Normative Modeling Reveals Functional Connectivity Heterogeneity and Subtypes in Internet Gaming Disorder

Normative Modeling Reveals Functional Connectivity Heterogeneity and Subtypes in Internet Gaming Disorder | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Normative Modeling Reveals Functional Connectivity Heterogeneity and Subtypes in Internet Gaming Disorder Zixiao Wang, Yong Xie, Yidan Wang, Lingxiao Wang, Guang-Heng Dong This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6938965/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 May, 2026 Read the published version in Annals of General Psychiatry → Version 1 posted 9 You are reading this latest preprint version Abstract Background Internet gaming disorder (IGD) is associated with abnormal functional connectivity (FC) in brain networks. However, findings from resting-state fMRI (RS-fMRI) studies are highly inconsistent, likely due to individual heterogeneity in IGD-related neural alterations—a feature commonly observed in other psychiatric disorders but understudied in IGD. Methods We applied normative modeling to RS-fMRI data to characterize individualized FC deviations between the nucleus accumbens (NAcc) and the rest of the brain in 173 IGD participants. Using a large sample of healthy controls (N = 232) to establish age- and sex-adjusted normative ranges, we quantified the abnormality of each IGD individual’s FC profile. Further, we performed k-means clustering on deviation values within the frontoparietal network (FPN) and basal ganglia network (BGN) to identify neurobiological subtypes. Results Our findings revealed considerable heterogeneity in FC abnormalities among IGD participants. Only a minority showed strong and widespread deviations, while most exhibited mild but diverse patterns. Clustering analysis identified five distinct IGD subtypes with varying hyper- and hypoconnectivity profiles in FPN and BGN. Conclusions These results highlight the heterogeneous neural basis of IGD and underscore the limitations of group-level comparisons. Normative modeling and FC-based subtyping offer a promising direction for individualized assessment and may inform personalized interventions, such as connectivity- and frequency-specific TMS targeting cognitive control and reward circuits. Internet gaming disorder functional connectivity heterogeneity normative modeling subtypes Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Background Internet gaming disorder (IGD) refers to the problematic and excessive use of internet games, which can be detrimental to various aspects of one's education and life ( 1 , 2 ). American psychiatric association (APA) has included IGD in Section III of fifth edition of diagnostic and statistical manual of mental disorders (DSM-5) as warranting more clinical research in 2013 ( 3 ). Using resting-state functional magnetic resonance imaging (RS-fMRI), researchers have conducted numerous studies on the characteristics and underlying neural basis of IGD, demonstrating some core neural variation of IGD, such as abnormal brain functional networks related to cognitive control, reward processing ( 4 – 7 ). However, after summarizing previous RS-fMRI studies on IGD, we found that it is hard to draw a uniform conclusion about the underlying neural mechanism of IGD. For example, some researchers found that compared to healthy control (HC) group, IGD group displayed decreased functional connectivity (FC) between the dorsolateral prefrontal cortex (DLPFC) and nucleus accumbens (NAcc) ( 8 ), decreased FC between the orbitofrontal cortex (OFC) and putamen ( 9 ), and decreased FC between the anterior cingulate cortex (ACC) and caudate ( 10 ). On the other hand, some researchers identified increased FC between the OFC and putamen ( 11 ), increased FC between the DLPFC and posterior cingulate cortex (PCC) ( 12 ), and increased FC between the middle frontal gyrus (MFG) and caudate ( 13 ) in IGD group in comparison with HC group. Both the direction and the focal area of abnormalities of IGD group exhibited different across different RS-fMRI studies. We inferred that the heterogeneity of IGD in the abnormalities of resting-state FC patterns is the main reason for these inconsistent results. Heterogeneity is one of core features of many disorders, including schizophrenia, autism, bipolar disorder and substance addiction ( 14 – 18 ). However, the heterogeneity of IGD is barely understood. The heterogeneity of IGD emphasizes the importance of identifying individualized abnormality of IGD, thus contributing to prompt the individualized treatment for IGD. In contrast, group-level differences from comparisons between IGD and healthy groups in previous studies might not necessarily be representative of any IGD gamer. They only represent a small part of the neural abnormalities that characterize IGD and the bulk of the small part of abnormalities are comprised of individual IGD who are highly deviated from the norm ( 19 ). Accordingly, the present study was set to map the individualized abnormalities of brain FC in IGD using a recent methodology named normative modeling. Normative modeling has been well documented and applied in studies of heterogeneity in other psychiatric disorders, such as autism, schizophrenia, depressive and bipolar disorder ( 14 , 16 , 20 , 21 ). A normative model can be understood as a statistical model that maps demographic or behavioral variables to a quantitative brain readout. Using brain data from a large number of typically developed individuals (i.e., HC individuals), a Bayesian process regression model was used to calculate the normative range within which brain metrics could be varied up or down for HC individuals at specific ages and sexes ( 22 ). Then, each IGD individual brain metric is compared to the normative model of same age and sex to compute their individualized abnormalities/deviations relative to HC individuals. This is the first goal of the present study. We assumed that different IGD individual would exhibit different abnormalities/deviations from normative model, i.e., heterogeneity, in resting-state brain FC, as we summarized from previous RS-fMRI studies. Additionally, the bilateral NAcc was selected as the region of interest (ROI) for the resting-state ROI-based brain FC analysis to compute the FC between the NAcc and the whole remaining brain regions in the current study. The NAcc, the main projection region of the mesolimbic pathway, has been confirmed to be involved in driving craving and development and maintain of addiction in many addicted types, including IGD ( 23 – 25 ). Moreover, many studies have revealed abnormal activation/FC of the NAcc in individuals with IGD ( 8 , 26 – 28 ). The second goal of the present study is to distinguish different subtypes of IGD using clustering algorithm. Clustering algorithm is also a common approach to biological heterogeneity. This approach is successful to some extent and is appropriate if the clinical cohort can be cleanly divided into a relatively small number of homogeneous subgroups based on the measures chosen ( 20 , 29 ). Accordingly, by using clustering algorithm, we expect to identify some homogeneous subtypes of IGD individuals based on their direction and magnitude of deviations from normative models regarding FC. Particularly, according to previous studies on IGD, the cognitive control and reward networks were most related to IGD ( 4 , 26 , 28 , 30 , 31 ). Thus, we aimed to identify the subtypes of IGD participants based on their abnormalities/deviations from normative model in the cognitive control and reward networks. In the current study, we used the frontoparietal network (FPN) and the basal ganglia network (BGN) from Shen’s brain functional atlas ( 32 ), which were implicated in the cognitive control and reward networks separately ( 33 – 35 ). Methods Participants The present study included 173 university students with IGD (males/females: 98/75) and 232 HCs (males/females: 139/93). The numbers of males and females were matched between the two groups (χ 2 =0.436, p =0.509). All the participants were free of any psychiatric/neurological disorders confirmed by structured psychiatric interviews (Mini International Neuropsychiatric Interview). No participant reported any illegal drug use and gambling, and excessive nicotine and alcohol uses. The diagnosis of IGD was based on the Young's Internet Addiction Test (IAT) (36) and DSM-5 criteria for IGD (37). Consistent with previous studies of IGD (4, 38-40), the diagnostic criteria for the IGD group included: 1) an IAT score >50; 2) a DSM-5 criteria score >5; and 3) spending most of their internet time in playing games. Participants with DSM-5 criteria scores <5 and IAT scores <50 were classified as the HC group. As shown in Table 1, the IGD group reported significant greater IAT scores and DSM-5 scores than the HC group. Table 1. Demographic information and group differences Items IGD (M=98, F=75) HC (M=139, F=93) t p Age (years) 21.13±2.33 21.53±2.43 -1.663 0.097 IAT score 66.15±9.13 40.32±10.92 25.223 <0.001 DSM-5 score 6.08±1.13 2.51±1.37 26.447 <0.001 Table values: mean ± standard deviation Abbreviations: IGD, internet gaming disorder; HC, healthy control; M, male; F, female; IAT, internet addiction test; DSM-5, the fifth edition of diagnostic and statistical manual of mental disorders. RS-fMRI data acquisition All participants underwent a RS-fMRI scan for 7 min using a 3 T scanner (Siemens Trio, Malvern, PA, USA) equipped for echo-planar imaging (EPI). During the scan, they were asked to open eyes and stay still and relaxed. The scan parameters were as follows: repetition time (TR) = 2000 ms, echo time (TE) = 30 ms, slice number = 33, interleaved sequence, slice thickness = 3 mm, voxel size = 3 × 3 × 3 mm 3 , field of view (FOV) = 220 × 220 mm 2 , flip angle = 90 ◦ and matrix of 64 × 64. Preprocessing of RS-fMRI data Data preprocessing was performed using DPABI V8.2 (http://rfmri.org/dpabi). The preprocessing steps included: 1) discarding the first 10 time points; 2) slice-timing; 3) realign for head-motion correction; 4) spatial normalization to Montreal Neurological Institute (MNI) standard space; 5) spatial smoothing (FWHM = 6 mm); 6) linear trend removal; 7) nuisance covariates regression, including head-motion covariates using the Friston 24-parameter model as well as signals from white matter, cerebrospinal fluid and global signals; 8) band-pass filtering with a range of 0.01-0.1 Hz. Constructing normative models of FC An overview of the normative modeling approach is provided in Figure 1; this approach has been described previously (22). First, ROI (bilateral NAcc)-based FC values were generated for each participant using DPABI v8.2. Second, warped Bayesian linear regression (41) was used to generate the normative models of FC at each specific voxel of the whole brain on the HC group (N=232), using age and gender as independent variables to predict FC, i.e., dependent variable, which was performed using the PCNtoolkit (Predictive Clinical Neuroscience toolkit) (42). The normative model of each voxel contains both the predicted FC and related predictive confidence at this voxel. The predictive confidence could be interpreted as centiles of variation within the norm population. Third, the normative model helps us place each participant within the normative range and then quantify their deviations (i.e., Z scores) of FC from the healthy range at each specific brain voxel. One participant-specific ( i ) Z score at each voxel ( j ) is calculated using the equation below: This equation combines three sources of information: 1) the difference between true response ( ) and predicted response ( ); 2) the predicted variance of the specific voxel ( ); 3) the variance of the normative data of the specific voxel ( ). Using this equation, a normative probability map (NPM) with all the Z scores of all the brain voxel for each participant is generated. Thus, the NPM of one participant provides a statistical estimate of how much the FC of the