Neural classification of internet gaming disorder and prediction of treatment response using a cue-reactivity fMRI task in young men. (January 2022)

• Record Type: Journal Article
• Authors: Wang, Zi-Liang; Potenza, Marc N.; Song, Kun-Ru; Fang, Xiao-Yi; Liu, Lu; Ma, Shan-Shan; Xia, Cui-Cui; Lan, Jing; Yao, Yuan-Wei; Zhang, Jin-Tao
• Is Part Of: Journal of psychiatric research. Volume 145(2022)
• Page Start: 309
• Page End: 316
• Publication Date: 2022-01
• Subjects: Internet -- Video games -- Craving -- Multi-voxel pattern analysis -- Addictive behaviors -- Behavior therapy; Psychiatry -- Periodicals; Mental Disorders -- Periodicals; Maladies mentales -- Périodiques; Psychiatry; Electronic journals; Periodicals; 616.89005
• DOI: 10.1016/j.jpsychires.2020.11.014
• Languages: English
• ISSNs: 0022-3956

Abstract

Abstract: Background: Neural mechanisms underlying internet gaming disorder (IGD) are important for diagnostic considerations and treatment development. However, neurobiological underpinnings of IGD remain relatively poorly understood. Methods: We employed multi-voxel pattern analysis (MVPA), a machine-learning approach, to examine the potential of neural features to statistically predict IGD status and treatment outcome (percentage change in weekly gaming time) for IGD. Cue-reactivity fMRI-task data were collected from 40 male IGD subjects and 19 male healthy control (HC) subjects. 23 IGD subjects received 6 weeks of craving behavioral intervention (CBI) treatment. MVPA was applied to classify IGD subjects from HCs and statistically predict clinical outcomes. Results: MVPA displayed a high (92.37%) accuracy (sensitivity of 90.00% and specificity of 94.74%) in the classification of IGD and HC subjects. The most discriminative brain regions that contribute to classification were the bilateral middle frontal gyrus, precuneus, and posterior lobe of the right cerebellum. MVPA statistically predicted clinical outcomes in the craving behavioral intervention (CBI) group ( r = 0.48, p = 0.0032). The most strongly implicated brain regions in the prediction model were the right middle frontal gyrus, superior frontal gyrus, supramarginal gyrus, anterior/posterior lobes of the cerebellum and left postcentral gyrus. Conclusions: The findings about cue-reactivity neural correlates could help identify IGD subjects and predict CBI-related treatment outcomes provide mechanistic insight into IGD and its treatment and may help promote treatment development efforts. Highlights: The present study is the first to investigate the utility of MVPA in analyzing cue-reactivity data. Cue-reactivity neural correlates could help identify IGD subjects. Cue-reactivity neural correlates could help predict CBI-related treatment outcomes.