Compromised Prefrontal Cognitive Control Over Emotional Interference in Adolescents with Internet Gaming Disorder

Abstract

Increased reports of impulsivity and aggression in male adolescents with Internet gaming might reflect their dysfunction in emotion regulation, particularly in suppression of negative emotions, which should affect the various stages of Internet gaming disorder. This study tested the hypothesis that adolescents with Internet gaming disorder would be more disturbed by the emotional interference and demonstrate compromised dorsal anterior cingulate cortex (dACC) activation during a Stroop Match-to-Sample task. In addition, functional connectivity analysis was conducted to examine the interplays between neural correlates involved in emotional processing and how they were altered in adolescents with Internet gaming disorder. The Internet gaming disorder group demonstrated weaker dACC activation and stronger insular activations to interfering angry facial stimuli compared with the healthy control group. Negative functional connectivity between stronger insular activation and weaker dorsolateral prefrontal activation correlated with higher cognitive impulsivity in adolescents with Internet gaming disorder. These findings provide evidence of the compromised prefrontal cognitive control over emotional interference in adolescents with Internet gaming disorder.

Introduction

Internet gaming disorder is a pattern of excessive and prolonged Internet gaming that results in a cluster of cognitive and behavioral symptoms, including progressive loss of control over gaming, tolerance, and withdrawal symptoms.¹ Internet gaming disorder, often referred to as Internet addiction, is considered a candidate of behavior addiction.² However, it has not yet been determined whether Internet addiction and substance addiction share a common neural pathway, that is, the disturbance of the mesolimbic reward pathway. Although Internet gaming disorder is one of the emerging social problems in adolescents,^3,4 the neural basis that underlies Internet gaming disorder is still under debate. Specifically, Internet addiction is known to be more prevalent during adolescence, and Internet gaming is the most popular reason for using the Internet in male adolescents, which leads to various social and occupational problems.^5,6

Impaired self-control, defined as the inability to control an impulse, emotion, or temptation, is proposed to be related to symptoms of Internet addiction,^7,8 whereas impulsivity and aggression are distinguishing characteristics of Internet addiction.^7,9 Impulsive aggression and violence are products of failure of emotion regulation. Previous studies have shown that a complex circuit consisting of the prefrontal cortex, amygdala, hippocampus, insular cortex, ventral striatum, and other interconnected regions are implicated in emotion regulation.¹⁰ Taken together, the increased reports of impulsivity and aggression in adolescents with Internet addiction might reflect their dysfunction in emotion regulation, particularly in suppression of negative emotions, which should affect the various stages of Internet addiction.

This study applied an emotional Stroop Match-to-Sample task¹¹ for use in a functional magnetic resonance imaging (fMRI) paradigm to investigate emotional regulation in adolescents with Internet gaming disorder. Previous studies have reported that the dorsal anterior cingulate cortex (dACC) is a key region for top-down cognitive control and selective attention while performing a Stroop task.^12–14 The dACC is the principal locus of conflict monitoring^15,16 and has been implicated in top-down control signaling to adjust behavior and guide performance.^17,18 The present task required subjects to ignore emotional inference and concentrate on a Stroop Match-to-Sample task. Given the role of dACC in cognitive control, the hypothesis that adolescents with Internet gaming disorder would be more disturbed by the emotional interference and demonstrate compromised dACC activation during the Stroop Match-to-Sample task was tested. In addition, functional connectivity analysis was conducted to examine the interplay between the neural correlates that are involved in emotional processing. Functional connectivity refers to the functionally integrated relationship between spatially separated brain regions. Unlike structural connectivity, which looks for physical connections in the brain, functional connectivity measures the temporal correlation of activations in remote brain regions.¹⁹ It was hypothesized that the functional connectivity of the dACC would be disrupted in adolescents with Internet gaming disorder.

