NeuroImage: Clinical 12 (2016) 845–851

Contents lists available at ScienceDirect

NeuroImage: Clinical

journal homepage: www.elsevier.com/locate/ynicl

Effects of outcome on the covariance between risk level and brain activity
in adolescents with internet gaming disorder
Xin Qi a,1, Yongxin Yang b,1, Shouping Dai c, Peihong Gaoc, Xin Du a, Yang Zhang a, Guijin Duc,
Xiaodong Li c,⁎⁎, Quan Zhang a,⁎
a
Department of Radiology and Tianjin Key Laboratory of Functional Imaging, Tianjin Medical University General Hospital, Tianjin 300052, China
b
Department of Psychology, Linyi Fourth People's Hospital, Linyi 276005, Shandong Province, China
c
Department of Radiology, Linyi People's Hospital, Linyi 276003, Shandong Province, China

a r t i c l e i n f o a b s t r a c t

Article history: Individuals with internet gaming disorder (IGD) often have impaired risky decision-making abilities, and IGD-re-
Received 15 May 2016 lated functional changes have been observed during neuroimaging studies of decision-making tasks. However, it
Received in revised form 30 October 2016 is still unclear how feedback (outcomes of decision-making) affects the subsequent risky decision-making in in-
Accepted 31 October 2016
dividuals with IGD. In this study, twenty-four adolescents with IGD and 24 healthy controls (HCs) were recruited
Available online 2 November 2016
and underwent functional magnetic resonance imaging while performing the balloon analog risk task (BART) to
Keywords:
evaluate the effects of prior outcomes on brain activity during subsequent risky decision-making in adolescents
Bart with IGD. The covariance between risk level and activation of the bilateral ventral medial prefrontal cortex, left
fMRI inferior frontal cortex, right ventral striatum (VS), left hippocampus/parahippocampus, right inferior occipital
Internet gaming disorder gyrus/fusiform gyrus and right inferior temporal gyrus demonstrated interaction effects of group by outcome
Risky decision-making (P b 0.05, AlphaSim correction). The regions with interactive effects were defined as ROI, and ROI-based inter-
group comparisons showed that the covariance between risk level and brain activation was significantly greater
in adolescents with IGD compared with HCs after a negative outcome occurred (P b 0.05). Our results indicated
that negative outcomes affected the covariance between risk level and activation of the brain regions related to
value estimation (prefrontal cortex), anticipation of rewards (VS), and emotional-related learning (hippocam-
pus/parahippocampus), which may be one of the underlying neural mechanisms of disadvantageous risky deci-
sion-making in adolescents with IGD.
© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction decision-making behavior, partly because of failure to utilize feedback

(Pawlikowski and Brand, 2011; Yao et al., 2014). IGD adolescents con-
Internet gaming disorder (IGD) is defined as persistent and recur- tinue to use internet games despite negative long-term consequences
rent use of the internet to engage in games, which may lead to signifi- in the social or work domains of life (Pawlikowski and Brand, 2011);
cant psychological distress and interfere with daily social life (Young, which may lead to physical dependence (Brand, 2008; Ersche et al.,
1999). IGD is the most prevalent form of internet addiction disorder in 2005), and eventually cause psychological, social, and/or work prob-
Asia (Dong et al., 2012b; Tang et al., 2014; Wu et al., 2013). Reduced lems similar to other addictions (Young, 1999). Therefore, it is impor-
risky decision-making ability is one of the most significant behavioral tant to investigate the neural mechanisms underlying feedback
impairments in IGD individuals (Pawlikowski and Brand, 2011; Yao et processing impairments during decision-making in individuals with
al., 2015). Risky decision-making is essential for human survival be- IGD.
cause risk is ubiquitous in the natural world and human life (Hastie, Optimal decision-making often requires the ability to learn from the
2001). However, IGD individuals tend to exhibit disadvantageous risky outcomes of previous choices, rewards or punishments and adjust fu-
ture choices accordingly (O'Doherty et al., 2003). The neural circuits re-
lated to the ability to interpret this feedback in health adults include the
⁎ Correspondence to: Department of Radiology and Tianjin Key Laboratory of Functional dorsal and ventral prefrontal cortex (PFC), striatum, anterior cingulate
Imaging, Tianjin Medical University General Hospital, No. 154, Anshan Road, Heping cortex (ACC), nucleus accumbens and insula (Balasubramani et al.,
District, Tianjin 300052, China.
⁎⁎ Corresponding author.
2015; Ernst and Steinhauser, 2015; Hauser et al., 2015; Kohno et al.,
E-mail addresses: [email protected] (X. Li), [email protected] (Q. Zhang). 2015; Liu et al., 2011; Rao et al., 2014; Tanabe et al., 2013). An impaired
1
These authors contributed equally to this work. ability to learn from outcomes has been demonstrated in individuals

http://dx.doi.org/10.1016/j.nicl.2016.10.024
2213-1582/© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
846 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851

