Accepted Manuscript Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of negative consequences Guangheng Dong, Marc N. Potenza PII:
S0022-3956(15)30011-X
DOI:
10.1016/j.jpsychires.2015.11.011
Reference:
PIAT 2767
To appear in:
Journal of Psychiatric Research
Received Date: 21 August 2015 Revised Date:
14 October 2015
Accepted Date: 19 November 2015
Please cite this article as: Dong G, Potenza MN, Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of negative consequences, Journal of Psychiatric Research (2015), doi: 10.1016/j.jpsychires.2015.11.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of
2
Department of Psychology, Zhejiang Normal University, Jinhua, Zhejiang Province, P.R. China Department of Psychiatry, Department of Neurobiology, Child Study Center, and CASAColumbia, Yale
M AN U
1
SC
Guangheng Dong 1 *, Marc N. Potenza 2
RI PT
negative consequences
University School of Medicine, New Haven, CT, USA
Corresponding author:
TE D
*
Guangheng Dong, Ph.D. Professor
Department of Psychology, Zhejiang Normal University, 688 Yingbin Road, Jinhua,
EP
Zhejiang Province 321004, PR China. Tel.: +86 15867949909.
AC C
E-mail address: [email protected]
1
ACCEPTED MANUSCRIPT
Abstract Individuals with Internet gaming disorder (IGD) continue gaming despite adverse consequences. However, the precise mechanism underlying this behavior remains
RI PT
unknown. In this study, data from 20 IGD subjects and 16 otherwise comparable healthy control subjects (HCs) were recorded and compared when they were
undergoing risk-taking and risky decision-making during functional magnetic
SC
resonance imaging (fMRI). During risk-taking and as compared to HCs, IGD subjects selected more risk-disadvantageous trials and demonstrated less activation of the
M AN U
anterior cingulate, posterior cingulate and middle temporal gyrus. During risky decision-making and as compared to HCs, IGD subjects showed shorter response times and less activations of the inferior frontal and superior temporal gyri. Taken together, data suggest that IGD subjects show impaired executive control in
TE D
selecting risk-disadvantageous choices, and they make risky decisions more hastily and with less recruitment of regions implicated in impulse control. These results suggest a possible neurobiological underpinning for why IGD subjects may exhibit
EP
poor control over their game-seeking behaviors even when encountering negative
AC C
consequences and provide possible therapeutic targets for interventions in this population.
Keywords: Internet gaming disorder; risk-taking; risky decision-making
2
ACCEPTED MANUSCRIPT
Introduction Internet gaming disorder (IGD) is a public health concern and a disorder that is
RI PT
included in Section 3 of the DSM-5 as a condition warranting further study
(American Psychiatric Association, 2013, Griffiths et al. , 2014). IGD shares features with substance (Dong et al. , 2015a, Dong et al. , 2014, Dong et al. , 2013d, Hare et al.
SC
, 2014, Lin et al. , 2015b) and gambling (Potenza, 2014a, b) addictions. Unlike drug addiction or substance abuse, no chemical or substance intake is involved in IGD,
M AN U
although excessive Internet gaming may lead to physical dependence, similar to other addictions (Dong et al. , 2013a, Holden, 2001). These findings suggest that online experiences may change brain function and related cognitive processes in manners that may perpetuate Internet gaming (Dong et al. , 2011, Holden, 2001,
TE D
Weinstein and Lejoyeux, 2010).
Risk-taking and risky decision-making contribute importantly to the development
EP
of addictions (Balogh et al. , 2013). Disadvantageous risk-taking and improper
AC C
decision-making may lead to problematic behaviors such as substance abuse (Rutherford et al. , 2010), pathological gambling (Potenza, 2014b), and IGD (Dong et al. , 2015b). Reduced cognitive capacity or willingness to avoid excessive behavioral engagement in pleasurable activities may contribute to the development of various clinical problems, including behavioral and substance addictions (Potenza et al. , 2013). Individuals with IGD may not fully consider outcomes when making decisions (Bechara et al. , 2002, Dong et al. , 2013b, Floros and Siomos,
3
ACCEPTED MANUSCRIPT
2012, Pawlikowski and Brand, 2011). Individuals with IGD may display a ‘myopia for the future’ in which they tend to pursue immediately rewarding experiences (e.g., playing online) and neglect long-term adverse consequences, as has been
RI PT
found in drug addictions (Bechara, Dolan, 2002, Floros and Siomos, 2012, Pawlikowski and Brand, 2011). Although studies have demonstrated
disadvantageous decision-making in association with IGD (Dong, Lin, 2015a, Lin et
SC
al. , 2015a), unanswered questions exist regarding the precise mechanisms that may lead individuals to game excessively or compulsively. This study aimed to
M AN U
examine the neural features underlying decision-making in IGD during performance of a risk-taking and risky decision-making task. Specific brain regions, including the inferior frontal gyrus (IFG) and orbitofrontal cortex (OFC), are involved in decision-making (Cazzell et al. , 2012, Rushworth et al.