Methods

Participants

The study participants consisted of 18 male adolescents with Internet gaming disorder (M_age = 3.6 years, SD = 0.9 years) and 18 age-matched male healthy controls (Table 1). All were recruited from the local community through flyers, announcements, or word of mouth. Adolescents with Internet game disorder were screened according to the Korean Internet Addiction Proneness scale, which consists of 15 items and four subdomains: (a) disturbance of adaptive function, (b) virtual life orientation, (c) withdrawal, and (d) tolerance. The scale indicated a high reliability, with a Cronbach's alpha of 0.838.²⁰ Only male adolescents who reported that online gaming was their main purpose of Internet use were included. Adolescents who reported using the Internet for other purposes, such as social networking or watching videos, were excluded. All participants were administered a structured interview by a clinical research psychologist to confirm the diagnosis according to the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-V). Participants with current or past psychiatric disorders, traumatic brain injury, neurological illness, relevant visual defects, or any radiological contraindications for MRI scanning were excluded. All participants were right-handed and were asked to complete psychometric self-reports, including the Barratt Impulsiveness Scale Version 11 (BIS-11)²¹ and the Aggression Questionnaire.²² The BIS-11²¹ was used to assess impulsivity. It includes subscales of cognitive impulsiveness, motor impulsiveness, and nonplanning impulsiveness. Especially, substance abusers are known to be highly impulsive and score higher on the BIS-11.²³ Aggression was assessed by the Aggression Questionnaire, which contains four subscales: physical aggression, verbal aggression, anger, and hostility.²² This study was carried out under the guidelines for the use of human participants established by the Institutional Review Board at Severance Mental Health Hospital, Yonsei University. Following a complete description of the scope of the study to all participants, written informed consent was obtained.

Table 1.

Demographic and Clinical Characteristics of Subjects

	Internet gaming disorder group	Healthy control group	T	p
Age (years)	13.6 (0.9)	13.4 (1.0)	0.350	0.729
Sex (male)	18	18	—	—
Verbal IQ^a	8.6 (2.6)	10.3 (3.3)	−1.771	0.086
Performance IQ^b	10.4 (3.0)	10.8 (3.3)	−0.420	0.677
BDI^c	11.5 (11.5)	4.4 (4.6)	2.392	0.023
BAI^d	8.9 (9.0)	5.0 (6.2)	1.491	0.145
BIS^e	65.1 (18.4)	56.2 (12.4)	1.700	0.098
Cognitive impulsiveness	18.2 (5.4)	14.8 (3.7)	2.189	0.036
Motor impulsiveness	20.9 (7.2)	18.0 (4.7)	1.428	0.162
Non-planning impulsiveness	26.1 (7.5)	23.4 (5.9)	1.157	0.255
Aggression Questionnaire	76.6 (27.8)	58.6 (11.4)	2.532	0.016
Physical aggression	23.9 (10.7)	16.7 (4.4)	2.674	0.011
Verbal aggression	13.2 (5.5)	12.8 (3.2)	0.296	0.769
Anger	17.2 (6.6)	14.9 (4.9)	1.199	0.239
Hostility	22.2 (9.0)	14.3 (3.8)	3.412	0.002

Values are expressed as mean (SD).

Verbal Intelligence Quotient (IQ) was assessed with the Wechsler Adult Intelligence Scale (Vocabulary).

Performance Intelligence Quotient (IQ) was assessed with the Wechsler Adult Intelligence Scale (Block design).

Beck Depression Inventory.

Beck Anxiety Inventory.

Barret Impulsiveness Scale.

Stimuli and experimental design

Stroop Match-to-Sample Task

The subjects underwent a Stroop Match-to-Sample Task¹² during MRI scanning. Stimuli were created and presented with E-prime software (Psychology Software Tools, Inc.). Subjects matched the color of a “cue (XXXX)” to the meaning of the “target word” or the font color of the “target word” (Fig. 1). The colors of the cue and target words were red, yellow, or blue. The cue's color either matched or did not match the target's color, which was either congruent (the word “RED” was written in red font) or incongruent (the word “BLUE” was written in yellow font). Subjects pressed a YES key for when the cue-target color matched and a NO key for non-matches using their right hand, yielding accuracy and reaction time (RT) measures. To test the effect of emotional interference on cognitive control and selective attention, the trials were designed in two different conditions. In trials with emotional interference, angry faces were presented between the cue and target word during the inter-stimulus interval for 1,700 milliseconds, whereas trials without emotional interference presented gray squares during the inter-stimulus interval. The angry faces were six grayscale images of emotionally expressive faces selected from a standard set of pictures of facial affect.²⁴ The total duration of each trial was 4,400 milliseconds, and the inter-trial interval was 1,200 milliseconds. The task was composed of eight blocks with emotional interference and eight blocks without emotional interference, which were presented in a pseudo-random order. The total number of trials was 96 (1 block = 6 trials, duration of each block = 26.4 milliseconds). The subject performed a practice session before entering the scanner. The percentage of correct response (accuracy) and reaction time data were recorded for each trial and subject. Subjects had to make at least 80% correct responses to be included for further analyses.