participant differs from the FC of healthy normative pattern at each voxel, i.e., Z scores. Lastly, to identify brain voxels with significant deviations from healthy normative pattern across the whole brain of each participant, the NPM of each participant was thresholded using FDR (false discovery rate) at p <0.01 (22). To examine the spatial spread of the significant deviations among the IGD group, an overlap map by counting all the participant-level FDR-corrected NPMs was generated. This can be used to identify which brain regions had positive deviation (increased FC) or negative deviation (decreased FC) among the IGD group compared to the HC group. Estimating regional deviations for each participant To better depict the important deviations of each participant’s FC, we calculated one summary index for each participant, i.e., regional deviation to capture the maximum deviation across specific brain regions or functional networks. To calculate regional deviations, we first parcellated the whole brain into 10 functional networks (the FPN, BGN, limbic network, default mode network, medial frontal network, motor network, visual association network, visual network I and II, and cerebellum network) from Shen’s brain functional atlas (32) (https://bioimagesuiteweb.github.io/ webapp/connviewer.html?species5human). Then, the trimmed mean of 1% of top positive/negative deviations across each functional network was taken as the regional deviation for each region of each participant. In the end, the greater trimmed mean deviation between positive and negative deviations was taken as the representative deviation for each participant in one functional network. Cluster analysis Cluster analysis was applied to identify subtypes of IGD participants. In the current study, the data-driven K-means cluster analysis (43) was conducted using the “kmeans” function in MATLAB to classify the IGD participants based on their regional deviations of the cognitive control and reward networks, i.e., FPN and BGN. The optimal number of clusters was determined by two methods: 1) the average Silhouette score for each value of cluster numbers, and the solution with the highest Silhouette score was considered as the optimal clustering configuration (44); 2) the within-cluster sum of squares (WCSS) and the elbow method, where the point of inflection indicated a suitable number of clusters (45). Results Individual deviation from normative models Figure 2 presents the summary maps of intra-individual FDR-corrected NPMs for the IGD and HC groups, highlighting voxels that showed significant positive or negative deviations from the normative models. As expected, the HC group showed few significant deviations, indicating that the normative model provided a good fit for this group. This fit was achieved under 10-fold cross-validation and was therefore unbiased. In contrast, the voxels with significant deviations in the IGD group were noticeably more than the HC group and were widespread across the brain. Figure 3 presents the top ten IGD participants with significant deviated voxels. The IGD participants with a large range of deviated voxels represented only a very small fraction of the IGD group, and their deviations were different from each other. This result emphasized that individual differences exist within the IGD group and cannot be ignored. Comparison between normative modeling and case-control Using traditional case-control method, we performed an independent sample t test on the FC of the IGD group and the HC group. FWE correction with voxel p =0.01 and cluster p =0.05 was applied to obtain significant group difference on FC. The results (Figure 4A) showed that the IGD group had significant stronger FC between NAcc and limbic lobe (i.e., the putamen, pallidum, thalamus, amygdala, and hippocampus than the HC group. Then, we removed the top 20 IGD participants with extreme deviations from the IGD group, and performed an independent sample t test on the FC of the remaining IGD participants (N=153) and the HC group. FWE correction with voxel p =0.01 and cluster p =0.05 was also applied. As shown in Figure 4B, no significant difference was found. This result further revealed the existence of individual differences, i.e., heterogeneity, within the IGD group. Moreover, the abnormalities of IGD found in previous studies using case-control method may only be caused by a small proportion of IGD individuals. The IGD subtype s based on K-means cluster analyses By using the custom code in MATLAB, we computed the average Silhouette score and the WCSS of cluster number for a range of cluster numbers from 2 to 10. Based on the results from both the Silhouette analysis (Figure 5A) and the elbow method applied to the WCSS curve (Figure 5B), the optimal number of clusters was determined to be 5. The k-means algorithm then iteratively assigned cases to clusters to minimize within-cluster variance. The final cluster solution demonstrated good validity, stability and interpretability. As shown in Figure 5C and Table 2, based on the regional deviations of the FPN and BGN, the IGD group was clustered into 5 subgroups: 1) the first subtype occupied 30.06% and showed negative deviations both in the FPN and in the BGN, and the average IAT score of this subgroup was 64.23, the average DSM-5 score was 5.98; 2) the second subtype occupied 17.92, showing positive deviations both in the FPN and in the BGN, and the average IAT score was 68.00, the average DSM-5 score was 6.23; 3) the third subtype occupied 25.43%, showing negative deviations in the FPN and positive deviations in the BGN, and the average IAT score was 67.07, the DSM-5 score was 6.13; 4) the fourth subtype occupied 26.01%, showing positive deviations in the FPN and negative deviations in the BGN, and the average IAT score was 65.80, the DSM-5 score was 6.09; 5) the fifth subtype p only have one IGD participant, showing extreme positive deviations in the FPN and extreme negative deviations in the BGN, and the IAT score was 84.00, the DSM-5 score was 7. Table 2. The results of k-means cluster analysis. Cluster 1 N=52 Cluster 2 N=31 Cluster 3 N=44 Cluster 4 N=45 Cluster 5 N=1 Proportion (%) 30.06 17.92 25.43 26.01 0.58 Mean regional deviations of the FPN -3.02 3.34 -2.95 3.10 11.52 Mean regional deviations of the BGN -2.90 3.44 3.04 -3.12 -10.68 Mean IAT scores 64.23 68.00 67.07 65.80 84.00 Mean DSM-5 scores 5.98 6.23 6.13 6.09 7.00 Abbreviations: N, Number; FPN, frontoparietal network; BGN, basal ganglia network; IAT, internet addiction test; DSM-5, the fifth edition of diagnostic and statistical manual of mental disorders. Discussion The heterogeneity in abnormal FC of IGD participants Our results provide compelling evidence that IGD is characterized by substantial heterogeneity in resting-state FC. The normative modeling approach revealed that a minority of IGD individuals exhibited widespread and diverse deviations in FC between NAcc and the remaining brain, the majority exhibited less and also diverse deviations. The ‘diverse’ could be seen from the overlap of all the IGD participants’ FDR-corrected deviation maps (Figure 2). Only 5/4 participants had significant positive/ negative deviation on a certain voxel. This is consistent with prior findings in psychiatric disorders such as autism spectrum disorder, major depression disorder and bipolar disorder, where patients show various deviations across brain in neural features, i.e., grey matter, cortical thickness), and only a small portion of patients show large deviations (14, 16, 20, 21). Such heterogeneity suggests that case-control studies, which average across individuals, likely highlight abnormalities present in only a subset—potentially explaining inconsistent findings among IGD RS-fMRI studies as we reviewed in Background. Indeed, when the top 20 most deviant IGD participants were removed, typical group-level FC differences (e.g., increased NAcc–limbic connectivity in IGD) vanished in our analysis. This underscores that normative modeling is vital for uncovering individual FC aberrations masked by averaged group contrasts, and aligns with findings in other psychiatric domains advocating for precision medicine approaches (46, 47). Beyond overall variability, heterogeneity of IGD manifested in distinct patterns of FC deviations across core functional networks related to IGD, i.e., the FPN responsible for cognitive control and the BGN responsible for reward processing (34, 35). By clustering the representative deviations of the two networks (regional deviation) in all the IGD participants, we identified five subtypes of IGD. Such an approach revealed that some IGD participants had negative deviations—a potential indicator of hypoconnectivity affecting cognitive control/reward processing, whereas others showed positive deviations, suggesting hyperconnectivity in cognitive control/reward circuits. This clustering results further explain why some researchers identified hyperconnectivity between the inferior frontal gyrus responsible for cognitive control and NAcc in IGD (48) whereas some researchers identified hypoconnectivity between the DLPFC responsible for cognitive control and NAcc in IGD (8, 31). Our findings suggest that deviations vary qualitatively across individuals with IGD, instead of one uniform pattern. Clinically, this implies that interventions such as cognitive training or reward-modulating therapies should be tailored—some IGD individuals may benefit from enhancing frontoparietal control, others from regulating striatal hyperactivity. Identification of IGD Subtypes and Clinical Implications The k‑means clustering on individualized FC deviations of the FPN and BGN yielded five distinct IGD subtypes. These were: (1) both hypoconnectivity between FPN and NAcc, as well as between BGN and NAcc; (2) both hyperconnectivity between FPN and NAcc, as well as between BGN and NAcc; (3) hypoconnectivity between FPN and NAcc–hyperconnectivity between BGN and NAcc; (4) hyperconnectivity between FPN and NAcc–hypoconnectivity between BGN and NAcc; and (5) extreme mode of subtype 4 (one IGD participant), the addiction severity of this participant was the highest (the highest IAT score and DSM-5 score). Interestingly, average addiction severity scores (IAT/DSM-5) were comparable across subtypes 1-4, suggesting that addiction severity does not align directly with FC deviation profiles. This resonates with normative modeling studies in autism and obsessive-compulsive disorder that identified cross-diagnostic subtypes not predicted by traditional symptom scales (20, 49). Our findings highlight the possible existence of meaningful neurobiological subtypes that conventional clinical assessments fail to detect. These findings point toward a precision psychiatry framework for IGD: mapping FC deviation patterns enables subtype-specific interventions, especially transcranial magnetic stimulation (TMS) intervention (50). For example, individuals in subtype 3 (reduced control, heightened reward connectivity) may benefit from a dual intervention approach: high-frequency TMS (≥10 Hz) targeting control-related areas such as the DLPFC to strengthen executive function, and low-frequency TMS (≤1 Hz) to reward-related areas (e.g., striatum, ventromedial PFC) to reduce hyperactive reward processing (51, 52). Conversely, subtype 4 (enhanced control but diminished reward responsiveness) may benefit from high-frequency TMS stimulation of reward-related circuits to restore motivational salience, while avoiding further stimulation of already over-engaged control networks. Our single-participant outlier (subtype 5), showing extreme FC deviation, underscores the need for individualized TMS protocols—tailored not only in anatomical targets but also in frequency—to address unique neurofunctional profiles. These insights support a frequency- and connectivity-informed TMS framework for personalized IGD treatment. Future research should validate these subtypes in longitudinal cohorts, link them with task-related fMRI and behavioral