with substance dependence disorders (Balconi et al., 2014; Tanabe et al., was used to exclude adolescents with diagnosis of a DSM-IV Axis I dis-
2013; Worbe et al., 2014). Studies using the Iowa gambling task (IGT) order. HCs were defined as adolescents not fitting the criteria for an
have identified decreased task performance in substance-dependent in- YDQ diagnosis, spending b 2 h/day on the internet, and having an IAT
dividuals, which may indicate deficits in learning from feedback score was b 50. Other selection criteria for the HCs were the same as
(Balconi et al., 2014; Tanabe et al., 2013). While performing the IGT, ac- those for the adolescents with IDG. The standard Raven's Progressive
tivation of the ventral striatum (VS) and medial orbitofrontal cortex Matrices (SPM) was used to test the intelligence quotient (IQ) of all par-
(OFC) was found to be deceased in substance-dependent individuals ticipants. The Barratt Impulsivity Scale (BIS) was used to test the impul-
compared to healthy subjects, which may be the underlying neuronal sivity of all participants. The Self-Rating Anxiety Scale (SAS) and Self-
mechanism of these deficits in learning from feedback (Tanabe et al., Rating Depression Scale (SDS) were used to test the levels of anxiety
2013). A study by Worbe et al. (2014) using an anticipatory risk-taking and depression among participants, respectively. Detailed clinical infor-
task found that binge drinkers showed a higher number of risky choices mation for the two groups is listed in Table 1.
in high-risk losses, and the high-risk attitude in the loss condition was The protocol of this study was approved by the Ethical Committee of
associated with greater activity in the dorsal lateral prefrontal cortex Tianjin Medical University General Hospital, and written informed con-
(DLPFC), the lateral OFC, and the superior parietal cortices. Failure to sent was obtained from each subject according to institutional
utilize feedback has also been identified in individuals with IGD by guidelines.
both behavioral and neuroimaging studies (Dong et al., 2013;
Pawlikowski and Brand, 2011; Yao et al., 2014). For instance, a behavior- 2.2. Task and procedure
al study by Yao et al. (2014) found that IGD individuals made more dis-
advantageous choices during the Game of Dice Task compared with The fMRI-adapted version of the BART used in this study was guided
healthy controls (HCs) and that individuals with IGD could not utilize by prior imaging work (Rao et al., 2008). The details of the experimental
feedback to optimize their decision-making and improve their perfor- task have previously been described (Qi et al., 2015). Briefly, the partic-
mance. A fMRI study using a continuous wins-and-losses task found ipants were presented a virtual balloon and asked to press one of two
that subjects with IGD exhibited enhanced sensitivity to win and de- buttons to either inflate (pump) the balloon or cash out. Before the ex-
creased sensitivity to loss (Dong et al., 2013). Although preliminary periment, participants were informed that they would receive the
studies have demonstrated that outcomes may affect brain activation equivalent amount of money earned during the experiment. As the bal-
during decision-making in individuals with IGD, no previous studies loon was inflated, both the monetary reward and the probability of ex-
have focused on the effect of different outcomes on the covariance be- plosion (risk level) increased. The value of the wagers corresponded to
tween risk level and brain activation during the risky decision-making the various balloon sizes, and the cumulative earnings for the tasks were
processing in individuals with IGD. displayed underneath the balloon stimuli. In a trial, participants could
In this study, twenty-four IGD adolescents and 24 HCs were enrolled, stop inflating the balloon at any point to win the wager or keep inflating
and fMRI data were obtained from the participants while performing until the balloon exploded (loss). The maximum number of pumps that
the Balloon Analog Risk Task (BART) (Lejuez et al., 2002) to investigate participants could use for each balloon was 12. The time of inflation was
the manner in which different outcomes affected the covariance be- controlled by a cue (the color of a small circle changing from red to
tween risk level and brain activation during decision-making processes green). After the participants successfully pressed a button, the small
in adolescents with IGD. The risk level in the BART was represented by circle immediately turned red at a random interval of between 1.5 and
the probability of balloon explosion, and task performance depended 2.5 s and then turned green again to indicate the next inflation. There
on to what extent participants learned from different previous out- was a varying 2–4 s interval between trials. Text indicating whether
comes (Kohno et al., 2015; Rao et al., 2014). Thus, the BART may then the trial was a “win” or “loss” was presented for 1.5 s. The picture of
be adapted to evaluate the effect of previous outcomes on the covari- the exploded balloon was presented for 20 ms. The number of balloons
ance between brain activation and risk level during subsequent risky depended on response speed rather than being pre-determined in the
decision-making processes. Based on the previous studies that the sub- experiment. After the experiment, the participants received the equiva-
jects with IGD failed to utilize feedback(Yao et al., 2014) and exhibited lent amount of money earned during the experiment.
different sensitivity to different outcomes (Dong et al., 2013), we hy-
pothesized that different outcomes would cause different effects on 2.3. Data acquisition
the covariance between risk level and brain activation in feedback-relat-
ed brain regions, mainly including the PFC and striatum in adolescents The functional MRI was conducting using a Siemens 3.0 T scanner
with IGD. This study may bring new insights into the understanding of (Magnetom Verio, Siemens, Erlangen, Germany). A gradient-recalled
the underlying neuronal mechanisms of impaired risky decision-mak- echo-planar imaging sequence was used with the following parame-
ing ability in adolescents with IGD. ters: repetition time (TR) = 2000 ms; echo time (TE) = 30 ms; field
of view = 220 × 220 mm; matrix = 64 × 64; slice thickness = 4 mm;
2. Material and method and slice gap = 1 mm. Through a mirror mounted on the head-coil,
the participants viewed the stimuli that was projected onto a screen
2.1. Participant selection
Table 1
In this study, twenty-four male adolescents with IGD and 24 age- Demographic and clinical characteristics of subjects (Mean ± SD).