FIG. 1.

Stroop Match-to-Sample task. In trials without emotional interference (A) gray squares were presented during the inter-stimulus interval. In trials with emotional interference (B), angry faces were presented between the cue and target word. Color images available online at www.liebertpub.com/cyber

Image acquisition

MR imaging was conducted on a 3T Siemens Magnetom MRI scanner equipped with an eight-channel head coil. Subject motion was minimized by following best practice for head fixation, and structural image series were inspected for residual motion. Whole-brain fMRI data were acquired with a T2*-weighted gradient echo planar pulse sequence (TE = 30 milliseconds, TR = 2,200 milliseconds, flip angle = 90°, field of view = 240 millimeters, matrix = 64 × 64, slice thickness = 4 millimeters) during the Stroop task. High-resolution anatomical images were acquired with a T1-weighted spoiled gradient echo sequence (TE = 2.19 milliseconds; TR = 1,780 milliseconds, flip angle = 9°, field of view = 256 mm, matrix = 256 × 256, slice thickness = 1 mm) to serves as an anatomical underlay for the fMRI data.

Image pre-processing and fMRI contrast analysis

Spatial preprocessing and statistical analysis of functional images were performed using SPM8 (Wellcome Trust Centre for Neuroimaging; www.fil.ion.ucl.ac.uk/spm). Motion artifacts were assessed in individual subjects by visual inspection of realignment parameter estimations to confirm that the maximum head motion in each axis was <3 millimeters, without any abrupt head motion. The anatomical volume was then segmented into gray matter, white matter, and cerebrospinal fluid. The gray matter image was used for determining the parameters of normalization onto the standard Montreal Neurological Institute (MNI) gray matter template. The spatial parameters were then applied to the realigned and unwarped functional volumes that were finally resampled to voxels of 2 × 2 × 2 millimeters and smoothed with an 8 millimeter full-width at half-maximum kernel.

Individual statistics were computed using a general linear model approach as implemented in SPM8. In order to reveal the activity of the blood oxygenation level dependent (BOLD) responses related to the emotional interference, individual contrast were generated by contrasting “angry face” blocks with “gray plate” blocks at the first-level analysis. The resulting set of contrast images were then entered into a second-level analysis using a full factorial model. To compare the contrast angry face > gray square between groups, exclusive masks were created for each group with SPM maps with a threshold at p < 0.05. Statistical inferences were thresholded using an uncorrected p < 0.001, k_E > 50 voxels for the whole brain.

Functional connectivity analyses

The CONN-fMRI functional connectivity toolbox (www.nitrc.org/projects/conn) was used to create individual seed-to-voxel functional connectivity maps. A right fronto-insula seed consisted of 5 millimeter radium sphere centered on coordinates (x = 46, y = 4, z = 14) that was identified in the contrast angry face > gray plate. Before averaging individual voxel data, the waveform of each brain voxel was filtered using a bandpass filter (0.008 Hertz < f < 0.09 Hertz) to reduce the effect of low-frequency drift and high-frequency noise. Signals from the ventricular regions and signal from the white matter were removed from the data through linear regression. To compare the functional connectivity between groups, exclusive masks were created for each group with SPM maps with a threshold at p < 0.05. Statistical inferences were thresholded using an uncorrected p < 0.001, k_E>50 voxels for the whole brain. The seed-to voxel functional conductivity maps were created according to both positive correlations and negative correlations individually.

Statistical analysis

For between-group comparisons, behavioral performances were conducted with repeated-measures analyses of variance (ANOVA). The ANOVAs were defined by one repeated-measures factor (with emotional interference vs. without emotional interference), and one between-subjects factor (Internet gaming disorder group vs. healthy control group).

To examine brain–behavior relationships, the functional connectivity strength was calculated by extracting the Fisher-transformed Z-value measures of functional connectivity between the right insula seed and the target regions that were identified through functional connectivity analysis of the task runs. Pearson correlation analysis tested relations between the functional seed-target connectivity and behavioral performance. Statistical analyses were conducted using SPSS 18.0 Statistics for Windows, with p < 0.05 (two-tailed).