TE D
, 2012, Sheth et al. , 2012). IFG activation may signal subjective risk and contribute to the formation of subjective feelings during decision-making (Craig, 2009). The OFC contributes importantly to value-based decision-making (Glascher et al. , 2012,
EP
Kable and Glimcher, 2009). In addiction, these prefrontal cortex regions and
AC C
associated brain structures have been implicated in the development and maintenance of problematic patterns of Internet use (Brand et al. , 2014, Dong and Potenza, 2014). A meta-analysis suggests that dysfunction of or anatomic deficits in the frontal cortex contribute to impaired impulse control (Meng et al. , 2015). Thus, in current study, we hypothesized that IGD subjects would show impaired decisionmaking that would relate to activations in these brain regions. Performance of risk-taking and decision-making tasks involves evaluation of risky
4
ACCEPTED MANUSCRIPT
features and avoidance of disadvantageous choices (Rothman and Salovey, 1997). In pathological gambling and substance-use disorders, reduced activation of reward-related circuitry is observed during gambling-related decision-making (de
RI PT
Ruiter et al., 2009; Reuter et al., 2005, Tanabe et al. , 2007). Thus, in the current
study, we hypothesized that IGD relative to healthy control subjects (HCs) would
of reward-related fronto-striatal brain regions.
SC
show disadvantageous decision-making that would relate to diminished activation
M AN U
During risk-taking or risky decision-making, executive inhibition contributes to better impulse control and advantageous decision-making. Some brain regions, such as the anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (DLPFC), contribute importantly in this regard. The ACC has been associated with error
TE D
monitoring, conflict detection and performance monitoring in decision-making (Holroyd and Coles, 2002, Platt and Huettel, 2008) and is involved in anticipating risks, especially potential losses (Krawitz et al. , 2010). The posterior cingulate
EP
cortex (PCC), and DLPFC were found to be more active during choice of risky versus
AC C
safe options (Paulus et al. , 2003, Schonberg et al. , 2011). Thus, we hypothesized that IGD versus HC subjects would show disadvantageous decision-making that would relate to activations in executive-function-related cortical brain regions.
Methods and Materials Participant Selections
5
ACCEPTED MANUSCRIPT
The experiment conforms to The Code of Ethics of the World Medical Association (Declaration of Helsinki). The Human Investigations Committee of Zhejiang Normal University approved this research. The methods were conducted in accordance with
RI PT
the approved guidelines. Participants were university students and were recruited through advertisements. Participants were right-handed males (20 IGD subjects, 16 HCs). IGD and HC groups did not significantly differ in age (IGD mean=21.33,
SC
SD=2.18 years; HC mean=21.90, SD=2.33 years; t= 0.66, p =0.48). All subjects had
normal or corrected to normal vision. Only males were included due to higher IGD
M AN U
prevalence in men than in women. All participants provided written informed consent and completed a structured psychiatric interviews (MINI) (Lecrubier et al. , 1997) that was performed by an experienced psychiatrist lasted approximately 15 minutes. All participants were free of Axis I psychiatric disorders assessed in the
TE D
MINI. We further assessed ‘depression’ with the Beck Depression Inventory (Beck et al. , 1961) and only participants scoring less than 5 were included. All participants were instructed not to use any substances of abuse, including caffeinated drinks, on
EP
the day of scanning. No participants reported previous use of illicit drugs (e.g.,
AC C
cocaine, marijuana).