profiles, and assess response to tailored interventions. Expanding normative models to incorporate dynamic FC and multimodal imaging—such as cortical thickness and molecular connectivity—could refine subtype definitions. Ultimately, embracing normative modeling and data-driven subtyping may enable meaningful stratification of IGD and pave the way for mechanism-based, personalized treatments (46). Conclusions The current study reveals two key findings related to IGD. First, using a normative modeling framework, we uncovered substantial individual heterogeneity in resting-state FC abnormalities among IGD participants. While some individuals showed pronounced FC deviations between the NAcc and widespread brain regions, most exhibited less extensive and highly variable patterns, suggesting that group-level analyses may obscure meaningful individual differences. Second, by clustering individualized deviations within the FPN and BGN, we identified five distinct FC-based IGD subtypes. These subtypes demonstrated divergent patterns of hyper- or hypoconnectivity in cognitive control and reward circuits, which were not aligned with clinical severity scores. These findings support a precision psychiatry approach for IGD, emphasizing the need for individualized neurobiological characterization and targeted interventions such as connectivity- and frequency-informed TMS. List of abbreviations Abbreviation Full Term IGD Internet Gaming Disorder APA American Psychiatric Association DSM-5 Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition HC Healthy Control RS-fMRI Resting-State Functional Magnetic Resonance Imaging FC Functional Connectivity DLPFC Dorsolateral Prefrontal Cortex NAcc Nucleus Accumbens OFC Orbitofrontal Cortex ACC Anterior Cingulate Cortex PCC Posterior Cingulate Cortex MFG Middle Frontal Gyrus ROI Region of Interest FPN Frontoparietal Network BGN Basal Ganglia Network NPM Normative Probability Map FDR False Discovery rate FWE Family-Wise Error WCSS Within-Cluster Sum of Squares TR Repetition Time TE Echo Time FOV Field of View MNI Montreal Neurological Institute FWHM Full Width at Half Maximum TMS Transcranial Magnetic Stimulation Declarations Ethics approval and consent to participate The Institutional Review Board of Hangzhou Normal University approved the present study. All procedures involving human participants were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants, and all of them completed the MRI safety screening questionnaire before the scan. Clinical trial number : not appliable. Consent for publication Not applicable. Availability of data and materials The datasets generated and analyzed during the current study are not publicly available due to restrictions related to participant privacy and institutional ethical regulations, but are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no competing interests. Funding This work was supported by the National Natural Science Foundation of China (NSFC) [number 32300925], Youth Fund for Humanities and Social Sciences Research of the Ministry of Education of China [number 23YJC190023], and the Construction Fund of Key Medical Disciplines of HangZhou [number 2025HZGF02]. The funding sources had no involvement in any of the manuscript. Authors’ contributions Z.W. and L.W. conducted the statistical analysis and wrote the manuscript. L.W. and G.D. collected the research data and modified the manuscript. Y.X. and Y.W. contributed to the data analysis. All authors read and approved the final manuscript. Acknowledgements Not applicable. References Kuss DJ, Griffiths MD. Internet Gaming Addiction: A Systematic Review of Empirical Research. Int J Ment Health Ad. 2012;10(2):278-96. Young. Internet addiction the emergence of a new clinical disorder. CyberPsychology & Behavior. 2009;1(3):237-44. APA. Diagnostic and statistical manual of mental disorders. 5th ed. Washington,DC: American Psychiatric Association; 2013. Wang L, Zhang Z, Wang S, Wang M, Dong H, Chen S, et al. Deficient dynamics of prefrontal-striatal and striatal-default mode network (DMN) neural circuits in internet gaming disorder. Journal of Affective Disorders. 2023;323:336-44. Dong H, Wang M, Zhang J, Hu Y, Potenza MN, Dong G. Reduced frontostriatal functional connectivity and associations with severity of Internet gaming disorder. Addiction Biology. 2021;26(4):e12985. Kuss DJ, Pontes HM, Griffiths MD. Neurobiological Correlates in Internet Gaming Disorder: A Systematic Literature Review. Frontiers in Psychiatry. 2018;9:166-. Yao Y, Liu L, Ma S, Shi X, Zhou N, Zhang J, et al. Functional and structural neural alterations in Internet gaming disorder: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews. 2017;83:313. Chen C, Yen J, Wang P, Liu G, Yen C, Ko C. Altered functional connectivity of the insula and nucleus accumbens in internet gaming disorder: A resting state fMRI study. European addiction research. 2016;22(4):192-200. Kim J, Chun J, Park C, Cho H, Choi J, Yang S, et al. The correlation between the frontostriatal network and impulsivity in internet gaming disorder. Scientific reports. 2019;9(1):1-9. Jin C, Zhang T, Cai C, Bi Y, Li Y, Yu D, et al. Abnormal prefrontal cortex resting state functional connectivity and severity of internet gaming disorder. Brain imaging and behavior. 2016;10(3):719-29. Kim S, Eom H, Jung YH, Kim M-K, Kim E, Kim J-J. Brain functional connectivity during and after imagery of gaming and alternative leisure activities in patients with internet gaming disorder. Neuroscience Letters. 2022;772:136451. Han X, Wu X, Wang Y, Sun Y, Ding W, Cao M, et al. Alterations of Resting-State Static and Dynamic Functional Connectivity of the Dorsolateral Prefrontal Cortex in Subjects with Internet Gaming Disorder. Frontiers in human neuroscience. 2018;12:41. Seok J, Sohn J. Altered gray matter volume and resting-state connectivity in individuals with internet gaming disorder: a voxel-based morphometry and resting-state functional magnetic resonance imaging study. Frontiers in psychiatry. 2018;9:77. Wolfers T, Doan NT, Kaufmann T, Alnæs D, Moberget T, Agartz I, et al. Mapping the Heterogeneous Phenotype of Schizophrenia and Bipolar Disorder Using Normative Models. JAMA Psychiatry. 2018;75(11). Cheng C, Ebrahimi OV, Luk JW. Heterogeneity of prevalence of social media addiction across multiple classification schemes: Latent profile analysis. Journal of Medical Internet Research. 2022;24(1):e27000. Zabihi M, Oldehinkel M, Wolfers T, Frouin V, Goyard D, Loth E, et al. Dissecting the heterogeneous cortical anatomy of autism spectrum disorder using normative models. Biological psychiatry: cognitive neuroscience and neuroimaging. 2019;4(6):567-78. Alexander-Bloch AF, Reiss PT, Rapoport J, McAdams H, Giedd JN, Bullmore ET, et al. Abnormal cortical growth in schizophrenia targets normative modules of synchronized development. Biological psychiatry. 2014;76(6):438-46. Carroll KM. The profound heterogeneity of substance use disorders: Implications for treatment development. Current directions in psychological science. 2021;30(4):358-64. White T, Schmidt M, Karatekin C. White matter ‘potholes’ in early-onset schizophrenia: A new approach to evaluate white matter microstructure using diffusion tensor imaging. Psychiatry Research: Neuroimaging. 2009;174(2):110-5. Shan X, Uddin LQ, Xiao J, He C, Ling Z, Li L, et al. Mapping the heterogeneous brain structural phenotype of autism spectrum disorder using the normative model. Biological Psychiatry. 2022;91(11):967-76. Sun X, Sun J, Lu X, Dong Q, Zhang L, Wang W, et al. Mapping neurophysiological subtypes of major depressive disorder using normative models of the functional connectome. Biological Psychiatry. 2023;94(12):936-47. Marquand AF, Rezek I, Buitelaar J, Beckmann CF. Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control Studies. Biological Psychiatry. 2016;80(7):552-61. Everitt BJ, Robbins TW. Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nature Neuroscience. 2005;8(11):1481-9. Wang M, Zheng H, Zhou W, Jiang Q, Dong GH. Persistent dependent behaviour is accompanied by dynamic switching between the ventral and dorsal striatal connections in internet gaming disorder. Addiction Biology. 2021;26(6). Yan H, Shlobin NA, Jung Y, Zhang KK, Warsi N, Kulkarni AV, et al. Nucleus accumbens: a systematic review of neural circuitry and clinical studies in healthy and pathological states. Journal of Neurosurgery. 2022;138(2):337-46. Dong, Li H, Wang L, Potenza MN. Cognitive control and reward/loss processing in Internet gaming disorder: Results from a comparison with recreational Internet game-users. European Psychiatry. 2020;44:30-8. Sun Y, Ying H, Seetohul RM, Xuemei W, Ya Z, Qian L, et al. Brain fMRI study of crave induced by cue pictures in online game addicts (male adolescents). Behavioural Brain Research. 2012;233(2):563-76. Wang L, Yang G, Zheng Y, Li Z, Qi Y, Li Q, et al. Enhanced neural responses in specific phases of reward processing in individuals with Internet gaming disorder. Journal of Behavioral Addictions. 2021;10(1):99-111. Shaw SY, Shah L, Jolly AM, Wylie JL. Identifying heterogeneity among injection drug users: a cluster analysis approach. American Journal of Public Health. 2008;98(8):1430-7. Wang L, Yang G, Zheng Y, Li Z, Wei P, Li Q, et al. Neural substrates of deficient cognitive control in individuals with severe internet gaming disorder. NeuroImage: Clinical. 2021;32:102828. Dong G, Lin X, Hu Y, Xie C, Du X. Imbalanced functional link between executive control network and reward network explain the online-game seeking behaviors in Internet gaming disorder. Sci Rep. 2015;5(5 Suppl 2):9197. Shen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. NeuroImage. 2013;82:403-15. Haber S, Knutson B. The reward circuit: linking primate anatomy and human imaging. Neuropsychopharmacology. 2010;35(1):4-26. Haber S. Parallel and integrative processing through the Basal Ganglia reward circuit: lessons from addiction. Biological psychiatry. 2008;64(3):173-4. Nee DE. Integrative frontal-parietal dynamics supporting cognitive control. elife. 2021;10:e57244. Young KS. Caught in the net: How to recognize the signs of internet addiction--and a winning strategy for recovery. New York: John Wiley & Sons, Inc; 1998. Petry NM, Rehbein F, Gentile DA, Lemmens JS, Rumpf H-J, Mößle T, et al. An international consensus for assessing internet gaming disorder using the new DSM-5 approach. Addiction. 2014;109(9):1399-406. Dong, Liu X, Zheng H, Du X, Potenza MN. Brain response features during forced break could predict subsequent recovery in internet gaming disorder: A longitudinal study. Journal of Psychiatric Research. 2019;113:17-26. Widyanto L, Mcmurran, M.,. The psychometric properties of the internet addiction test. Cyberpsychology & Behavior. 2004;7(4):443. Wang L, Zheng H, Wang M, Chen SY, Du XX, Dong GH. Sex differences in neural substrates of risk taking: Implications for sex-specific vulnerabilities to internet gaming disorder. Journal of Behavioral Addictions. 2022;11(3):778-95. Fraza CJ, Dinga R, Beckmann CF, Marquand AF. Warped Bayesian linear regression for normative modelling of big data. NeuroImage. 2021;245:118715. Rutherford S, Kia SM, Wolfers T, Fraza C, Zabihi M, Dinga R, et al. The normative modeling framework for computational psychiatry. Nature protocols. 2022;17(7):1711–34. Sinaga KP, Yang M-S. Unsupervised K-means clustering algorithm. IEEE access. 2020;8:80716-27. Shahapure KR, Nicholas C, editors. Cluster quality analysis using silhouette score. 2020 IEEE 7th international conference on data science and advanced analytics (DSAA); 2020: IEEE. Cui M. Introduction to the k-means clustering algorithm based on the elbow method. Accounting, Auditing and Finance. 2020;1(1):5-8. Ginsburg GS, Phillips KA. Precision medicine: from science to value. Health affairs. 2018;37(5):694-701. Leopold JA, Loscalzo J. Emerging role of precision medicine in cardiovascular disease. Circulation research. 