Results

Behavioral performance: accuracy and reaction time

The behavioral performances are presented in Table 2. Repeated-measures ANOVA revealed no significant difference between the two groups in accuracy, F(1, 34) = 0.12, p = 0.913. There was no significant effect of condition (emotional interference; F(1, 34) = 1.235, p = 0.274, and the group–condition interaction was not significant, F(1, 34) = 1.089, p = 0.304.

Table 2.

Behavioral Performance of Subjects

	Internet gaming disorder group	Healthy control group	T	p
Accuracy (%)
Without emotional interference^a	93.9 ± 4.7	92.5 ± 6.0	0.723	0.475
With emotional interference^a	91.3 ± 5.6	92.4 ± 7.1	−0.488	0.629
Reaction time (milliseconds)
Without emotional interference	501.6 ± 80.4	510.4 ± 80.4	−0.363	0.719
With emotional Interference	534.9 ± 82.9	514.3 ± 58.9	0.860	0.396

In trials with emotional interference, angry faces were presented between the cue and target word during the inter-stimulus interval. By contrast, in trials without emotional interference, gray squares were presented during the inter-stimulus interval.

Repeated-measures ANOVA revealed no significant difference between the two groups in reaction time, F(1, 34) = 0.072, p = 0.791. There was no significant effect of condition, F(1, 34) = 3.326, p = 0.077, and the group–condition interaction was not significant, F(1, 34) = 2.090, p = 0.157.

Brain activations associated with emotionally interfering angry faces

The healthy control group demonstrated significant BOLD responses to emotionally interfering angry faces in the dACC, primary motor cortex, superior temporal sulcus, and posterior parietal cortex. By contrast, the Internet gaming disorder group showed significant BOLD responses to emotionally interfering angry faces in the anterior insula, supplementary motor area, and fusiform gyrus (Table 3 and Fig. 2).

FIG. 2.

Brain activations associated with emotionally interfering angry faces. Green: brain activations stronger in healthy control group than in the Internet gaming disorder group. Red: brain activations stronger in the Internet gaming disorder group than in the healthy control group. Color images available online at www.liebertpub.com/cyber

Table 3.

Brain Activations ^a Associated with Emotionally Interfering Angry Faces

Region		BA	k_E	T_max	x	y	z
Healthy control group > Internet gaming disorder group
Anterior cingulate cortex		24	110	3.64	0	22	40
Supplementary motor area	Left	6	60	3.80	−26	−6	46
Dorsolateral prefrontal cortex	Left	4	91	3.92	−52	−6	34
Superior temporal gyrus	Right	22	62	3.54	66	−8	10
Posterior cingulate cortex	Left	23	503	4.78	−6	−26	28
Posterior parietal cortex	Right	7	103	4.00	8	−82	48
Internet gaming disorder group > Healthy control group
Anterior insula	Right	48	57	4.18	46	4	14
Supplementary motor area	Right	6	115	4.64	2	−6	68
	Left	6	404	4.70	−18	−10	72
Fusiform gyrus	Left	37	81	3.86	−44	−60	−22

To compare the contrast (angry face> gray square) between groups, exclusive masks were created for each group with SPM maps with a threshold at p < 0.05. Statistical inferences were thresholded using an uncorrected p < 0.001, k_E > 50 voxels for the whole brain.

Functional connectivity between the right insula and the dorsal attention network

The insula activations synchronized with activations negatively in the dorsolateral prefrontal cortex, middle temporal gyrus, cerebellum, and posterior parietal cortex in the Internet gaming disorder group (Table 4 and Fig. 3A).

FIG. 3.

Brain–behavior relationship in the Internet gaming disorder group. (A) Brain regions exhibiting negative functional connectivity with the right insula activation. (B) Correlation between insular-prefrontal functional connectivity and reaction time. (C) Correlation between insular-prefrontal functional connectivity and cognitive impulsivity. Color images available online at www.liebertpub.com/cyber

Table 4.

Brain Regions ^a Exhibiting Negative Functional Connectivity with the Right Insula in Internet Gaming Disorder Group

Region		BA	k_E	T_max	x	y	z
Dorsolateral prefrontal cortex	Right	9	59	4.71	16	38	54
			61	7.10	42	18	58
	Left		202	5.08	−36	18	56
Middle temporal gyrus	Left	21	261	5.43	−50	−2	−24
	Right		558	7.74	68	−26	−12
Fusiform gyrus	Left	37	64	5.66	−44	−40	−6
Posterior parietal cortex	Left	39	196	6.00	−44	−68	42
	Right	7	88	4.54	40	−72	50
Cerebellum	Right		146	4.61	40	−66	−50
	Left		376	6.02	−34	−66	−42
	Left		204	6.51	−6	−82	−34

Statistical inferences were thresholded using an uncorrected p < 0.001, k_E > 50 voxels for the whole brain.