and education-matched HCs were recruited. Only male adolescents IGD (N = 24) HCs (N = 24) t P
were enrolled because the prevalence of IGD is substantially higher in Age (year) 17.17 ± 3.51 17.42 ± 3.05 −2.263 0.793
men than in women. All participants were recruited based on the find- Education (year) 10.08 ± 2.98 11.25 ± 2.88 −1.380 0.174
ings of diagnostic interviews by a senior psychiatrist. Adolescents with IQ (SPM) 48.92 ± 6.79 48.58 ± 6.26 0.177 0.860
IGD were defined according to the following criteria: five or more IAT score 70.71 ± 10.76 33.42 ± 7.75 13.852 b0.001
BIS 68.79 ± 11.83 54.13 ± 8.05 5.022 b0.001
“yes” responses to the eight questions on the Young Diagnostic Ques-
SAS 43.13 ± 8.90 35.42 ± 5.02 3.680 0.001
tionnaire for Internet Addiction (YDQ) (Young, 1998), a score ≥ 50 on SDS 49.57 ± 5.02 39.38 ± 9.16 3.442 0.001
Young's Online Internet Addiction Test (IAT), spending an average of
Two-sample two-tailed t-tests. Significant level is set as P b 0.05.
four or more hours per day playing internet games, right-handed, no al- BIS, Barratt impulsivity scale; HCs, healthy controls; IAT, internet addiction test; IGD, inter-
cohol or drug abuse, no neurologic or psychiatric diseases, and medica- net game disorder; IQ, intelligence quotient; SAS, Self-Rating Anxiety Scale; SDS, Self-rat-
tion-free. The MINI-International Neuropsychiatric Interview (MINI) ing depression scale; SPM, standard Raven's progressive matrices.
 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851 847

in front of the magnet bore. The participants were asked to respond to 5 mm; and the mask of global gray matter). The brain regions with in-
the task by pressing a button on the fMRI-compatible response box. A teraction effects were defined as regions of interest (ROI). The average
10-minute BART fMRI was performed on all participants after they β estimates within ROIs were extracted and a ROI-based intergroup
learned and practiced the tasks. comparison was conducted using t-tests. Pearson's correlation analysis
was performed to test the correlations between the average β estimates
2.4. Behavioral analysis within ROIs and behavioral performance and IAT scores in adolescents
with IGD. The significance level was set at P b 0.05.
In the fMRI experiment, the behavioral variables collected during the
BART included the trial number, total and mean number of pumps, 3. Results
number of wins and losses, adjusted number of pumps (defined as the
average number of pumps excluding the balloons that exploded), ad- 3.1. Behavioral results
justed number of pumps after wins or losses, the reward collection
rate (the number of win trials divided by the number of total trials), Table 2 shows the behavioral results obtained during the fMRI ex-
and average response time (RT). A two-sample t-test was used to com- periment. The two-sample t-test revealed that the average RT was sig-
pare behavioral data between the two groups. Statistical analyses were nificantly shorter (P = 0.033) and the number of the total pumps was
conducted with SPSS 21.0, and the significance level was set at P b 0.05. significantly greater in adolescents with IGD than in HCs (P = 0.003).
There was no significant intergroup difference in the adjusted number
2.5. Functional MRI data preprocessing of pumps; mean numbers of pumps after wins, losses, or overall; trial
number; number of wins/losses, or reward collection rate (P N 0.05).
Statistical Parametric Mapping software (SPM8; http://www.fil.ion.
ucl.ac.uk/spm/software/spm8) was used to preprocess functional MRI 3.2. Imaging results
data. First, the functional images of each subject were corrected for
the acquisition time delay between different slices. Second, geometric A repeated ANOVA revealed a significant interaction effect of group
displacements due to estimated head movement were corrected by (IGD and HCs) by post-outcome (“after win” and “after loss”) on the co-
realigning all images to the first volume. Participants who had a maxi- variance between risk level and activation of the bilateral ventral medial
mum displacement in any direction (x, y, or z) N2 mm or N2° of angular prefrontal cortex (VMPFC), left IFC, right VS, left hippocampus/
rotation (x, y, or z) were excluded from the study. Third, all realigned parahippocampus, right inferior occipital gyrus/fusiform and right infe-
images were spatially normalized to the Montreal Neurological Institute rior temporal gyrus (P b 0.05, AlphaSim correction) (Table 3) (Fig. 1).
(MNI) EPI template and resampled to 3 × 3 × 3 mm3. Finally, normal- ROI-based intergroup comparisons revealed that the reason for the in-
ized images were smoothed with a Gaussian kernel of 6 mm full- teraction effects was that the covariance between risk level and brain
width at half-maximum (FWHM). activation in these regions had no significant intergroup differences
after wins, whereas higher covariance was found in the adolescents
2.6. Statistical analysis with IGD after losses (Fig. 2). The main effects of group and outcome
were insignificant after correction for multiple comparisons (P N 0.05,
In the general linear model (GLM), preprocessed fMRI data were AlphaSim correction).
modeled using a standard hemodynamic response function (HRF) There was no significant correlation between mean β estimates
with a time derivative. The head movement parameters of each partic- within ROIs and behavioral performance or IAT scores in adolescents
ipant were introduced as covariates of no-interest. with IGD (P N 0.05).
In our BART experiment, three events resulting from a button press Brain activations that co-varied with the risk levels in HCs are illus-
could be modeled: an inflation of the balloon, a win outcome, or a loss trated in Fig. 3. A threshold of whole brain false discovery rate (FDR)
outcome. Because we aimed to investigate the effect of prior outcomes corrected P b 0.005 and cluster size larger than 50 voxels was used to
on the brain activity during subsequent decision-making, four regres- identify activation areas.
sors were included in the GLM: win outcome, inflation after win, loss
outcome, and inflation after loss. The risk level associated with each in- 4. Discussion
flation (i.e., the probability of explosion, orthogonalized by a mean cen-
tral correction) was also entered into the model as a linear parametric In the present study, the effects of outcome on the covariance be-
modulation of the balloon inflation regressor. For each subject, the β es- tween the brain activation and risk level during subsequent risky deci-
timates for the parametric modulation of the risk of balloon inflation sion-making in adolescents with IGD were evaluated using BART-
after wins and after losses represented the covariance between the fMRI. Our findings in the HCs demonstrated that the BART was a reliable
risk and brain activation. method for evaluating changes in brain activation during risky decision-
Second level random effect analyses were conducted using a 2 making. There was a significant interaction effect of group by post-out-
(group: IGD and HCs) × 2 (post-outcome: “after win” and “after loss”) come (“after win” and “after loss”) on the subsequent covariance
ANOVA on the covariance between risk and brain activation with full
factorial analysis, embedded in SPM8. The covariance between risk Table 2
and brain activation “after win” and “after loss” within the same partic- The behavioral results of the BART during fMRI experiment (Mean ± SD).