Internet gaming disorder was determined based on scores of 50 or more on Young’s online Internet addiction test (IAT) (Young, 2009) and, at the same time, meeting a proposed IGD diagnosis per DSM-5 criteria (Petry et al. , 2014). Young's IAT consists of 20 items assessing problematic Internet use, including psychological dependence, compulsive use, withdrawal, problems in school or work, sleep, family or time
6
ACCEPTED MANUSCRIPT
management (Young, 2009). The IAT is a valid and reliable instrument that can be used in classifying Internet addiction (Widyanto et al. , 2011, Widyanto and McMurran, 2004). For each item, a graded response is selected from 1 = “Rarely” to
RI PT
5 = “Always”, or “Does not Apply”. Scores over 50 indicate occasional or frequent
internet-related problems (www.netaddiction.com). When recruiting IGD subjects, we added an additional criterion on Young’s established measures of IAT, ‘you
SC
spend ___% of your online time playing online games’ (>80%).
M AN U
Task and Procedure
The fMRI task used an event-related design. This task consisted of 80 trials. Each trial was divided into three stages: Decision stage (risk-taking), gamble stage (risky
TE D
decision-making), and feedback stage. Figure 1a shows the event sequence of each trial during the task. First, a white cross was presented at the center of a black screen for 500ms to cue the beginning of a new trial. During the subsequent risk-
EP
taking stage, participants were asked to select one from of two risky options (see
AC C
details on decision stage in Figure 1b). This selection process lasted for 4000 ms at most and disappeared once the participant made a selection. After a variable period of delay (mean 3000ms, ranging from 1000 to 5000ms), the risky decision-making stage followed (Figure 1c). During this stage, participants would see 4 backs of cards and were asked to guess which one was red and indicate their response by a button press within 2000 ms (the order of the cards during the gamble stage was randomized). If they missed, they would lose 15 Chinese Yuan (about $2.5 USD).
7
ACCEPTED MANUSCRIPT
After the response and a delay ranging from 1000-3000 ms (mean=2000 ms), the selected card would turn over and show participants the outcome, which was presented for a period of 1000 ms. Participants won/lost the amount according to
RI PT
the card color and the number on the card. The next trial begun after a jittered delay (mean 3000 ms, ranged from 2000-4500 ms). Subjects were asked to make
software (Psychology Software Tools, Inc.).
M AN U
Risk–taking
SC
responses with their right hands. The experiment was presented using E-prime
During the risk-taking stage, two lines of cards (each line consisting of 4 cards) were presented on the computer screen (see Figure 1b), with the red cards and the amount on it suggesting winning some amount, and the yellow ones suggesting
TE D
losing some amount. The cards were shown in colors to indicate the results (red, win; yellow, lose), win/lose rates (the proportions of different color cards), and win/lose amount (the number on cards).
EP
The probability and magnitude of the gains/losses were manipulated to create
AC C
advantageous/disadvantageous risky selections. The advantageous risk-taking means the sum of numbers on red cards (win) are larger than those on yellow cards (loss) (risk-advantageous). It suggests that although participants have opportunities to lose a big amount, they are more likely to win money in the long run. In Figure 1b, the first line is risk-advantageous selection ((45*0.75)+(-35*0.25) = 25 Yuan). In contrast, the second line in Figure 1b is risk-disadvantageous: the numbers on red card (win) are smaller than the sum in all yellow cards
8
ACCEPTED MANUSCRIPT
((35*0.25)+(-45*0.75) = -25 Yuan). It suggests that although participants have opportunities to win a big amount, they are more likely to lose money in the long
Insert Figure 1 about here
RI PT
run. Participants practiced using the same task before fMRI.
SC
Participants were told they had 80 opportunities to win some money and trials were presented in a random order. Each participant was provided with 200 Chinese Yuan
M AN U
as the initial balance before the task and was explicitly informed that he would receive the entire balance in cash at the end of the task.
Participants who chose the same selections for more than 80 percent of all trials
TE D
(who might have selective bias) or chose the selections using the same button 10 or more times consecutively (who might lack motivation to perform properly) were excluded from further analysis. Participants who took less than 10 trials in one of
EP
these conditions were excluded from further analysis to keep the statistical power.
AC C
No subjects were excluded in the risk-taking stage. However, six subjects (4 IGD, 2 HC) were excluded in the gamble stage based on these criteria.