Correlation between functional connectivity and behavioral performance

In the Internet gaming disorder group, delayed reaction time correlated with stronger negative functional connectivity between right insula and right dorsolateral prefrontal cortex (Pearson's r = −0.609, p = −0.007; Fig. 3B). A higher cognitive impulsivity score also correlated with stronger negative functional connectivity between the right insula and right dorsolateral prefrontal cortex (Pearson's r = −0.467, p = −0.051; Fig. 3C).

Discussion

This study showed that the prefrontal cognitive control of emotional interference was compromised in adolescents with Internet gaming disorder. The adolescents with Internet gaming disorder demonstrated weaker dACC activation and stronger insular activations to interfering angry facial stimuli, which implied difficulty in emotional regulation. Reciprocal interaction between stronger insular activation and weaker dorsolateral prefrontal activation correlated with higher cognitive impulsivity in adolescents with Internet gaming disorder.

These results support the hypothesis that the dACC activation during the Stroop Match-to-Sample task would be compromised in adolescents with Internet gaming disorder. As a key region for top-down cognitive control and selective attention,^25,26 stronger dACC activation was demonstrated in the healthy control group compared with the Internet gaming disorder group. In addition, the dorsolateral prefrontal cortex and posterior parietal cortex, which make up the dorsal attention network,²⁷ were significantly activated in the healthy control group compared with the Internet gaming disorder group. The dorsal attention network is involved in the cognitive selection of sensory information and responses, and is thought to generate and maintain endogenous signals based on current goals.^27,28 The dACC is a part of the salience network, which initiates switching between other large-scale brain networks to facilitate access to attention and working memory resources when a salient event is detected.²⁹ Therefore, it is speculated that the weaker activation of the dorsal attention network in the Internet gaming disorder group should be linked to the weaker activation of the dACC.

The adolescents with Internet gaming disorder showed stronger activation in the right anterior insula and the fusiform gyrus. The anterior insula, which is the hub of the ventral attention system, is proposed to function as a link between attention-related problem solving and salience systems during the coordination of task performance.³⁰ In contrast to the dorsal attention network, the ventral attention network makes important contributions to stimulus-driven reorienting of covert visual spatial attention.²⁸ Most importantly, stimulus-driven ventral attention network interferes with the top-down dorsal attention network as a “circuit breaker” during task performance coordination.^27,31 Therefore, the stronger activation of the right anterior insula could be related to the weaker activation of the dorsal attention network in the Internet gaming disorder group. Taken together, it is assumed that the adolescents with Internet gaming disorder were more distracted by the emotional interference of the angry facial stimuli. From the same viewpoint, it is noteworthy that fusiform gyrus activation was demonstrated in the Internet gaming disorder group. Previous studies suggest that the fusiform gyrus plays a crucial role in facial recognition, especially invariant aspects of faces.^32,33 Compared with the healthy control group, the activation in the fusiform gyrus indicates the failure of the Internet gaming group to ignore the emotional interfering angry faces during the Stroop Sample-to-Match task.

The speculations that angry face stimulus-driven attention disrupted goal-directed attention of the Internet gaming disorder group were supported the functional connectivity analysis, which demonstrated reciprocal interaction between stronger insular activation and weaker dorsolateral prefrontal activation. In addition, stronger negative functional connectivity between insular and dorsolateral prefrontal activation correlated with delayed reaction time and with a higher cognitive impulsivity score. Although adolescents with Internet gaming disorder demonstrated a tendency for longer reaction times in emotional interfering conditions, the difference was not significant. Therefore, the findings show that the changes of the cortical activation pattern might precede noticeable changes of the behavioral performance in adolescents with Internet gaming disorder.