ipant were processed as repeated measures. In this study, the main aim IGD (N = 24) HCs (N = 24) t P
was to evaluate the intergroup difference in the different effects of out-
Mean pumps 7.77 ± 1.01 7.32 ± 0.93 1.621 0.112
come on the covariance between risk and brain activation during deci- Adjusted pumps 6.28 ± 1.12 5.71 ± 1.10 1.783 0.081
sion-making following outcomes. Therefore, the interaction effect of Mean pumps after win 6.91 ± 1.05 6.53 ± 1.03 1.266 0.212
group by post-outcome, HCs (“after loss” - “after win”) - IGD (“after Mean pumps after loss 6.42 ± 1.21 5.82 ± 1.15 1.777 0.082
Total pumps 210.67 ± 9.77 199.21 ± 15.16 3.112 0.003
loss” - “after win”), was analyzed in this study. Additionally, the main ef-
Trial number 31.21 ± 5.00 31.58 ± 5.03 −0.259 0.797
fects of group and post-outcome were also tested. A correction for mul- Win trials 23.21 ± 5.97 25.00 ± 7.02 −0.952 0.346
tiple comparisons was performed using the Monte Carlo simulation Pop trials 8.00 ± 3.17 6.58 ± 2.81 1.636 0.109
method, resulting in a corrected threshold of P b 0.05 (AlphaSim pro- Reward collection rate 0.74 ± 0.11 0.78 ± 0.12 −1.233 0.224
gram, parameters including: single voxel P = 0.005; 1000 simulations; Response time (ms) 494.29 ± 71.84 553.21 ± 110.23 −2.194 0.033

full width at half maximum = 6 mm; cluster connection radius r = Two-sample two-tailed t-tests. Significant level is set as P b 0.05.
848 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851

Table 3
Brain regions exhibited a significant interaction effect of group by outcome on the risk-related activation.

Brain regions Cluster size (voxels) Peak T value MNI coordinates of cluster centroid (mm)

x y z

VMPFC 99 3.983 3 57 15
IFC (L) 25 3.489 −48 33 12
VS (R) 22 3.165 15 12 −3
Hippocampus/parahippocampus (L) 70 3.562 −15 −21 −15
Inferior occipital gyrus/fusiform (R) 50 3.378 30 −87 −12
Inferior temporal gyrus (R) 28 4.078 54 −33 −15

MNI: Montreal Neurological Institute; R: right side; L: left side. IFC, inferior frontal cortex; VMPFC, ventral medial prefrontal cortex; VS, ventral striatum.