2018;122(9):1302-15. Zhang Z, Wang M, Meng G, Qi Y, Wang L, Dong G-H. Altered dynamic reconfiguration of brain functional networks during gaming and deprivation in individuals with internet gaming disorder. Journal of Behavioral Addictions. 2025. Liu L, Jia D, He Z, Wen B, Zhang X, Han S. Individualized functional connectome abnormalities obtained using two normative model unveil neurophysiological subtypes of obsessive compulsive disorder. Progress in Neuro-Psychopharmacology and Biological Psychiatry. 2024;135:111122. Ekhtiari H, Tavakoli H, Addolorato G, Baeken C, Bonci A, Campanella S, et al. Transcranial electrical and magnetic stimulation (tES and TMS) for addiction medicine: a consensus paper on the present state of the science and the road ahead. Neuroscience & Biobehavioral Reviews. 2019;104:118-40. Gorelick DA, Zangen A, George MS. Transcranial magnetic stimulation in the treatment of substance addiction. Annals of the New York Academy of Sciences. 2014;1327(1):79-93. Yuan J, Liu W, Liang Q, Cao X, Lucas MV, Yuan T-F. Effect of low-frequency repetitive transcranial magnetic stimulation on impulse inhibition in abstinent patients with methamphetamine addiction: a randomized clinical trial. JAMA network open. 2020;3(3):e200910-e. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 02 May, 2026 Read the published version in Annals of General Psychiatry → Version 1 posted Editorial decision: Revision requested 19 Oct, 2025 Reviews received at journal 18 Oct, 2025 Reviewers agreed at journal 27 Sep, 2025 Reviews received at journal 03 Sep, 2025 Reviewers agreed at journal 20 Aug, 2025 Reviewers invited by journal 27 Jul, 2025 Editor assigned by journal 20 Jun, 2025 Submission checks completed at journal 20 Jun, 2025 First submitted to journal 20 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6938965","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":492574217,"identity":"a53017ad-327e-4282-9240-9e3f13f5166a","order_by":0,"name":"Zixiao Wang","email":"","orcid":"","institution":"the Affiliated Hospital of Hangzhou Normal University","correspondingAuthor":false,"prefix":"","firstName":"Zixiao","middleName":"","lastName":"Wang","suffix":""},{"id":492574218,"identity":"bdb0383a-2fd8-4d78-a588-7a20d1633f25","order_by":1,"name":"Yong Xie","email":"","orcid":"","institution":"the Affiliated Hospital of Hangzhou Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yong","middleName":"","lastName":"Xie","suffix":""},{"id":492574219,"identity":"383b683c-0501-4794-9b31-bbd777574a08","order_by":2,"name":"Yidan Wang","email":"","orcid":"","institution":"the Affiliated Hospital of Hangzhou Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yidan","middleName":"","lastName":"Wang","suffix":""},{"id":492574220,"identity":"c4a3f76a-dd5a-4a97-a941-436834491356","order_by":3,"name":"Lingxiao Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBElEQVRIiWNgGAWjYLCCBwZAghmIPzZABCQIakmAamGcSbwWKM3MS4wWg+NnD79IKLhjt+E488PHtjvq8gwOMB+8zcNgl4dTy5m8NIsEg2fJM5vZjI1zz7AVGxxgS7bmYUguxqXF7ECOmUGCweFkfmYGM+ncNp7EDQd4zKR5GA4kNuDScv4NRAsbM/s3acs2CaAW/m/4tdzIMX4A1GLHzww0nLHNAGQLG14t9jfemAED+XCCZDNPsWFvW0LizMNsxpZzDJJxapHszzH+8OHPYXuD88c3PvjZVpfYd7z54Y03FXY4tQABGygWkBSAkgGDAW71ICUfQA7Eq2QUjIJRMApGNgAAK7NXQydUYC4AAAAASUVORK5CYII=","orcid":"","institution":"the Affiliated Hospital of Hangzhou Normal University","correspondingAuthor":true,"prefix":"","firstName":"Lingxiao","middleName":"","lastName":"Wang","suffix":""},{"id":492574221,"identity":"87ef50e4-d84e-4832-9d8d-b750d2cb89c6","order_by":4,"name":"Guang-Heng Dong","email":"","orcid":"","institution":"Yunnan Normal University","correspondingAuthor":false,"prefix":"","firstName":"Guang-Heng","middleName":"","lastName":"Dong","suffix":""}],"badges":[],"createdAt":"2025-06-20 12:38:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6938965/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6938965/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12991-026-00664-3","type":"published","date":"2026-05-02T15:57:46+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":88030459,"identity":"ede50e64-3cc0-4179-ada5-0cdf73876244","added_by":"auto","created_at":"2025-07-31 15:24:24","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":269500,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe overview of normative modeling approach\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: NAcc, nucleus accumbens; FC, functional connectivity; IGD, internet gaming disorder; HCs, healthy controls; FPN, frontoparietal network; BGN, basal ganglia network; FDR, false discovery rates.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/08225e1221732d3dd5f985be.png"},{"id":88030455,"identity":"a0501146-4426-4d58-b618-5ed08c23e5c8","added_by":"auto","created_at":"2025-07-31 15:24:24","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":589294,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverlap of voxel-wise deviations across each group.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA. Overlap of positive deviations among the IGD group after FDR correction (\u003cem\u003ep\u003c/em\u003e=0.01).\u003c/p\u003e\n\u003cp\u003eB. Overlap of negative deviation among the HC group after FDR correction (\u003cem\u003ep\u003c/em\u003e=0.01).\u003c/p\u003e\n\u003cp\u003eAbbreviations: IGD, internet gaming disorder; HC, healthy control; FDR, false discovery rates.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/5894dcb3322e038d1500f844.png"},{"id":88031023,"identity":"85cbe473-4969-43a8-8f85-2d9a139ca9bf","added_by":"auto","created_at":"2025-07-31 15:32:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":936441,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe NPMs of top ten IGD participants with significant deviated voxels\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: IGD, internet gaming disorder; NPMs, normative probability map.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/734d5a7f2350c135a8e4105d.png"},{"id":88030478,"identity":"03cdcbd0-a8c2-42a6-9d58-bd02443750ea","added_by":"auto","created_at":"2025-07-31 15:24:24","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":245905,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe results of case-control method\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA. The significant difference of FC between the IGD group and the HC group after using FWE correction with voxel \u003cem\u003ep\u003c/em\u003e=0.01 and cluster \u003cem\u003ep\u003c/em\u003e=0.05. The red indicates stronger FC between the regions and NAcc in the IGD group than the HC group.\u003c/p\u003e\n\u003cp\u003eB. No significant difference of FC between the IGD group with removing the top 20 IGD participants with extreme deviations and the HC group.\u003c/p\u003e\n\u003cp\u003eAbbreviations: IGD, internet gaming disorder; HC, healthy control; L, left; R, right; FC, functional connectivity; NAcc, nucleus accumbens.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/71f2af3554e4da5c87324cff.png"},{"id":88030514,"identity":"f47ed5bd-b26a-4b21-b294-5a5143a3d509","added_by":"auto","created_at":"2025-07-31 15:24:26","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":187972,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe results of k-means cluster analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA. The average Silhouette score was calculated for cluster numbers (K) ranging from 2 to 10.\u003c/p\u003e\n\u003cp\u003eB. The within-cluster sum of squares (WCSS) was calculated for cluster numbers (K) ranging from 2 to 10.\u003c/p\u003e\n\u003cp\u003eC. The scatter plot of the cluster results based on the IGD participants’ regional deviations of the FPN and BGN.\u003c/p\u003e\n\u003cp\u003eAbbreviations: FPN, frontoparietal network; BGN, basal ganglia network.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/831804beae00d54a017a1ca2.png"},{"id":108498190,"identity":"bff108d5-ef40-4c86-b828-22686b3fbefa","added_by":"auto","created_at":"2026-05-05 10:14:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2536637,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6938965/v1/09586f86-ca41-4faa-861c-2f685c68232a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Normative Modeling Reveals Functional Connectivity Heterogeneity and Subtypes in Internet Gaming Disorder","fulltext":[{"header":"Background","content":"\u003cp\u003eInternet gaming disorder (IGD) refers to the problematic and excessive use of internet games, which can be detrimental to various aspects of one's education and life (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). American psychiatric association (APA) has included IGD in Section III of fifth edition of diagnostic and statistical manual of mental disorders (DSM-5) as warranting more clinical research in 2013 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Using resting-state functional magnetic resonance imaging (RS-fMRI), researchers have conducted numerous studies on the characteristics and underlying neural basis of IGD, demonstrating some core neural variation of IGD, such as abnormal brain functional networks related to cognitive control, reward processing (\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHowever, after summarizing previous RS-fMRI studies on IGD, we found that it is hard to draw a uniform conclusion about the underlying neural mechanism of IGD. For example, some researchers found that compared to healthy control (HC) group, IGD group displayed decreased functional connectivity (FC) between the dorsolateral prefrontal cortex (DLPFC) and nucleus accumbens (NAcc) (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e), decreased FC between the orbitofrontal cortex (OFC) and putamen (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e), and decreased FC between the anterior cingulate cortex (ACC) and caudate (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). On the other hand, some researchers identified increased FC between the OFC and putamen (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e), increased FC between the DLPFC and posterior cingulate cortex (PCC) (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e), and increased FC between the middle frontal gyrus (MFG) and caudate (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e) in IGD group in comparison with HC group. Both the direction and the focal area of abnormalities of IGD group exhibited different across different RS-fMRI studies. We inferred that the heterogeneity of IGD in the abnormalities of resting-state FC patterns is the main reason for these inconsistent results. Heterogeneity is one of core features of many disorders, including schizophrenia, autism, bipolar disorder and substance addiction (\u003cspan additionalcitationids=\"CR15 CR16 CR17\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). However, the heterogeneity of IGD is barely understood. The heterogeneity of IGD emphasizes the importance of identifying individualized abnormality of IGD, thus contributing to prompt the individualized treatment for IGD. In contrast, group-level differences from comparisons between IGD and healthy groups in previous studies might not necessarily be representative of any IGD gamer. They only represent a small part of the neural abnormalities that characterize IGD and the bulk of the small part of abnormalities are comprised of individual IGD who are highly deviated from the norm (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAccordingly, the present study was set to map the individualized abnormalities of brain FC in IGD using a recent methodology named normative modeling. Normative modeling has been well documented and applied in studies of heterogeneity in other psychiatric disorders, such as autism, schizophrenia, depressive and bipolar disorder (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). A normative model can be understood as a statistical model that maps demographic or behavioral variables to a quantitative brain readout. Using brain data from a large number of typically developed individuals (i.e., HC individuals), a Bayesian process regression model was used to calculate the normative range within which brain metrics could be varied up or down for HC individuals at specific ages and sexes (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e). Then, each IGD individual brain metric is compared to the normative model of same age and sex to compute their individualized abnormalities/deviations relative to HC individuals. This is the first goal of the present study. We assumed that different IGD individual would exhibit different abnormalities/deviations from normative model, i.e., heterogeneity, in resting-state brain FC, as we summarized from previous RS-fMRI studies.