These findings are in line with previous studies,^7,9 which report a higher incidence of impulsivity and aggression in adolescents with Internet gaming disorder. However, other factors that account for the impulsive and aggressive behaviors in adolescents with Internet addiction should be considered. For instance, top-down control of cortical inhibition is also reduced in attention deficit hyperactivity disorder, which is one of the most common comorbidities in adolescents with Internet addiction.³⁴ The influence of combined psychiatric disorders should be evaluated in further studies and should be treated in order to prevent their deleterious effect on the prognosis of adolescents with Internet gaming disorder. Another factor to consider should be the content and type of Internet game. A meta-analytic review reported that violent video games (in which the predominant goal is to harm another game character) increase aggression, whereas prosocial video games (in which the predominant goal is to benefit another game character) have the opposite effect.³⁵ Among Korean adolescents, role playing games, first-person shooting games, real-time strategy games, and sports games are the most popular types of Internet games.³⁶ In the present study, the majority of participants reported enjoying role playing PC games, especially massively multiplayer online role-playing games (MMORPGs). It was not possible to examine the effect of the game type on the behavioral responses or brain activations.

There are several limitations to this study. First, there were no statistically significant differences in behavioral results between groups. Studies with larger populations should be followed to confirm the trend of these results. Second, diverse results were not seen following different emotional stimuli because only an angry face was used as emotional interruption. The effects of different emotions, such as fear, on these results are a matter of debate. Third, the data showed significant difference of Beck Depression Inventory (BDI) score between adolescents with Internet gaming disorder and the healthy control group. This might reflect the comorbidity of Internet gaming disorder and depression, which has been reported in previous studies.^34,37

In conclusion, adolescents with Internet gaming disorder demonstrated weaker dACC activation and stronger insular activations during a Stroop task, which implied compromised prefrontal cognitive control over emotional interference. “Unsuccessful attempts to control the behavior” is one of the main criteria of Internet gaming disorder according to the DSM-5. The present findings showed that this impaired prefrontal control was not limited to Internet use, but could be applied to broader areas of emotion regulation.

Footnotes

Acknowledgments

This study was supported by a grant of the Korean Mental Health Technology R&D Project, Ministry of Health & Welfare, Korea (HM14C2578).

Author Disclosure Statement

No competing financial interests exist.

References

American Psychiatric Association. (2013) Diagnostic and statistical manual of mental disorders: DSM-5. 5th ed. Washington, DC: American Psychiatric Association.

Grant

, Potenza

, Weinstein

, et al. Introduction to behavioral addictions. American Journal of Drug & Alcohol Abuse, 2010; 36:233–241.

Grusser

, Thalemann

, Griffiths

. Excessive computer game playing: evidence for addiction and aggression?. CyberPsychology & Behavior, 2007; 10:290–292.

Kim

, Han

, Park

, et al. The relationships between online game player biogenetic traits, playing time, and the genre of the game being played. Psychiatry Investigation, 2010; 7:17–23.

Mak

, Lai

, Watanabe

, et al. Epidemiology of Internet behaviors and addiction among adolescents in six Asian countries. Cyberpsychology, Behavior, & Social Networking, 2014; 17:720–728.

Müller

, Janikian

, Dreier

, et al. Regular gaming behavior and Internet gaming disorder in European adolescents: results from a cross-national representative survey of prevalence, predictors, and psychopathological correlates. European Child & Adolescent Psychiatry, 2015; 24:565–574.

Kim

, Namkoong

, Ku

, et al. The relationship between online game addiction and aggression, self-control and narcissistic personality traits. European Psychiatry, 2008; 23:212–218.

Park

, Park

, Lee

, et al. The effects of behavioral inhibition/approach system as predictors of Internet addiction in adolescents. Personality & Individual Differences, 2013; 54:7–11.

Lim

, Gwak

, Park

, et al. Are adolescents with Internet addiction prone to aggressive behavior? The mediating effect of clinical comorbidities on the predictability of aggression in adolescents with Internet addiction. Cyberpsychology, Behavior, & Social Networking, 2015; 18:260–267.

10.

Davidson

, Putnam

, Larson

. Dysfunction in the neural circuitry of emotion regulation—a possible prelude to violence. Science, 2000; 289:591–594.

11.

Schulte

, Mueller-Oehring

, Rosenbloom

, et al. Differential effect of HIV infection and alcoholism on conflict processing, attentional allocation, and perceptual load: evidence from a Stroop Match-to-Sample task. Biological Psychiatry, 2005; 57:67–75.

12.

Bush

, Frazier

, Rauch

, et al. Anterior cingulate cortex dysfunction in attention-deficit/hyperactivity disorder revealed by fMRI and the Counting Stroop. Biological Psychiatry, 1999; 45:1542–1552.