between risk level and activation of the bilateral VMPFC, left IFC, right the VMPFC decreased in proportion to the value of an anticipated loss
VS, left hippocampus/parahippocampus, right inferior occipital gyrus/ (Tom et al., 2007). A previous BART study found robust activation of
fusiform and right inferior temporal gyrus. The cause of this interaction the VMPFC when participants chose to continue inflating the balloon
may have been that there was no significant intergroup difference in the (risky option) (Fukunaga et al., 2012). The IFC is responsible for the
covariance between brain activation and risk level after wins, while ad- modulation between reward and risky level (Christopoulos et al.,
olescents with IGD showed significantly higher covariance between 2009; Lin et al., 2015). A previous study by Christopoulos et al. (2009)
brain activation and risk level after losses compared with HCs. found a positive correlation between the activation of the IFC and the
It has been demonstrated that various behaviors that repeatedly re- risk aversion. A study by Dong et al. (2013) found that subjects with
inforce the reward circuitry are part of the disease of addiction (Grant et IGD showed increased activation of the inferior frontal gyrus after con-
al., 2010; Karim and Chaudhri, 2012; Love et al., 2015). When a negative tinuous wins or losses. In present study, the covariance between risk
outcome occurs, the resulting negative emotional state leads to activa- level and activation of the left IFC and bilateral VMPFC increased after
tion and dysregulation of brain reward systems, which leads to further negative feedback during decision-making, which may suggest that
negative reinforcement as the individual continues to engage in the ad- the higher the risk level was, the higher brain activation in adolescent
dictive behavior to avoid the negative affect (Koob and Volkow, 2010; with IGD. In our study, the average RT was shorter in adolescents with
Volkow et al., 2011). Therefore, a negative emotional state is an impor- IGD, and the number of total pumps was significantly higher in adoles-
tant component in the process of behavioral addiction. This may be the cents with IGD. The number of mean pumps after a loss in adolescents
reason why the covariance between risk level and brain activation in ad- with IGD was higher than in HCs. Additionally, the maximum number
olescents was affected more dramatically after losses in our study. of possible balloon pumps in our BART experiment was 12; only approx-
Previous studies have demonstrated that the IFC and VMPFC are in- imately 30 balloon trials were performed during the BOLD scanning,
volved in risk-evaluation and value calculation (Cardinal, 2006; which may have decreased the sensitivity of detecting intergroup differ-
Christopoulos et al., 2009; Rushworth et al., 2011; Schonberg et al., ences in behavioral performance (Rao et al., 2010). It is cannot be ex-
2012; Tobler et al., 2007). The VMPFC is responsible for processing neg- cluded that the intergroup difference in behavioral performance
ative performance outcomes and receiving punishment feedback would become more significant with an increase in the number of pos-
(Elliott et al., 2000; O'Doherty et al., 2003; O'Doherty et al., 2001; sible balloon pumps. Combined, these suggest that the adolescents with
Rushworth et al., 2011; van Leijenhorst et al., 2006), and activation of IGD were more impulsive than HCs, which was consistent with the

Fig. 1. Brain regions with a significant interaction effect of group by post-outcome on the covariance between the risk level and the brain activation. A, left IFC; B, VMPFC; C, right ventral
striatum; D, right inferior occipital cortex/fusiform; E, right inferior temporal gyrus; F, left hippocampus/parahippocampus. T value ranges from 0 to 5 presented by color bar.
 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851 849

Fig. 2. ROI-based post hoc analysis of the brain regions with a significant interaction effect of group by post-outcome on the covariance between the risk level and the brain activation. A, left
IFC; B, VMPFC; C, right ventral striatum; D, right inferior occipital gyrus/fusiform; E, right inferior temporal gyrus; F, left hippocampus/parahippocampus. The covariance between the risk
level and the brain activation of these regions after loss significantly increased in IGD group compared with the HCs group.

observed higher scores on the BIS in adolescents with IGD. Our results and activation of the VS increased after receiving negative feedback,
indicated that adolescents with IGD and high impulsivity might need which indicated that adolescents with IGD might experience higher an-
to recruit more PFC resources to evaluate the risk values of the subse- ticipation of rewards during subsequent risky decision-making after a
quent options after receiving negative feedback. loss. Moreover, previous studies revealed strong interconnections be-
The VS is a key structure of the brain's reward system, and has been tween the VS and several parts of the medial PFC, namely the fronto-
implicated in anticipating and processing different types of rewards, as striatal circuits, which are associated to the addictive behavior, and
well as producing learning signals known as prediction errors (Dong et the PFC exerts top-down control of the VS that has been found to be in-
al., 2013; Haber and Knutson, 2010; McClure et al., 2003; Seymour et al., volved in reward-processing (Ashby et al., 2010; Haber and Knutson,
2007; van Duijvenvoorde et al., 2014). Neuroimaging studies have re- 2010). Therefore, it is also plausible that the increased covariance be-
vealed increased reward-related activation of the VS in individuals tween risk level and activation of the PFC after receiving negative feed-
with substance dependence and behavioral addiction. For instance, back might be due to the need to control VS activation.
fMRI studies have found hyperactivity in the VS when processing re- The hippocampus is a key structure in the memory system and has
ward stimuli (Garcia-Garcia et al., 2014) and risky decision-making been implicated in encoding recent experiences (Ferbinteanu and
(Yamamoto et al., 2015) in substance-dependent individuals. In addi- Shapiro, 2003). The parahippocampus provides a contextual represen-
tion, activity of the VS increased in pathological gambling patients dur- tation function and is an important afferent pathway to the hippocam-
ing loss anticipation in a monetary incentive delay task (Romanczuk- pus (Rudy, 2009). Human decisions cannot be explained solely by
Seiferth et al., 2015). In our study, the covariance between risk level rational imperatives but are strongly influenced by emotion (Coricelli