Image acquisition and pre-processing Scanning was performed in the Shanghai Key Laboratory of Magnetic Resonance, East China Normal University. Structural images covering the whole brain were collected, using a T1-weighted three-dimensional spoiled gradient-recalled
9
ACCEPTED MANUSCRIPT
sequence (176 slices, TR=1700 ms, echo time (TE)=3.93 ms, slice thickness=1.0 mm, skip=0 mm, flip angle=15°, inversion time= 1100 ms, field of view (FOV)=240×240 mm, in-plane resolution=256×256). Functional MRI was performed on a 3T scanner
RI PT
(Siemens Trio) with a gradient-echo EPI T2 sensitive pulse sequence in 33 slices
(interleaved sequence, 3mm thickness, TR=2000ms, TE=30ms, flip angle =90°, field of view =220 × 220 mm2, matrix =64 ×64). Stimuli were presented using Invivo
SC
synchronous system (Invivo Company, www.invivocorp.com/) through a screen in the head coil, enabling participants to view the stimuli. A total of 630 volumes was
First-level regression analysis
M AN U
acquired for each participant during the 1260 seconds of task performance.
The functional data were analyzed using SPM8 and Neuroelf (http://neuroelf.net) as
TE D
described previously (DeVito et al. , 2012, Dong, Hu, 2013b, Krishnan-Sarin et al. , 2013). Images were slice-timed, corrected, reoriented (manually), and realigned to the first volume. T1-co-registered volumes were normalized to a MNI template and
EP
spatially smoothed with a 6mm FWHM Gaussian kernel. A general linear model
AC C
(GLM) was applied to identify blood oxygen level dependence (BOLD) activation in relation to separating event types. The six head-movement parameters derived from the realignment stage were included as covariates of no interest. All types of trials (2 decision* 2 results) and reward history (cumulative won-lost amounts before the present trial) were included as conditions in the model to account for potential influences on the results. GLM was independently applied to each voxel to identify voxels that were significantly activated for the different events of each condition.
10
ACCEPTED MANUSCRIPT
Second-level group analysis Second-level analysis treated inter-subject variability as a random effect. First, we
RI PT
determined voxels showing a main effect in different conditions. Second, we tested for voxels that showed higher or lower activity between IGD and HC groups. We first identified clusters of contiguously significant voxels at an uncorrected threshold
SC
p
S0022-3956(15)30011-X
DOI:
10.1016/j.jpsychires.2015.11.011
Reference:
PIAT 2767
To appear in:
Journal of Psychiatric Research
Received Date: 21 August 2015 Revised Date:
14 October 2015
Accepted Date: 19 November 2015
Please cite this article as: Dong G, Potenza MN, Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of negative consequences, Journal of Psychiatric Research (2015), doi: 10.1016/j.jpsychires.2015.11.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Risk-taking and risky decision-making in Internet gaming disorder: Implications regarding online gaming in the setting of
2
Department of Psychology, Zhejiang Normal University, Jinhua, Zhejiang Province, P.R. China Department of Psychiatry, Department of Neurobiology, Child Study Center, and CASAColumbia, Yale
M AN U
1
SC
Guangheng Dong 1 *, Marc N. Potenza 2
RI PT
negative consequences
University School of Medicine, New Haven, CT, USA
Corresponding author:
TE D
*
Guangheng Dong, Ph.D. Professor
Department of Psychology, Zhejiang Normal University, 688 Yingbin Road, Jinhua,
EP
Zhejiang Province 321004, PR China. Tel.: +86 15867949909.
AC C
E-mail address: [email protected]
1
ACCEPTED MANUSCRIPT
Abstract Individuals with Internet gaming disorder (IGD) continue gaming despite adverse consequences. However, the precise mechanism underlying this behavior remains
RI PT
unknown. In this study, data from 20 IGD subjects and 16 otherwise comparable healthy control subjects (HCs) were recorded and compared when they were
undergoing risk-taking and risky decision-making during functional magnetic
SC
resonance imaging (fMRI). During risk-taking and as compared to HCs, IGD subjects selected more risk-disadvantageous trials and demonstrated less activation of the
M AN U
anterior cingulate, posterior cingulate and middle temporal gyrus. During risky decision-making and as compared to HCs, IGD subjects showed shorter response times and less activations of the inferior frontal and superior temporal gyri. Taken together, data suggest that IGD subjects show impaired executive control in
TE D
selecting risk-disadvantageous choices, and they make risky decisions more hastily and with less recruitment of regions implicated in impulse control. These results suggest a possible neurobiological underpinning for why IGD subjects may exhibit
EP
poor control over their game-seeking behaviors even when encountering negative
AC C
consequences and provide possible therapeutic targets for interventions in this population.