\u003c/p\u003e\u003cp\u003eAdditionally, the bilateral NAcc was selected as the region of interest (ROI) for the resting-state ROI-based brain FC analysis to compute the FC between the NAcc and the whole remaining brain regions in the current study. The NAcc, the main projection region of the mesolimbic pathway, has been confirmed to be involved in driving craving and development and maintain of addiction in many addicted types, including IGD (\u003cspan additionalcitationids=\"CR24\" citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). Moreover, many studies have revealed abnormal activation/FC of the NAcc in individuals with IGD (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe second goal of the present study is to distinguish different subtypes of IGD using clustering algorithm. Clustering algorithm is also a common approach to biological heterogeneity. This approach is successful to some extent and is appropriate if the clinical cohort can be cleanly divided into a relatively small number of homogeneous subgroups based on the measures chosen (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). Accordingly, by using clustering algorithm, we expect to identify some homogeneous subtypes of IGD individuals based on their direction and magnitude of deviations from normative models regarding FC. Particularly, according to previous studies on IGD, the cognitive control and reward networks were most related to IGD (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Thus, we aimed to identify the subtypes of IGD participants based on their abnormalities/deviations from normative model in the cognitive control and reward networks. In the current study, we used the frontoparietal network (FPN) and the basal ganglia network (BGN) from Shen\u0026rsquo;s brain functional atlas (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e), which were implicated in the cognitive control and reward networks separately (\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e).\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eParticipants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe present study included 173 university students with IGD (males/females: 98/75) and 232 HCs (males/females: 139/93). The numbers of males and females were matched between the two groups (\u0026chi;\u003csup\u003e2\u0026nbsp;\u003c/sup\u003e=0.436, \u003cem\u003ep\u003c/em\u003e =0.509). All the participants were free of any psychiatric/neurological disorders confirmed by structured psychiatric interviews (Mini International Neuropsychiatric Interview). No participant reported any illegal drug use and gambling, and excessive nicotine and alcohol uses. The diagnosis of IGD was based on the Young\u0026apos;s Internet Addiction Test (IAT) (36) and DSM-5 criteria for IGD (37). Consistent with previous studies of IGD (4, 38-40), the diagnostic criteria for the IGD group included: 1) an IAT score \u0026gt;50; 2) a DSM-5 criteria score \u0026gt;5; and 3) spending most of their internet time in playing games. Participants with DSM-5 criteria scores \u0026lt;5 and IAT scores \u0026lt;50 were classified as the HC group. As shown in Table 1, the IGD group reported significant greater IAT scores and DSM-5 scores than the HC group.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1. Demographic information and group differences\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"614\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 198px;\"\u003e\n \u003cp\u003eItems\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eIGD\u003c/p\u003e\n \u003cp\u003e(M=98, F=75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 123px;\"\u003e\n \u003cp\u003eHC\u003c/p\u003e\n \u003cp\u003e(M=139, F=93)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 198px;\"\u003e\n \u003cp\u003eAge (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e21.13\u0026plusmn;2.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e21.53\u0026plusmn;2.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-1.663\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e0.097\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 198px;\"\u003e\n \u003cp\u003eIAT score\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e66.15\u0026plusmn;9.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e40.32\u0026plusmn;10.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e25.223\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 198px;\"\u003e\n \u003cp\u003eDSM-5 score\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e6.08\u0026plusmn;1.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e2.51\u0026plusmn;1.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e26.447\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable values: mean \u0026plusmn; standard deviation\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAbbreviations: IGD, internet gaming disorder; HC, healthy control; M, male; F, female; IAT, internet addiction test; DSM-5, the fifth edition of diagnostic and statistical manual of mental disorders.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRS-fMRI data acquisition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll participants underwent a RS-fMRI scan for 7 min using a 3 T scanner (Siemens Trio, Malvern, PA, USA) equipped for echo-planar imaging (EPI). During the scan, they were asked to open eyes and stay still and relaxed. The scan parameters were as follows: repetition time (TR) = 2000 ms, echo time (TE) = 30 ms, slice number = 33, interleaved sequence, slice thickness = 3 mm, voxel size = 3 \u0026times; 3 \u0026times; 3 mm\u003csup\u003e3\u003c/sup\u003e, field of view (FOV) = 220 \u0026times; 220 mm\u003csup\u003e2\u003c/sup\u003e, flip angle = 90\u003csup\u003e◦\u003c/sup\u003e and matrix of 64 \u0026times; 64.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreprocessing of RS-fMRI data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData preprocessing was performed using DPABI V8.2 (http://rfmri.org/dpabi). The preprocessing steps included: 1) discarding the first 10 time points; 2) slice-timing; 3) realign for head-motion correction; 4) spatial normalization to Montreal Neurological Institute (MNI) standard space; 5) spatial smoothing (FWHM = 6 mm); 6) linear trend removal; 7) nuisance covariates regression, including head-motion covariates using the Friston 24-parameter model as well as signals from white matter, cerebrospinal fluid and global signals; 8) band-pass filtering with a range of 0.01-0.1 Hz.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConstructing normative models of FC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAn overview of the normative modeling approach is provided in Figure 1; this approach has been described previously (22). First, ROI (bilateral NAcc)-based FC values were generated for each participant using DPABI v8.2. Second, warped Bayesian linear regression (41) was used to generate the normative models of FC at each specific voxel of the whole brain on the HC group (N=232), using age and gender as independent variables to predict FC, i.e., dependent variable, which was performed using the PCNtoolkit (Predictive Clinical Neuroscience toolkit) (42). The normative model of each voxel contains both the predicted FC and related predictive confidence at this voxel. The predictive confidence could be interpreted as centiles of variation within the norm population. Third, the normative model helps us place each participant within the normative range and then quantify their deviations (i.e., Z scores) of FC from the healthy range at each specific brain voxel. One participant-specific (\u003cem\u003ei\u003c/em\u003e) Z score at each voxel (\u003cem\u003ej\u003c/em\u003e) is calculated using the equation below:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"126\" height=\"45\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e\u003c/p\u003e\n\u003cp\u003eThis equation combines three sources of information: 1) the difference between true response (\u003cimg width=\"19\" height=\"21\" src=\"data:image/png;base64,R0lGODlhHAAfAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAADAAaABMAhAAAAAAAAAAAOgAAZgA6ZgA6kABmtjoAADoAOjqQ22YAAGY6AGY6ZmZmOma222a2/5A6AJDb/7ZmALZmOrbb/7b/27b//9uQOtv///+2Zv/bkP/btv//tv//2wECAwECAwWYILAtgRABaBYEg4W+8MsxlFpg7xUYcR9zilYu8PAZURLiS3I7GnUJlOZQRAphqmpqh7o0AR3Id3mVHm4axMlpDLckPPZRImiUdcqXKhCN4bVdZVuAW3EwTDgwF4JgYokyCn2HY0gmPVOEQH0XCWGWPRefL5hIKwKEZBUTWqFrEAEEFEcZBwIOk49yPkCGuj0dE6K+h7BrKCEAOw==\" alt=\"image\"\u003e) and predicted response (\u003cimg width=\"19\" height=\"21\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e); 2) the predicted variance of the specific voxel (\u003cimg width=\"19\" height=\"21\" src=\"data:image/png;base64,R0lGODlhHAAfAHcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAADAAaABMAhAAAAAAAAAAAOgAAZgA6OgA6ZgA6kABmtjoAADo6ADqQ22YAAGY6AGZmOmaQtma222a2/5A6AJC225Db/7ZmALZmOrb/27b//9uQOtu2Ztu2kNv///+2Zv/bkP//tv//2wWSICCKWpMEKGpsY+uKHZPOwfHeQIcEhQRggwsO51nUWDmEbfjCBFYjHVRECbo4Acgnkm1JkdvpqHrZWkdYMRNQFG8DinXLLBRhz84uGhUHYPsZCAITLUB1exAjGicBBA5IY2o/Z3JFfS0UkkM6iS2WciMYgy6coFQDFhWdP6OmHIIPmJqmIkVLtC4fFa24YzyEIyEAOw==\" alt=\"image\"\u003e); 3) the variance of the normative data of the specific voxel (\u003cimg width=\"22\" height=\"21\" src=\"data:image/png;base64,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\" alt=\"image\"\u003e). Using this equation, a normative probability map (NPM) with all the Z scores of all the brain voxel for each participant is generated. Thus, the NPM of one participant provides a statistical estimate of how much the FC of the participant differs from the FC of healthy normative pattern at each voxel, i.e., Z scores. Lastly, to identify brain voxels with significant deviations from healthy normative pattern across the whole brain of each participant, the NPM of each participant was thresholded using FDR (false discovery rate) at \u003cem\u003ep\u003c/em\u003e\u0026lt;0.01 (22). To examine the spatial spread of the significant deviations among the IGD group, an overlap map by counting all the participant-level FDR-corrected NPMs was generated. This can be used to identify which brain regions had positive deviation (increased FC) or negative deviation (decreased FC) among the IGD group compared to the HC group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEstimating regional deviations for each participant\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo better depict the important deviations of each participant\u0026rsquo;s FC, we calculated one summary index for each participant, i.e., regional deviation to capture the maximum deviation across specific brain regions or functional networks. To calculate regional deviations, we first parcellated the whole brain into 10 functional networks (the FPN, BGN, limbic network, default mode network, medial frontal network, motor network, visual association network, visual network I and II, and cerebellum network) from Shen\u0026rsquo;s brain functional atlas (32) (https://bioimagesuiteweb.github.io/ webapp/connviewer.html?species5human). Then, the trimmed mean of 1% of top positive/negative deviations across each functional network was taken as the regional deviation for each region of each participant. In the end, the greater trimmed mean deviation between positive and negative deviations was taken as the representative deviation for each participant in one functional network.