Fig. 3. Brain activations covaried with the level of risk in HCs. T value ranges from 0 to 8 presented by color bar.
850 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851

et al., 2007). Previous studies have found that hippocampal activation Author contributions
was affected by emotionality and regret (Addis et al., 2004; Coricelli et
al., 2007). The hippocampus/parahippocampus involvement in the pro- X.Q., Y.Y., X.L. and Q.Z. designed research; X.Q., S.D., P.G., Y.Z., X.D.
cess of learning from experiences (Eichenbaum, 2004; Steidl et al., and Q.Z. performed research; Y.Y., S.D., P.G. was involved in the clinical
2006) might play a role in the recollection of a previous decision-mak- assessment; X.Q., Y.Z., G.D., and Q.Z. analyzed data; X.Q., Y.Z., X.L., Y.Y.
ing experiences, especially the emotional experiences (Ferbinteanu and Q.Z. wrote the paper.
and Shapiro, 2003; Kohno et al., 2015). Thus, after a negative feedback,
the lesson to be learnt is: ‘in the future, pay more attention to the poten- Disclosure
tial consequences of your choices’ (Coricelli et al., 2007). In our study,
the increased covariance between risk level and activation of the hippo- All authors have reported no conflicts of interest.
campus/parahippocampus in the adolescents with IGD after negative
feedback might be associated with learning following emotional experi-
ences during risky decision-making. References
In this study, adolescents with IGD also showed increased covari- Addis, D.R., Moscovitch, M., Crawley, A.P., McAndrews, M.P., 2004. Recollective qualities
ance between risk level and activation of the right inferior occipital modulate hippocampal activation during autobiographical memory retrieval. Hippo-
gyrus/fusiform and the right inferior temporal gyrus after receiving neg- campus 14:752–762. http://dx.doi.org/10.1002/hipo.10215.
Ashby, F.G., Turner, B.O., Horvitz, J.C., 2010. Cortical and basal ganglia contributions to
ative feedback during risky decision-making. The occipitotemporal area habit learning and automaticity. Trends Cogn. Sci. 14:208–215. http://dx.doi.org/10.
is related to visual perceptual learning, spatial attention and perceptual 1016/j.tics.2010.02.001.
judgments (LaBar et al., 1999; Li et al., 2012), and the inferior temporal Association AP, 2013. Diagnostic and Statistical Manual of Mental Disorders. fifth ed.
American Psychiatric Association, Washington, DC.
gyrus/fusiform are a component of the fronto-temporal circuits, which Balasubramani, P.P., Chakravarthy, V.S., Ravindran, B., Moustafa, A.A., 2015. A network
are associated with working memory and attention processing. The in- model of basal ganglia for understanding the roles of dopamine and serotonin in re-
creased covariance between risk level and activation of the ward-punishment-risk based decision making. Front. Comput. Neurosci. 9:76. http://
dx.doi.org/10.3389/fncom.2015.00076.
occipitotemporal areas in adolescents with IGD after receiving negative Balconi, M., Finocchiaro, R., Campanella, S., 2014. Reward sensitivity, decisional bias, and
feedback might indicate that these adolescents made more attempts to metacognitive deficits in cocaine drug addiction. J. Addict. Med. 8:399–406. http://dx.
remember previous feedback and focus on subsequent risky decision- doi.org/10.1097/adm.0000000000000065.
Bjork, J.M., Knutson, B., Fong, G.W., Caggiano, D.M., Bennett, S.M., Hommer, D.W., 2004.
making simultaneously; however, which should be confirmed in future
Incentive-elicited brain activation in adolescents: similarities and differences from
studies. young adults. J. Neurosci. 24:1793–1802. http://dx.doi.org/10.1523/JNEUROSCI.
There was no significant correlation between the average β esti- 4862-03.2004.
Brand, M., 2008. Does the feedback from previous trials influence current decisions? A
mates within ROIs and behavioral performance or IAT scores in adoles-
study on the role of feedback processing in making decisions under explicit risk con-
cents with IGD. One possible explanation is that altered covariance ditions. J. Neuropsychol. 2, 431–443.
between risk level and brain activation is a trait index rather than Cardinal, R.N., 2006. Neural systems implicated in delayed and probabilistic reinforce-
state index, which means that the changed index has no quantitative re- ment. Neural Netw. 19:1277–1301. http://dx.doi.org/10.1016/j.neunet.2006.03.004.
Christopoulos, G.I., Tobler, P.N., Bossaerts, P., Dolan, R.J., Schultz, W., 2009. Neural corre-
lationship with the extent of disease. It's also worth noting that the ad- lates of value, risk, and risk aversion contributing to decision making under risk.
olescent brain is still developing compared to the adult brain, so risk J. Neurosci. 29:12574–12583. http://dx.doi.org/10.1523/JNEUROSCI.2614-09.2009.