13.

Peterson

, Skudlarski

, Gatenby

, et al. An fMRI study of Stroop word–color interference: evidence for cingulate subregions subserving multiple distributed attentional systems. Biological Psychiatry, 1999; 45:1237–1258.

14.

Harrison

, Shaw

, Yucel

, et al. Functional connectivity during Stroop task performance. Neuroimage, 2005; 24:181–191.

15.

Carter

, Braver

, Barch

, et al. Anterior cingulate cortex, error detection, and the online monitoring of performance. Science, 1998; 280:747–749.

16.

Botvinick

, Braver

, Barch

, et al. Conflict monitoring and cognitive control. Psychological Review, 2001; 108:624.

17.

Ridderinkhof

, Ullsperger

, Crone

, et al. The role of the medial frontal cortex in cognitive control. Science, 2004; 306:443–447.

18.

Venkatraman

, Rosati

, Taren

, et al. Resolving response, decision, and strategic control: evidence for a functional topography in dorsomedial prefrontal cortex. Journal of Neuroscience, 2009; 29:13158–13164.

19.

Friston

. Functional and effective connectivity: a review. Brain Connectivity, 2011; 1:13–36.

20.

Shin

, Kim

, Jeong

. (2011) Third standardization of Korea Internet Addiction Proneness Scale. Seoul, Korea: National Information Society Agency.

21.

Patton

, Stanford

, Barratt

. Factor structure of the Barratt Impulsiveness Scale. Journal of Clinical Psychology, 1995; 51:768–774.

22.

Buss

, Perry

. The aggression questionnaire. Journal of Personality & Social Psychology, 1992; 63:452–459.

23.

Stanford

, Mathias

, Dougherty

, et al. Fifty years of the Barratt Impulsiveness Scale: an update and review. Personality & Individual Differences, 2009; 47:385–395.

24.

Park

, Oh

, Kim

et al., Korean facial expressions of emotion (KOFEE). Section of Affect & Neuroscience, Institute of Behavioral Science in Medicine, Yonsei University College of Medicine, Seoul, South Korea, 2011.

25.

Petersen

, Posner

. The attention system of the human brain: 20 years after. Annual Review of Neuroscience, 2012; 35:73–89.

26.

Bush

, Luu

, Posner

. Cognitive and emotional influences in anterior cingulate cortex. Trends in Cognitive Sciences, 2000; 4:215–222.

27.

Corbetta

, Shulman

. Control of goal-directed and stimulus-driven attention in the brain. Nature Reviews Neuroscience, 2002; 3:201–215.

28.

Corbetta

, Kincade

, Ollinger

, et al. Voluntary orienting is dissociated from target detection in human posterior parietal cortex. Nature Neuroscience, 2000; 3:292–297.

29.

Menon

, Uddin

. Saliency, switching, attention and control: a network model of insula function. Brain Structure & Function, 2010; 214:655–667.

30.

Eckert

, Menon

, Walczak

, et al. At the heart of the ventral attention system: the right anterior insula. Human Brain Mapping, 2009; 30:2530–2541.

31.

Weissman

, Prado

. Heightened activity in a key region of the ventral attention network is linked to reduced activity in a key region of the dorsal attention network during unexpected shifts of covert visual spatial attention. Neuroimage, 2012; 61:798–804.

32.

Calder

, Young

. Understanding the recognition of facial identity and facial expression. Nature Reviews Neuroscience, 2005; 6:641–651.

33.

Jehna

, Neuper

, Ischebeck

, et al. The functional correlates of face perception and recognition of emotional facial expressions as evidenced by fMRI. Brain Research, 2011; 1393:73–83.

34.

, Yen

, et al. The association between Internet addiction and psychiatric disorder: a review of the literature. European Psychiatry, 2012; 27:1–8.

35.

Greitemeyer

, Mügge

. Video games do affect social outcomes: a meta-analytic review of the effects of violent and prosocial video game play. Personality & Social Psychology Bulletin, 2014; 40:578–589.

36.

Korean National Information Society Agency. www.nia.or.kr/bbs/board_view.asp?boardid=201408061323065914&Order=020403&id=15626 (accessed May 14, 2015).

37.

Morrison

, Gore

. The relationship between excessive Internet use and depression: a questionnaire-based study of 1,319 young people and adults. Psychopathology, 2010; 43:121–126.