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCluster analysis\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCluster analysis was applied to identify subtypes of IGD participants. In the current study, the data-driven K-means cluster analysis (43) was conducted using the \u0026ldquo;kmeans\u0026rdquo; function in MATLAB to classify the IGD participants based on their regional deviations of the cognitive control and reward networks, i.e., FPN and BGN. The optimal number of clusters was determined by two methods: 1) the average Silhouette score for each value of cluster numbers, and the solution with the highest Silhouette score was considered as the optimal clustering configuration (44); 2) the within-cluster sum of squares (WCSS) and the elbow method, where the point of inflection indicated a suitable number of clusters (45).\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eIndividual deviation from normative models\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 2 presents the summary maps of intra-individual FDR-corrected NPMs for the IGD and HC groups, highlighting voxels that showed significant positive or negative deviations from the normative models. As expected, the HC group showed few significant deviations, indicating that the normative model provided a good fit for this group. This fit was achieved under 10-fold cross-validation and was therefore unbiased. In contrast, the voxels with significant deviations in the IGD group were noticeably more than the HC group and were widespread across the brain.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFigure 3 presents the top ten IGD participants with significant deviated voxels. The IGD participants with a large range of deviated voxels represented only a very small fraction of the IGD group, and their deviations were different from each other. This result emphasized that individual differences exist within the IGD group and cannot be ignored.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComparison between normative modeling and case-control\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing traditional case-control method, we performed an independent sample \u003cem\u003et\u0026nbsp;\u003c/em\u003etest on the FC of the IGD group and the HC group. FWE correction with voxel \u003cem\u003ep\u003c/em\u003e=0.01 and cluster \u003cem\u003ep\u003c/em\u003e=0.05 was applied to obtain significant group difference on FC. The results (Figure 4A) showed that the IGD group had significant stronger FC between NAcc and limbic lobe (i.e., the putamen, pallidum, thalamus, amygdala, and hippocampus than the HC group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThen, we removed the top 20 IGD participants with extreme deviations from the IGD group, and performed an independent sample \u003cem\u003et\u0026nbsp;\u003c/em\u003etest on the FC of the remaining IGD participants (N=153) and the HC group. FWE correction with voxel \u003cem\u003ep\u003c/em\u003e=0.01 and cluster \u003cem\u003ep\u003c/em\u003e=0.05 was also applied. As shown in Figure 4B, no significant difference was found. This result further revealed the existence of individual differences, i.e., heterogeneity, within the IGD group. Moreover, the abnormalities of IGD found in previous studies using case-control method may only be caused by a small proportion of IGD individuals.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe IGD\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003esubtype\u003c/strong\u003e\u003cstrong\u003es based on K-means cluster analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBy using the custom code in MATLAB, we computed the average Silhouette score and the WCSS of cluster number for a range of cluster numbers from 2 to 10. Based on the results from both the Silhouette analysis (Figure 5A) and the elbow method applied to the WCSS curve (Figure 5B), the optimal number of clusters was determined to be 5. The k-means algorithm then iteratively assigned cases to clusters to minimize within-cluster variance. The final cluster solution demonstrated good validity, stability and interpretability.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs shown in Figure 5C and Table 2, based on the regional deviations of the FPN and BGN, the IGD group was clustered into 5 subgroups: 1) the first subtype occupied 30.06% and showed negative deviations both in the FPN and in the BGN, and the average IAT score of this subgroup was 64.23, the average DSM-5 score was 5.98; 2) the second subtype occupied 17.92, showing positive deviations both in the FPN and in the BGN, and the average IAT score was 68.00, the average DSM-5 score was 6.23; 3) the third subtype occupied 25.43%, showing negative deviations in the FPN and positive deviations in the BGN, and the average IAT score was 67.07, the DSM-5 score was 6.13; 4) the fourth subtype occupied 26.01%, showing positive deviations in the FPN and negative deviations in the BGN, and the average IAT score was 65.80, the DSM-5 score was 6.09; 5) the fifth subtype p only have one IGD participant, showing extreme positive deviations in the FPN and extreme negative deviations in the BGN, and the IAT score was 84.00, the DSM-5 score was 7.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. The results of k-means cluster analysis.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"643\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 265px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 1 N=52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 2\u003c/p\u003e\n \u003cp\u003eN=31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 3\u003c/p\u003e\n \u003cp\u003eN=44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 4\u003c/p\u003e\n \u003cp\u003eN=45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003eCluster 5\u003c/p\u003e\n \u003cp\u003eN=1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 265px;\"\u003e\n \u003cp\u003eProportion (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e30.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e17.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 94px;\"\u003e\n \u003cp\u003e25.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e26.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 265px;\"\u003e\n \u003cp\u003eMean regional deviations of the FPN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-3.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e3.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-2.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e3.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e11.52\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 265px;\"\u003e\n \u003cp\u003eMean regional deviations of the BGN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e-2.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e3.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e3.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-3.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e-10.68\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 265px;\"\u003e\n \u003cp\u003eMean IAT scores\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e64.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e68.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e67.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e65.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e84.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 265px;\"\u003e\n \u003cp\u003eMean DSM-5 scores\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e5.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e6.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e6.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e6.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e7.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eAbbreviations: N, Number; FPN, frontoparietal network; BGN, basal ganglia network; IAT, internet addiction test; DSM-5, the fifth edition of diagnostic and statistical manual of mental disorders.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cstrong\u003eThe heterogeneity in abnormal FC of IGD participants\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur results provide compelling evidence that IGD is characterized by substantial heterogeneity in resting-state FC. The normative modeling approach revealed that a minority of IGD individuals exhibited widespread and diverse deviations in FC between NAcc and the remaining brain, the majority exhibited less and also diverse deviations. The \u0026lsquo;diverse\u0026rsquo; could be seen from the overlap of all the IGD participants\u0026rsquo; FDR-corrected deviation maps (Figure 2). Only 5/4 participants had significant positive/ negative deviation on a certain voxel. This is consistent with prior findings in psychiatric disorders such as autism spectrum disorder, major depression disorder and bipolar disorder, where patients show various deviations across brain in neural features, i.e., grey matter, cortical thickness), and only a small portion of patients show large deviations (14, 16, 20, 21). Such heterogeneity suggests that case-control studies, which average across individuals, likely highlight abnormalities present in only a subset\u0026mdash;potentially explaining inconsistent findings among IGD RS-fMRI studies as we reviewed in Background. Indeed, when the top 20 most deviant IGD participants were removed, typical group-level FC differences (e.g., increased NAcc\u0026ndash;limbic connectivity in IGD) vanished in our analysis. This underscores that normative modeling is vital for uncovering individual FC aberrations masked by averaged group contrasts, and aligns with findings in other psychiatric domains advocating for precision medicine approaches (46, 47).\u003c/p\u003e\n\u003cp\u003eBeyond overall variability, heterogeneity of IGD manifested in distinct patterns of FC deviations across core functional networks related to IGD, i.e., the FPN responsible for cognitive control and the BGN responsible for reward processing (34, 35). By clustering the representative deviations of the two networks (regional deviation) in all the IGD participants, we identified five subtypes of IGD. Such an approach revealed that some IGD participants had negative deviations\u0026mdash;a potential indicator of hypoconnectivity affecting cognitive control/reward processing, whereas others showed positive deviations, suggesting hyperconnectivity in cognitive control/reward circuits. This clustering results further explain why some researchers identified hyperconnectivity between the inferior frontal gyrus responsible for cognitive control and NAcc in IGD (48) whereas some researchers identified hypoconnectivity between the DLPFC responsible for cognitive control and NAcc in IGD (8, 31). Our findings suggest that deviations vary qualitatively across individuals with IGD, instead of one uniform pattern. Clinically, this implies that interventions such as cognitive training or reward-modulating therapies should be tailored\u0026mdash;some IGD individuals may benefit from enhancing frontoparietal control, others from regulating striatal hyperactivity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of IGD Subtypes and Clinical Implications\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe k‑means clustering on individualized FC deviations of the FPN and BGN yielded five distinct IGD subtypes. These were: (1) both hypoconnectivity between FPN and NAcc, as well as between BGN and NAcc; (2) both hyperconnectivity between FPN and NAcc, as well as between BGN and NAcc; (3) hypoconnectivity between FPN and NAcc\u0026ndash;hyperconnectivity between BGN and NAcc; (4) hyperconnectivity between FPN and NAcc\u0026ndash;hypoconnectivity between BGN and NAcc; and (5) extreme mode of subtype 4 (one IGD participant), the addiction severity of this participant was the highest (the highest IAT score and DSM-5 score). Interestingly, average addiction severity scores (IAT/DSM-5) were comparable across subtypes 1-4, suggesting that addiction severity does not align directly with FC deviation profiles. This resonates with normative modeling studies in autism and obsessive-compulsive disorder that identified cross-diagnostic subtypes not predicted by traditional symptom scales (20, 49). Our findings highlight the possible existence of meaningful neurobiological subtypes that conventional clinical assessments fail to detect.