decision-making strategies and the responses to feedback may differ be- Coricelli, G., Dolan, R.J., Sirigu, A., 2007. Brain, emotion and decision making: the paradig-
matic example of regret. Trends Cogn. Sci. 11:258–265. http://dx.doi.org/10.1016/j.
tween the two (Bjork et al., 2004; Smith et al., 2015; van Leijenhorst et tics.2007.04.003.
al., 2006). Thus, the developmental heterogeneity of adolescents might Dong, G., DeVito, E., Huang, J., Du, X., 2012a. Diffusion tensor imaging reveals thalamus
be a factor that affects decision-making-related brain activity, and a lon- and posterior cingulate cortex abnormalities in internet gaming addicts. J. Psychiatr.
Res. 46:1212–1216. http://dx.doi.org/10.1016/j.jpsychires.2012.05.015.
gitudinal study would be warranted. Dong, G., Devito, E.E., Du, X., Cui, Z., 2012b. Impaired inhibitory control in ‘internet addic-
Several limitations need be noted in the present study. First, a rela- tion disorder’: a functional magnetic resonance imaging study. Psychiatry Res. 203:
tive small sample size may have reduced the statistical power. Second, 153–158. http://dx.doi.org/10.1016/j.pscychresns.2012.02.001.
Dong, G., Hu, Y., Lin, X., Lu, Q., 2013. What makes Internet addicts continue playing online
our results are limited to male adolescents because of the substantially
even when faced by severe negative consequences? Possible explanations from an
higher IGD prevalence in young males; conditions in female adolescents fMRI study. Biol. Psychol. 94:282–289. http://dx.doi.org/10.1016/j.biopsycho.2013.
and other age groups need to be evaluated in future studies. Third, the 07.009.
Eichenbaum, H., 2004. Hippocampus: cognitive processes and neural representations that
length and amount of experience across video game genres were not
underlie declarative memory. Neuron 44:109–120. http://dx.doi.org/10.1016/j.
controlled. Finally, although the widely used YDQ and YIAT were neuron.2004.08.028.
adopted as diagnostic criteria, other standards have been adopted by Elliott, R., Friston, K.J., Dolan, R.J., 2000. Dissociable neural responses in human reward
other researchers (Dong et al., 2012a; Ko et al., 2014; Yao et al., 2014), systems. J. Neurosci. 20, 6159–6165.
Ernst, B., Steinhauser, M., 2015. Effects of invalid feedback on learning and feedback-relat-
which may limit the comparability among the studies to some extent. ed brain activity in decision-making. Brain Cogn. 99:78–86. http://dx.doi.org/10.
With the IGD now listed in the DSM-V as a disorder requiring further re- 1016/j.bandc.2015.07.006.
search (Association AP, 2013), this problem may be corrected in the Ersche, K.D., Roiser, J.P., Clark, L., London, M., Robbins, T.W., Sahakian, B.J., 2005. Punish-
ment induces risky decision-making in methadone-maintained opiate users but not
near future. in heroin users or healthy volunteers. Neuropsychopharmacology 30:2115–2124.
http://dx.doi.org/10.1038/sj.npp.1300812.
Ferbinteanu, J., Shapiro, M.L., 2003. Prospective and retrospective memory coding in the
hippocampus. Neuron 40:1227–1239. http://dx.doi.org/10.1016/S0896-
5. Conclusions 6273(03)00752-9.
Fukunaga, R., Brown, J.W., Bogg, T., 2012. Decision making in the Balloon Analogue Risk
To the best of our knowledge, this is the first study to investigate Task (BART): anterior cingulate cortex signals loss aversion but not the infrequency
of risky choices. Cogn. Affect. Behav. Neurosci. 12:479–490. http://dx.doi.org/10.
the effect of previous outcomes on the covariance between brain ac-
3758/s13415-012-0102-1.
tivation and risk level during risky decision-making processing ado- Garcia-Garcia, I., Horstmann, A., Jurado, M.A., Garolera, M., Chaudhry, S.J., Margulies, D.S.,
lescents with IGD using the BART-fMRI. Our results indicated that Villringer, A., Neumann, J., 2014. Reward processing in obesity, substance addiction
and non-substance addiction. Obes. Rev. 15:853–869. http://dx.doi.org/10.1111/obr.
negative outcomes affected the covariance between risk level and
12221.
activation of the brain regions related to value estimation (PFC), an- Grant, J.E., Potenza, M.N., Weinstein, A., Gorelick, D.A., 2010. Introduction to behavioral
ticipation of rewards (VS), and emotion-related learning (hippocam- addictions. Am. J. Drug Alcohol Abuse 36:233–241. http://dx.doi.org/10.3109/
pus/parahippocampus), which may be one of the underlying neural 00952990.2010.491884.
Haber, S.N., Knutson, B., 2010. The reward circuit: linking primate anatomy and human
mechanisms of disadvantageous risky decision-making in adoles- imaging. Neuropsychopharmacology 35:4–26. http://dx.doi.org/10.1038/npp.2009.
cents with IGD. 129.
 X. Qi et al. / NeuroImage: Clinical 12 (2016) 845–851 851