Keywords: Internet gaming disorder; risk-taking; risky decision-making
2
ACCEPTED MANUSCRIPT
Introduction Internet gaming disorder (IGD) is a public health concern and a disorder that is
RI PT
included in Section 3 of the DSM-5 as a condition warranting further study
(American Psychiatric Association, 2013, Griffiths et al. , 2014). IGD shares features with substance (Dong et al. , 2015a, Dong et al. , 2014, Dong et al. , 2013d, Hare et al.
SC
, 2014, Lin et al. , 2015b) and gambling (Potenza, 2014a, b) addictions. Unlike drug addiction or substance abuse, no chemical or substance intake is involved in IGD,
M AN U
although excessive Internet gaming may lead to physical dependence, similar to other addictions (Dong et al. , 2013a, Holden, 2001). These findings suggest that online experiences may change brain function and related cognitive processes in manners that may perpetuate Internet gaming (Dong et al. , 2011, Holden, 2001,
TE D
Weinstein and Lejoyeux, 2010).
Risk-taking and risky decision-making contribute importantly to the development
EP
of addictions (Balogh et al. , 2013). Disadvantageous risk-taking and improper
AC C
decision-making may lead to problematic behaviors such as substance abuse (Rutherford et al. , 2010), pathological gambling (Potenza, 2014b), and IGD (Dong et al. , 2015b). Reduced cognitive capacity or willingness to avoid excessive behavioral engagement in pleasurable activities may contribute to the development of various clinical problems, including behavioral and substance addictions (Potenza et al. , 2013). Individuals with IGD may not fully consider outcomes when making decisions (Bechara et al. , 2002, Dong et al. , 2013b, Floros and Siomos,
3
ACCEPTED MANUSCRIPT
2012, Pawlikowski and Brand, 2011). Individuals with IGD may display a ‘myopia for the future’ in which they tend to pursue immediately rewarding experiences (e.g., playing online) and neglect long-term adverse consequences, as has been
RI PT
found in drug addictions (Bechara, Dolan, 2002, Floros and Siomos, 2012, Pawlikowski and Brand, 2011). Although studies have demonstrated
disadvantageous decision-making in association with IGD (Dong, Lin, 2015a, Lin et
SC
al. , 2015a), unanswered questions exist regarding the precise mechanisms that may lead individuals to game excessively or compulsively. This study aimed to
M AN U
examine the neural features underlying decision-making in IGD during performance of a risk-taking and risky decision-making task. Specific brain regions, including the inferior frontal gyrus (IFG) and orbitofrontal cortex (OFC), are involved in decision-making (Cazzell et al. , 2012, Rushworth et al.
TE D
, 2012, Sheth et al. , 2012). IFG activation may signal subjective risk and contribute to the formation of subjective feelings during decision-making (Craig, 2009). The OFC contributes importantly to value-based decision-making (Glascher et al. , 2012,
EP
Kable and Glimcher, 2009). In addiction, these prefrontal cortex regions and
AC C
associated brain structures have been implicated in the development and maintenance of problematic patterns of Internet use (Brand et al. , 2014, Dong and Potenza, 2014). A meta-analysis suggests that dysfunction of or anatomic deficits in the frontal cortex contribute to impaired impulse control (Meng et al. , 2015). Thus, in current study, we hypothesized that IGD subjects would show impaired decisionmaking that would relate to activations in these brain regions. Performance of risk-taking and decision-making tasks involves evaluation of risky
4
ACCEPTED MANUSCRIPT
features and avoidance of disadvantageous choices (Rothman and Salovey, 1997). In pathological gambling and substance-use disorders, reduced activation of reward-related circuitry is observed during gambling-related decision-making (de
RI PT
Ruiter et al., 2009; Reuter et al., 2005, Tanabe et al. , 2007). Thus, in the current
study, we hypothesized that IGD relative to healthy control subjects (HCs) would
of reward-related fronto-striatal brain regions.
