worry” (GAD3) also exhibited high centrality within the network of depressive, anxiety, and NSSI symptoms among adolescents, establishing it as a key central symptom in the network. This finding aligns with previous network studies focusing on depression and anxiety in Chinese adolescent populations [31]. Within the Chinese cultural context, “excessive worry” in adolescents is frequently associated with factors such as academic pressure and gradually generalizes into persistent concerns spanning multiple domains including current academic performance, future educational prospects, employment opportunities, and anticipated income [51]. Furthermore, adolescents experiencing “excessive worry” often present with comorbid sleep disturbances [52]. Insomnia itself can function as a stressor that impairs neuroplasticity [53], pushing individuals into a pathological state of overload and inducing anxiety-like effects [54]. This condition further disrupts emotional regulation and decision-making functions, resulting in diminished impulse control, increased aggressive behaviors, and heightened risk-taking tendencies [55]. Consequently, “excessive worry” likely serves as a critical hub connecting the depressive, anxiety, and NSSI symptom networks. A comprehensive multi-system intervention approach—incorporating regular stress management activities in schools [56], emotional and material support from parents [57], and connections with the medical system (e.g., timely screening and intervention) —can effectively alleviate excessive worry in adolescents [58]. “Suicidal ideation” (PHQ9) and “restlessness” (GAD5) both occupied important positions within the depression-anxiety symptom network, consistent with previous research [37, 50]. “Negative thoughts,” encompassing self-harm and suicidal ideation, function as a cognitive core that may be driven by feelings of hopelessness and depressed mood, subsequently contributing to the occurrence of NSSI. Meanwhile, “restlessness,” as a core somatic manifestation of anxiety, externalizes and intensifies inner tension. Providing adolescents with effective coping strategies is crucial; for instance, self-compassion techniques can help them manage distressing emotions and promote psychological well-being [59]. We identified nodes PHQ8 (psychomotor agitation/retardation) and GAD6 (irritability) as key bridge symptoms in the current network. “Psychomotor agitation/retardation” (PHQ8), one of the most recognizable psychomotor symptoms in adolescent depression and consistent with prior studies [18], emerged as a significant bridge connecting depression and NSSI. Although these manifestations represent opposite phenotypes—retardation characterized by slowed speech and lethargic gait, while agitation manifests as restlessness, skin picking, or frequent position changes—they can coexist within the same patient [60]. Recent fMRI findings indicate that depressed patients presenting with retardation show heightened thalamocortical connectivity, whereas those with agitation primarily exhibit increased pallidal-cortical connectivity [61]. Furthermore, Kang et al. discovered that patients with major depressive disorder (MDD) and a history of NSSI demonstrate abnormalities in emotional circuits, sensory systems, and pain modulation regions, including the putamen, orbitofrontal cortex, globus pallidus, olfactory epithelium, amygdala, insula, and lingual gyrus [62]. These regions notably overlap with the pallidal-cortical psychomotor circuit involved in agitated movement, suggesting that the pallidal-cortex pathway may underlie the connection between agitated motor phenotypes and self-injurious impulses—a hypothesis warranting further validation. “Irritability” (GAD6) served as another critical bridge connecting anxiety and NSSI, aligning with findings from Liu et al. [63]. Irritability represents one of the most common presenting issues in child and adolescent psychiatric settings [64, 65]. It is defined as “a heightened tendency to react with anger compared to peers” [66]. Behavioral experiments indicate that such adolescents exhibit excessive responses to both obtaining and missing rewards, with a potential neural basis in dysregulation of the prefrontal-striatal-amygdala circuit [66]. Among adolescents, anxiety does not always manifest as pure tension or fear but may present as “irritability” [67]. Irritability may increase the risk of impulsive NSSI [68]. Meta-analysis evidence indicates that difficulties in impulse control and insufficient adaptive strategies are the key mechanisms by which emotional dysregulation is associated with self-harm [69]. Previous research also suggests that interventions targeting impulsivity should focus on disrupting the emotion-action transition, such as implementing measures to enhance cognitive flexibility in adolescents [39]. Therefore, parents, educators, and clinicians should prioritize “irritability” as a key intervention target. This study has several limitations. First, regarding measurement: the use of a total NSSI score and single-item assessments may have oversimplified constructs and introduced measurement ambiguity. Second, regarding sample: the single-region sample (N = 200) limits generalizability, and while meeting minimum technical requirements, the sample size may not fully offset clinical heterogeneity and measurement—reducing power to detect subtle meaningful connections. It should also be noted that our sample had a high prevalence of NSSI, suggesting a clinically severe subgroup; thus, findings may not be fully generalizable to less severe outpatient populations. Third, regarding methodology: self-report measures are subject to recall and social desirability biases; the cross-sectional design precludes causal inference; and the absence of a semi-structured diagnostic interview (e.g., K-SADS-PL) may affect diagnostic reliability. Furthermore, information on participants’ medication status was not collected. This limits our ability to differentiate between primary depressive symptoms and potential medication side effects (e.g., restlessness as a result of antidepressant use), introducing potential confounding in the interpretation of specific nodes. Fourth, regarding comorbidities: although bipolar features were excluded, ADHD and borderline personality traits were not assessed, and their potential confounding effects cannot be ruled out. Fifth, regarding contextual factors: important environmental variables known to contribute to NSSI—such as childhood adversity and family environment—were not included in the analysis. These omitted factors may influence the symptom network. Future studies should employ larger, multi-site samples; use multi-item assessments and specific NSSI behaviors as nodes; adopt longitudinal designs; systematically evaluate common comorbidities; incorporate contextual factors; and collect detailed medication and treatment phase information to provide a more comprehensive understanding of the clinical picture. In summary, this study identified central symptoms (e.g., “inability to control worrying,” “excessive worry,” “restlessness,” and “suicidal ideation”) and key bridge symptoms (e.g., “psychomotor agitation/retardation” and “irritability”) within the depression-anxiety-NSSI network among Chinese adolescent patients with depression. These findings may inform effective targets for preventive and interventional measures in clinical populations of adolescents with depression. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements We would like to express my sincere gratitude to Dr. Chen Xianbin, Dr. Liang Qi and Dr. Wang Chunyan for their assistance. In addition, I would like to sincerely thank the participants in this study and their parents or guardiansts. Abbreviations Body mass index Cognitive Behavioral Therapy Correlation stability coefficient Experiential Avoidance Model Extended Bayesian Information Criterion Generalized Anxiety Disorder Scale-7 Gaussian Graphical Model Modified version of the Adolescents Self-Harm Scale Major depressive disorder Non-suicidal self-injury Patient Health Questionnaire-9 Author contributions WJD, YMY and JXY conceived of the study, participated in its design and coordination; JXY, WZY, ZZK and SGD performed the measurement; JXY performed the statistical analysis; JXY and WJD drafted the manuscript; All authors read and approved the final manuscript. Funding This work was supported by the Natural Science Foundation of Guangxi Autonomous Region [Grant numbers: 2022GXNSFDA035077]. Data availability All data supporting the findings of this study are available within the paper and its Supplementary Information, see Paper date. Declarations Ethics approval and consent to participate The study was approved by the Ethics Review Committee of the Second Affiliated Hospital of Guangxi Medical University (Approval No. KY-0632) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from participants aged 16–18 years, and from parents or legal guardians for those aged 12–15 years, together with written assent from the minors. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Miaoyu Yu, Email: yuliuyu01@sina.com. Junduan Wu, Email: wujunduan@gxmu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement All data supporting the findings of this study are available within the paper and its Supplementary Information, see Paper date.