\u003c/p\u003e\n\u003cp\u003eThese findings point toward a precision psychiatry framework for IGD: mapping FC deviation patterns enables subtype-specific interventions, especially transcranial magnetic stimulation (TMS) intervention (50). For example, individuals in subtype 3 (reduced control, heightened reward connectivity) may benefit from a dual intervention approach: high-frequency TMS (\u0026ge;10 Hz) targeting control-related areas such as the DLPFC to strengthen executive function, and low-frequency TMS (\u0026le;1 Hz) to reward-related areas (e.g., striatum, ventromedial PFC) to reduce hyperactive reward processing (51, 52). Conversely, subtype 4 (enhanced control but diminished reward responsiveness) may benefit from high-frequency TMS stimulation of reward-related circuits to restore motivational salience, while avoiding further stimulation of already over-engaged control networks. Our single-participant outlier (subtype 5), showing extreme FC deviation, underscores the need for individualized TMS protocols\u0026mdash;tailored not only in anatomical targets but also in frequency\u0026mdash;to address unique neurofunctional profiles. These insights support a frequency- and connectivity-informed TMS framework for personalized IGD treatment. Future research should validate these subtypes in longitudinal cohorts, link them with task-related fMRI and behavioral profiles, and assess response to tailored interventions. Expanding normative models to incorporate dynamic FC and multimodal imaging\u0026mdash;such as cortical thickness and molecular connectivity\u0026mdash;could refine subtype definitions. Ultimately, embracing normative modeling and data-driven subtyping may enable meaningful stratification of IGD and pave the way for mechanism-based, personalized treatments (46).\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe current study reveals two key findings related to IGD. First, using a normative modeling framework, we uncovered substantial individual heterogeneity in resting-state FC abnormalities among IGD participants. While some individuals showed pronounced FC deviations between the NAcc and widespread brain regions, most exhibited less extensive and highly variable patterns, suggesting that group-level analyses may obscure meaningful individual differences. Second, by clustering individualized deviations within the FPN and BGN, we identified five distinct FC-based IGD subtypes. These subtypes demonstrated divergent patterns of hyper- or hypoconnectivity in cognitive control and reward circuits, which were not aligned with clinical severity scores. These findings support a precision psychiatry approach for IGD, emphasizing the need for individualized neurobiological characterization and targeted interventions such as connectivity- and frequency-informed TMS.\u0026nbsp;\u003c/p\u003e"},{"header":"List of abbreviations","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"650\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eAbbreviation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFull Term\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eIGD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eInternet Gaming Disorder\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eAPA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eAmerican Psychiatric Association\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eDSM-5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eDiagnostic and Statistical Manual of Mental Disorders, Fifth Edition\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eHC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eHealthy Control\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eRS-fMRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eResting-State Functional Magnetic Resonance Imaging\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFunctional Connectivity\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eDLPFC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eDorsolateral Prefrontal Cortex\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eNAcc\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eNucleus Accumbens\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eOFC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eOrbitofrontal Cortex\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eACC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eAnterior Cingulate Cortex\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003ePCC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003ePosterior Cingulate Cortex\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eMFG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eMiddle Frontal Gyrus\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eROI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eRegion of Interest\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFPN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFrontoparietal Network\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eBGN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eBasal Ganglia Network\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eNPM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eNormative Probability Map\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFDR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFalse Discovery rate\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFWE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFamily-Wise Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eWCSS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eWithin-Cluster Sum of Squares\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eTR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eRepetition Time\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eTE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eEcho Time\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFOV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eField of View\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eMNI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eMontreal Neurological Institute\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eFWHM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eFull Width at Half Maximum\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eTMS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 463px;\"\u003e\n \u003cp\u003eTranscranial Magnetic Stimulation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Institutional Review Board of Hangzhou Normal University approved the present study. All procedures involving human participants were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants, and all of them completed the MRI safety screening questionnaire before the scan.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u003c/strong\u003e: not appliable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and analyzed during the current study are not publicly available due to restrictions related to participant privacy and institutional ethical regulations, but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (NSFC) [number 32300925], Youth Fund for Humanities and Social Sciences Research of the Ministry of Education of China [number 23YJC190023], and the Construction Fund of Key Medical Disciplines of HangZhou [number 2025HZGF02]. The funding sources had no involvement in any of the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eZ.W. and L.W. conducted the statistical analysis and wrote the manuscript. L.W. and G.D. collected the research data and modified the manuscript. Y.X. and Y.W. contributed to the data analysis.\u0026nbsp;All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eKuss DJ, Griffiths MD. Internet Gaming Addiction: A Systematic Review of Empirical Research. Int J Ment Health Ad. 2012;10(2):278-96.\u003c/li\u003e\n\u003cli\u003eYoung. Internet addiction the emergence of a new clinical disorder. CyberPsychology \u0026amp; Behavior. 2009;1(3):237-44.\u003c/li\u003e\n\u003cli\u003eAPA. Diagnostic and statistical manual of mental disorders. 5th ed. Washington,DC: American Psychiatric Association; 2013.\u003c/li\u003e\n\u003cli\u003eWang L, Zhang Z, Wang S, Wang M, Dong H, Chen S, et al. Deficient dynamics of prefrontal-striatal and striatal-default mode network (DMN) neural circuits in internet gaming disorder. Journal of Affective Disorders. 2023;323:336-44.\u003c/li\u003e\n\u003cli\u003eDong H, Wang M, Zhang J, Hu Y, Potenza MN, Dong G. Reduced frontostriatal functional connectivity and associations with severity of Internet gaming disorder. Addiction Biology. 2021;26(4):e12985.\u003c/li\u003e\n\u003cli\u003eKuss DJ, Pontes HM, Griffiths MD. Neurobiological Correlates in Internet Gaming Disorder: A Systematic Literature Review. Frontiers in Psychiatry. 2018;9:166-.\u003c/li\u003e\n\u003cli\u003eYao Y, Liu L, Ma S, Shi X, Zhou N, Zhang J, et al. Functional and structural neural alterations in Internet gaming disorder: A systematic review and meta-analysis. Neuroscience \u0026amp; Biobehavioral Reviews. 2017;83:313.\u003c/li\u003e\n\u003cli\u003eChen C, Yen J, Wang P, Liu G, Yen C, Ko C. Altered functional connectivity of the insula and nucleus accumbens in internet gaming disorder: A resting state fMRI study. European addiction research. 2016;22(4):192-200.\u003c/li\u003e\n\u003cli\u003eKim J, Chun J, Park C, Cho H, Choi J, Yang S, et al. The correlation between the frontostriatal network and impulsivity in internet gaming disorder. Scientific reports. 2019;9(1):1-9.\u003c/li\u003e\n\u003cli\u003eJin C, Zhang T, Cai C, Bi Y, Li Y, Yu D, et al. Abnormal prefrontal cortex resting state functional connectivity and severity of internet gaming disorder. Brain imaging and behavior. 2016;10(3):719-29.\u003c/li\u003e\n\u003cli\u003eKim S, Eom H, Jung YH, Kim M-K, Kim E, Kim J-J. Brain functional connectivity during and after imagery of gaming and alternative leisure activities in patients with internet gaming disorder. Neuroscience Letters. 2022;772:136451.\u003c/li\u003e\n\u003cli\u003eHan X, Wu X, Wang Y, Sun Y, Ding W, Cao M, et al. Alterations of Resting-State Static and Dynamic Functional Connectivity of the Dorsolateral Prefrontal Cortex in Subjects with Internet Gaming Disorder. Frontiers in human neuroscience. 2018;12:41.\u003c/li\u003e\n\u003cli\u003eSeok J, Sohn J. Altered gray matter volume and resting-state connectivity in individuals with internet gaming disorder: a voxel-based morphometry and resting-state functional magnetic resonance imaging study. Frontiers in psychiatry. 2018;9:77.\u003c/li\u003e\n\u003cli\u003eWolfers T, Doan NT, Kaufmann T, Aln\u0026aelig;s D, Moberget T, Agartz I, et al. Mapping the Heterogeneous Phenotype of Schizophrenia and Bipolar Disorder Using Normative Models. JAMA Psychiatry. 2018;75(11).\u003c/li\u003e\n\u003cli\u003eCheng C, Ebrahimi OV, Luk JW. Heterogeneity of prevalence of social media addiction across multiple classification schemes: Latent profile analysis. Journal of Medical Internet Research. 2022;24(1):e27000.\u003c/li\u003e\n\u003cli\u003eZabihi M, Oldehinkel M, Wolfers T, Frouin V, Goyard D, Loth E, et al. Dissecting the heterogeneous cortical anatomy of autism spectrum disorder using normative models. Biological psychiatry: cognitive neuroscience and neuroimaging. 2019;4(6):567-78.\u003c/li\u003e\n\u003cli\u003eAlexander-Bloch AF, Reiss PT, Rapoport J, McAdams H, Giedd JN, Bullmore ET, et al. Abnormal cortical growth in schizophrenia targets normative modules of synchronized development. Biological psychiatry. 2014;76(6):438-46.