SC
show disadvantageous decision-making that would relate to diminished activation
M AN U
During risk-taking or risky decision-making, executive inhibition contributes to better impulse control and advantageous decision-making. Some brain regions, such as the anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (DLPFC), contribute importantly in this regard. The ACC has been associated with error
TE D
monitoring, conflict detection and performance monitoring in decision-making (Holroyd and Coles, 2002, Platt and Huettel, 2008) and is involved in anticipating risks, especially potential losses (Krawitz et al. , 2010). The posterior cingulate
EP
cortex (PCC), and DLPFC were found to be more active during choice of risky versus
AC C
safe options (Paulus et al. , 2003, Schonberg et al. , 2011). Thus, we hypothesized that IGD versus HC subjects would show disadvantageous decision-making that would relate to activations in executive-function-related cortical brain regions.
Methods and Materials Participant Selections
5
ACCEPTED MANUSCRIPT
The experiment conforms to The Code of Ethics of the World Medical Association (Declaration of Helsinki). The Human Investigations Committee of Zhejiang Normal University approved this research. The methods were conducted in accordance with
RI PT
the approved guidelines. Participants were university students and were recruited through advertisements. Participants were right-handed males (20 IGD subjects, 16 HCs). IGD and HC groups did not significantly differ in age (IGD mean=21.33,
SC
SD=2.18 years; HC mean=21.90, SD=2.33 years; t= 0.66, p =0.48). All subjects had
normal or corrected to normal vision. Only males were included due to higher IGD
M AN U
prevalence in men than in women. All participants provided written informed consent and completed a structured psychiatric interviews (MINI) (Lecrubier et al. , 1997) that was performed by an experienced psychiatrist lasted approximately 15 minutes. All participants were free of Axis I psychiatric disorders assessed in the
TE D
MINI. We further assessed ‘depression’ with the Beck Depression Inventory (Beck et al. , 1961) and only participants scoring less than 5 were included. All participants were instructed not to use any substances of abuse, including caffeinated drinks, on
EP
the day of scanning. No participants reported previous use of illicit drugs (e.g.,
AC C
cocaine, marijuana).
Internet gaming disorder was determined based on scores of 50 or more on Young’s online Internet addiction test (IAT) (Young, 2009) and, at the same time, meeting a proposed IGD diagnosis per DSM-5 criteria (Petry et al. , 2014). Young's IAT consists of 20 items assessing problematic Internet use, including psychological dependence, compulsive use, withdrawal, problems in school or work, sleep, family or time
6
ACCEPTED MANUSCRIPT
management (Young, 2009). The IAT is a valid and reliable instrument that can be used in classifying Internet addiction (Widyanto et al. , 2011, Widyanto and McMurran, 2004). For each item, a graded response is selected from 1 = “Rarely” to
RI PT
5 = “Always”, or “Does not Apply”. Scores over 50 indicate occasional or frequent
internet-related problems (www.netaddiction.com). When recruiting IGD subjects, we added an additional criterion on Young’s established measures of IAT, ‘you
SC
spend ___% of your online time playing online games’ (>80%).
M AN U
Task and Procedure
The fMRI task used an event-related design. This task consisted of 80 trials. Each trial was divided into three stages: Decision stage (risk-taking), gamble stage (risky
TE D
decision-making), and feedback stage. Figure 1a shows the event sequence of each trial during the task. First, a white cross was presented at the center of a black screen for 500ms to cue the beginning of a new trial. During the subsequent risk-
EP
taking stage, participants were asked to select one from of two risky options (see
AC C
details on decision stage in Figure 1b). This selection process lasted for 4000 ms at most and disappeared once the participant made a selection. After a variable period of delay (mean 3000ms, ranging from 1000 to 5000ms), the risky decision-making stage followed (Figure 1c). During this stage, participants would see 4 backs of cards and were asked to guess which one was red and indicate their response by a button press within 2000 ms (the order of the cards during the gamble stage was randomized). If they missed, they would lose 15 Chinese Yuan (about $2.5 USD).
7
ACCEPTED MANUSCRIPT
After the response and a delay ranging from 1000-3000 ms (mean=2000 ms), the selected card would turn over and show participants the outcome, which was presented for a period of 1000 ms. Participants won/lost the amount according to
RI PT
the card color and the number on the card. The next trial begun after a jittered delay (mean 3000 ms, ranged from 2000-4500 ms). Subjects were asked to make
software (Psychology Software Tools, Inc.).
M AN U
Risk–taking
SC
responses with their right hands. The experiment was presented using E-prime
During the risk-taking stage, two lines of cards (each line consisting of 4 cards) were presented on the computer screen (see Figure 1b), with the red cards and the amount on it suggesting winning some amount, and the yellow ones suggesting
TE D
losing some amount. The cards were shown in colors to indicate the results (red, win; yellow, lose), win/lose rates (the proportions of different color cards), and win/lose amount (the number on cards).
EP
The probability and magnitude of the gains/losses were manipulated to create
AC C
advantageous/disadvantageous risky selections. The advantageous risk-taking means the sum of numbers on red cards (win) are larger than those on yellow cards (loss) (risk-advantageous). It suggests that although participants have opportunities to lose a big amount, they are more likely to win money in the long run. In Figure 1b, the first line is risk-advantageous selection ((45*0.75)+(-35*0.25) = 25 Yuan). In contrast, the second line in Figure 1b is risk-disadvantageous: the numbers on red card (win) are smaller than the sum in all yellow cards
8
ACCEPTED MANUSCRIPT
((35*0.25)+(-45*0.75) = -25 Yuan). It suggests that although participants have opportunities to win a big amount, they are more likely to lose money in the long
Insert Figure 1 about here
RI PT
run. Participants practiced using the same task before fMRI.
SC
Participants were told they had 80 opportunities to win some money and trials were presented in a random order. Each participant was provided with 200 Chinese Yuan
M AN U
as the initial balance before the task and was explicitly informed that he would receive the entire balance in cash at the end of the task.
Participants who chose the same selections for more than 80 percent of all trials
TE D
(who might have selective bias) or chose the selections using the same button 10 or more times consecutively (who might lack motivation to perform properly) were excluded from further analysis. Participants who took less than 10 trials in one of
EP
these conditions were excluded from further analysis to keep the statistical power.
AC C
No subjects were excluded in the risk-taking stage. However, six subjects (4 IGD, 2 HC) were excluded in the gamble stage based on these criteria.
Image acquisition and pre-processing Scanning was performed in the Shanghai Key Laboratory of Magnetic Resonance, East China Normal University. Structural images covering the whole brain were collected, using a T1-weighted three-dimensional spoiled gradient-recalled
9
ACCEPTED MANUSCRIPT
sequence (176 slices, TR=1700 ms, echo time (TE)=3.93 ms, slice thickness=1.0 mm, skip=0 mm, flip angle=15°, inversion time= 1100 ms, field of view (FOV)=240×240 mm, in-plane resolution=256×256). Functional MRI was performed on a 3T scanner
RI PT
(Siemens Trio) with a gradient-echo EPI T2 sensitive pulse sequence in 33 slices
(interleaved sequence, 3mm thickness, TR=2000ms, TE=30ms, flip angle =90°, field of view =220 × 220 mm2, matrix =64 ×64). Stimuli were presented using Invivo
SC
synchronous system (Invivo Company, www.invivocorp.com/) through a screen in the head coil, enabling participants to view the stimuli. A total of 630 volumes was
First-level regression analysis
M AN U
acquired for each participant during the 1260 seconds of task performance.
The functional data were analyzed using SPM8 and Neuroelf (http://neuroelf.net) as
TE D
described previously (DeVito et al. , 2012, Dong, Hu, 2013b, Krishnan-Sarin et al. , 2013). Images were slice-timed, corrected, reoriented (manually), and realigned to the first volume. T1-co-registered volumes were normalized to a MNI template and
EP
spatially smoothed with a 6mm FWHM Gaussian kernel. A general linear model
AC C
(GLM) was applied to identify blood oxygen level dependence (BOLD) activation in relation to separating event types. The six head-movement parameters derived from the realignment stage were included as covariates of no interest. All types of trials (2 decision* 2 results) and reward history (cumulative won-lost amounts before the present trial) were included as conditions in the model to account for potential influences on the results. GLM was independently applied to each voxel to identify voxels that were significantly activated for the different events of each condition.
10
ACCEPTED MANUSCRIPT
Second-level group analysis Second-level analysis treated inter-subject variability as a random effect. First, we
RI PT
determined voxels showing a main effect in different conditions. Second, we tested for voxels that showed higher or lower activity between IGD and HC groups. We first identified clusters of contiguously significant voxels at an uncorrected threshold
SC
p