\u003c/li\u003e\n\u003cli\u003eCarroll KM. The profound heterogeneity of substance use disorders: Implications for treatment development. Current directions in psychological science. 2021;30(4):358-64.\u003c/li\u003e\n\u003cli\u003eWhite T, Schmidt M, Karatekin C. White matter \u0026lsquo;potholes\u0026rsquo; in early-onset schizophrenia: A new approach to evaluate white matter microstructure using diffusion tensor imaging. Psychiatry Research: Neuroimaging. 2009;174(2):110-5.\u003c/li\u003e\n\u003cli\u003eShan X, Uddin LQ, Xiao J, He C, Ling Z, Li L, et al. Mapping the heterogeneous brain structural phenotype of autism spectrum disorder using the normative model. Biological Psychiatry. 2022;91(11):967-76.\u003c/li\u003e\n\u003cli\u003eSun X, Sun J, Lu X, Dong Q, Zhang L, Wang W, et al. Mapping neurophysiological subtypes of major depressive disorder using normative models of the functional connectome. Biological Psychiatry. 2023;94(12):936-47.\u003c/li\u003e\n\u003cli\u003eMarquand AF, Rezek I, Buitelaar J, Beckmann CF. Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control Studies. Biological Psychiatry. 2016;80(7):552-61.\u003c/li\u003e\n\u003cli\u003eEveritt BJ, Robbins TW. Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nature Neuroscience. 2005;8(11):1481-9.\u003c/li\u003e\n\u003cli\u003eWang M, Zheng H, Zhou W, Jiang Q, Dong GH. Persistent dependent behaviour is accompanied by dynamic switching between the ventral and dorsal striatal connections in internet gaming disorder. Addiction Biology. 2021;26(6).\u003c/li\u003e\n\u003cli\u003eYan H, Shlobin NA, Jung Y, Zhang KK, Warsi N, Kulkarni AV, et al. Nucleus accumbens: a systematic review of neural circuitry and clinical studies in healthy and pathological states. Journal of Neurosurgery. 2022;138(2):337-46.\u003c/li\u003e\n\u003cli\u003eDong, Li H, Wang L, Potenza MN. Cognitive control and reward/loss processing in Internet gaming disorder: Results from a comparison with recreational Internet game-users. European Psychiatry. 2020;44:30-8.\u003c/li\u003e\n\u003cli\u003eSun Y, Ying H, Seetohul RM, Xuemei W, Ya Z, Qian L, et al. Brain fMRI study of crave induced by cue pictures in online game addicts (male adolescents). Behavioural Brain Research. 2012;233(2):563-76.\u003c/li\u003e\n\u003cli\u003eWang L, Yang G, Zheng Y, Li Z, Qi Y, Li Q, et al. Enhanced neural responses in specific phases of reward processing in individuals with Internet gaming disorder. Journal of Behavioral Addictions. 2021;10(1):99-111.\u003c/li\u003e\n\u003cli\u003eShaw SY, Shah L, Jolly AM, Wylie JL. Identifying heterogeneity among injection drug users: a cluster analysis approach. American Journal of Public Health. 2008;98(8):1430-7.\u003c/li\u003e\n\u003cli\u003eWang L, Yang G, Zheng Y, Li Z, Wei P, Li Q, et al. Neural substrates of deficient cognitive control in individuals with severe internet gaming disorder. NeuroImage: Clinical. 2021;32:102828.\u003c/li\u003e\n\u003cli\u003eDong G, Lin X, Hu Y, Xie C, Du X. Imbalanced functional link between executive control network and reward network explain the online-game seeking behaviors in Internet gaming disorder. Sci Rep. 2015;5(5 Suppl 2):9197.\u003c/li\u003e\n\u003cli\u003eShen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. NeuroImage. 2013;82:403-15.\u003c/li\u003e\n\u003cli\u003eHaber S, Knutson B. The reward circuit: linking primate anatomy and human imaging. Neuropsychopharmacology. 2010;35(1):4-26.\u003c/li\u003e\n\u003cli\u003eHaber S. Parallel and integrative processing through the Basal Ganglia reward circuit: lessons from addiction. Biological psychiatry. 2008;64(3):173-4.\u003c/li\u003e\n\u003cli\u003eNee DE. Integrative frontal-parietal dynamics supporting cognitive control. elife. 2021;10:e57244.\u003c/li\u003e\n\u003cli\u003eYoung KS. Caught in the net: How to recognize the signs of internet addiction--and a winning strategy for recovery. New York: John Wiley \u0026amp; Sons, Inc; 1998.\u003c/li\u003e\n\u003cli\u003ePetry NM, Rehbein F, Gentile DA, Lemmens JS, Rumpf H-J, M\u0026ouml;\u0026szlig;le T, et al. An international consensus for assessing internet gaming disorder using the new DSM-5 approach. Addiction. 2014;109(9):1399-406.\u003c/li\u003e\n\u003cli\u003eDong, Liu X, Zheng H, Du X, Potenza MN. Brain response features during forced break could predict subsequent recovery in internet gaming disorder: A longitudinal study. Journal of Psychiatric Research. 2019;113:17-26.\u003c/li\u003e\n\u003cli\u003eWidyanto L, Mcmurran, M.,. The psychometric properties of the internet addiction test. Cyberpsychology \u0026amp; Behavior. 2004;7(4):443.\u003c/li\u003e\n\u003cli\u003eWang L, Zheng H, Wang M, Chen SY, Du XX, Dong GH. Sex differences in neural substrates of risk taking: Implications for sex-specific vulnerabilities to internet gaming disorder. Journal of Behavioral Addictions. 2022;11(3):778-95.\u003c/li\u003e\n\u003cli\u003eFraza CJ, Dinga R, Beckmann CF, Marquand AF. Warped Bayesian linear regression for normative modelling of big data. NeuroImage. 2021;245:118715.\u003c/li\u003e\n\u003cli\u003eRutherford S, Kia SM, Wolfers T, Fraza C, Zabihi M, Dinga R, et al. The normative modeling framework for computational psychiatry. Nature protocols. 2022;17(7):1711\u0026ndash;34.\u003c/li\u003e\n\u003cli\u003eSinaga KP, Yang M-S. Unsupervised K-means clustering algorithm. IEEE access. 2020;8:80716-27.\u003c/li\u003e\n\u003cli\u003eShahapure KR, Nicholas C, editors. Cluster quality analysis using silhouette score. 2020 IEEE 7th international conference on data science and advanced analytics (DSAA); 2020: IEEE.\u003c/li\u003e\n\u003cli\u003eCui M. Introduction to the k-means clustering algorithm based on the elbow method. Accounting, Auditing and Finance. 2020;1(1):5-8.\u003c/li\u003e\n\u003cli\u003eGinsburg GS, Phillips KA. Precision medicine: from science to value. Health affairs. 2018;37(5):694-701.\u003c/li\u003e\n\u003cli\u003eLeopold JA, Loscalzo J. Emerging role of precision medicine in cardiovascular disease. Circulation research. 2018;122(9):1302-15.\u003c/li\u003e\n\u003cli\u003eZhang Z, Wang M, Meng G, Qi Y, Wang L, Dong G-H. Altered dynamic reconfiguration of brain functional networks during gaming and deprivation in individuals with internet gaming disorder. Journal of Behavioral Addictions. 2025.\u003c/li\u003e\n\u003cli\u003eLiu L, Jia D, He Z, Wen B, Zhang X, Han S. Individualized functional connectome abnormalities obtained using two normative model unveil neurophysiological subtypes of obsessive compulsive disorder. Progress in Neuro-Psychopharmacology and Biological Psychiatry. 2024;135:111122.\u003c/li\u003e\n\u003cli\u003eEkhtiari H, Tavakoli H, Addolorato G, Baeken C, Bonci A, Campanella S, et al. Transcranial electrical and magnetic stimulation (tES and TMS) for addiction medicine: a consensus paper on the present state of the science and the road ahead. Neuroscience \u0026amp; Biobehavioral Reviews. 2019;104:118-40.\u003c/li\u003e\n\u003cli\u003eGorelick DA, Zangen A, George MS. Transcranial magnetic stimulation in the treatment of substance addiction. Annals of the New York Academy of Sciences. 2014;1327(1):79-93.\u003c/li\u003e\n\u003cli\u003eYuan J, Liu W, Liang Q, Cao X, Lucas MV, Yuan T-F. Effect of low-frequency repetitive transcranial magnetic stimulation on impulse inhibition in abstinent patients with methamphetamine addiction: a randomized clinical trial. JAMA network open. 2020;3(3):e200910-e.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"annals-of-general-psychiatry","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"agps","sideBox":"Learn more about [Annals of General Psychiatry](http://annals-general-psychiatry.biomedcentral.com/)","snPcode":"12991","submissionUrl":"https://submission.nature.com/new-submission/12991/3","title":"Annals of General Psychiatry","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Internet gaming disorder, functional connectivity, heterogeneity, normative modeling, subtypes","lastPublishedDoi":"10.21203/rs.3.rs-6938965/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6938965/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eInternet gaming disorder (IGD) is associated with abnormal functional connectivity (FC) in brain networks. However, findings from resting-state fMRI (RS-fMRI) studies are highly inconsistent, likely due to individual heterogeneity in IGD-related neural alterations\u0026mdash;a feature commonly observed in other psychiatric disorders but understudied in IGD.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eWe applied normative modeling to RS-fMRI data to characterize individualized FC deviations between the nucleus accumbens (NAcc) and the rest of the brain in 173 IGD participants. Using a large sample of healthy controls (N\u0026thinsp;=\u0026thinsp;232) to establish age- and sex-adjusted normative ranges, we quantified the abnormality of each IGD individual\u0026rsquo;s FC profile. Further, we performed k-means clustering on deviation values within the frontoparietal network (FPN) and basal ganglia network (BGN) to identify neurobiological subtypes.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eOur findings revealed considerable heterogeneity in FC abnormalities among IGD participants. Only a minority showed strong and widespread deviations, while most exhibited mild but diverse patterns. Clustering analysis identified five distinct IGD subtypes with varying hyper- and hypoconnectivity profiles in FPN and BGN.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eThese results highlight the heterogeneous neural basis of IGD and underscore the limitations of group-level comparisons. Normative modeling and FC-based subtyping offer a promising direction for individualized assessment and may inform personalized interventions, such as connectivity- and frequency-specific TMS targeting cognitive control and reward circuits.\u003c/p\u003e","manuscriptTitle":"Normative Modeling Reveals Functional Connectivity Heterogeneity and Subtypes in Internet Gaming Disorder","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-31 15:24:15","doi":"10.21203/rs.3.rs-6938965/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-19T11:26:45+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-19T02:51:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"114524786588649718717701526301102080357","date":"2025-09-27T16:09:13+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-03T13:14:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"186703142939877914743867271512895375014","date":"2025-08-20T11:37:01+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-27T11:37:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-20T14:11:53+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-20T14:09:23+00:00","index":"","fulltext":""},{"type":"submitted","content":"Annals of General Psychiatry","date":"2025-06-20T12:29:35+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"annals-of-general-psychiatry","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"agps","sideBox":"Learn more about [Annals of General Psychiatry](http://annals-general-psychiatry.biomedcentral.com/)","snPcode":"12991","submissionUrl":"https://submission.nature.com/new-submission/12991/3","title":"Annals of General Psychiatry","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"aa15fcd4-0a85-4016-a1e5-9ffdd2b39b83","owner":[],"postedDate":"July 31st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-05-05T10:00:43+00:00","versionOfRecord":{"articleIdentity":"rs-6938965","link":"https://doi.org/10.1186/s12991-026-00664-3","journal":{"identity":"annals-of-general-psychiatry","isVorOnly":false,"title":"Annals of General Psychiatry"},"publishedOn":"2026-05-02 15:57:46","publishedOnDateReadable":"May 2nd, 2026"},"versionCreatedAt":"2025-07-31 15:24:15","video":"","vorDoi":"10.1186/s12991-026-00664-3","vorDoiUrl":"https://doi.org/10.1186/s12991-026-00664-3","workflowStages":[]},"version":"v1","identity":"rs-6938965","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6938965","identity":"rs-6938965","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}