Hastie, R., 2001. Problems of judgment and decision making. Annu. Rev. Psychol. 52: Romanczuk-Seiferth, N., Koehler, S., Dreesen, C., Wüstenberg, T., Heinz, A., 2015. Patho-
653–683. http://dx.doi.org/10.1146/annurev.psych.52.1.653. logical gambling and alcohol dependence: neural disturbances in reward and loss
Hauser, T.U., Hunt, L.T., Iannaccone, R., Walitza, S., Brandeis, D., Brem, S., Dolan, R.J., 2015. avoidance processing. Addict. Biol. 20:557–569. http://dx.doi.org/10.1111/adb.12144.
Temporally dissociable contributions of human medial prefrontal Subregions to re- Rudy, J.W., 2009. Context representations, context functions, and the parahippocampal-
ward-guided learning. J. Neurosci. 35:11209–11220. http://dx.doi.org/10.1523/ hippocampal system. Learn. Mem. 16:573–585. http://dx.doi.org/10.1101/lm.
JNEUROSCI.0560-15.2015. 1494409.
Karim, R., Chaudhri, P., 2012. Behavioral addictions: an overview. J. Psychoactive Drugs Rushworth, M.F., Noonan, M.P., Boorman, E.D., Walton, M.E., Behrens, T.E., 2011. Frontal
44:5–17. http://dx.doi.org/10.1080/02791072.2012.662859. cortex and reward-guided learning and decision-making. Neuron 70:1054–1069.
Ko, C.H., Hsieh, T.J., Chen, C.Y., Yen, C.F., Chen, C.S., Yen, J.Y., Wang, P.W., Liu, G.C., 2014. http://dx.doi.org/10.1016/j.neuron.2011.05.014.
Altered brain activation during response inhibition and error processing in subjects Schonberg, T., Fox, C.R., Mumford, J.A., Congdon, E., Trepel, C., Poldrack, R.A., 2012. De-
with Internet gaming disorder: a functional magnetic imaging study. Eur. Arch. Psy- creasing ventromedial prefrontal cortex activity during sequential risk-taking: an
chiatry Clin. Neurosci. 264:661–672. http://dx.doi.org/10.1007/s00406-013-0483-3. fMRI investigation of the balloon analog risk task. Front. Neurosci. 6:80. http://dx.
Kohno, M., Ghahremani, D.G., Morales, A.M., Robertson, C.L., Ishibashi, K., Morgan, A.T., doi.org/10.3389/fnins.2012.00080.
Mandelkern, M.A., London, E.D., 2015. Risk-taking behavior: dopamine d2/d3 recep- Seymour, B., Daw, N., Dayan, P., Singer, T., Dolan, R., 2007. Differential encoding of losses
tors, feedback, and frontolimbic activity. Cereb. Cortex 25:236–245. http://dx.doi. and gains in the human striatum. J. Neurosci. 27:4826–4831. http://dx.doi.org/10.
org/10.1093/cercor/bht218. 1523/JNEUROSCI.0400-07.2007.
Koob, G.F., Volkow, N.D., 2010. Neurocircuitry of addiction. Neuropsychopharmacology Smith, A.R., Steinberg, L., Strang, N., Chein, J., 2015. Age differences in the impact of peers
35:217–238. http://dx.doi.org/10.1038/npp.2009.110. on adolescents' and adults' neural response to reward. Dev. Cogn. Neurosci. 11:
LaBar, K.S., Gitelman, D.R., Parrish, T.B., Mesulam, M., 1999. Neuroanatomic overlap of 75–82. http://dx.doi.org/10.1016/j.dcn.2014.08.010.
working memory and spatial attention networks: a functional MRI comparison with- Steidl, S., Mohi-uddin, S., Anderson, A.K., 2006. Effects of emotional arousal on multiple
in subjects. NeuroImage 10:695–704. http://dx.doi.org/10.1006/nimg.1999.0503. memory systems: evidence from declarative and procedural learning. Learn. Mem.
Lejuez, C.W., Read, J.P., Kahler, C.W., Richards, J.B., Ramsey, S.E., Stuart, G.L., Strong, D.R., 13:650–658. http://dx.doi.org/10.1101/lm.324406.
Brown, R.A., 2002. Evaluation of a behavioral measure of risk taking: the Balloon An- Tanabe, J., Reynolds, J., Krmpotich, T., Claus, E., Thompson, L.L., Du, Y.P., Banich, M.T., 2013.
alogue Risk Task (BART). J. Exp. Psychol. Appl. 8:75–84. http://dx.doi.org/10.1037// Reduced neural tracking of prediction error in substance-dependent individuals. Am.
1076-898X.8.2.75. J. Psychiatry 170:1356–1363. http://dx.doi.org/10.1176/appi.ajp.2013.12091257.
Li, S., Mayhew, S.D., Kourtzi, Z., 2012. Learning shapes spatiotemporal brain patterns for Tang, J., Yu, Y., Du, Y., Ma, Y., Zhang, D., Wang, J., 2014. Prevalence of internet addiction
flexible categorical decisions. Cereb. Cortex 22:2322–2335. http://dx.doi.org/10. and its association with stressful life events and psychological symptoms among ad-
1093/cercor/bhr309. olescent internet users. Addict. Behav. 39:744–747. http://dx.doi.org/10.1016/j.
Lin, X., Zhou, H., Dong, G., Du, X., 2015. Impaired risk evaluation in people with Internet addbeh.2013.12.010.
gaming disorder: fMRI evidence from a probability discounting task. Prog. Neuro- Tobler, P.N., O'Doherty, J.P., Dolan, R.J., Schultz, W., 2007. Reward value coding distinct
Psychopharmacol. Biol. Psychiatry 56:142–148. http://dx.doi.org/10.1016/j.pnpbp. from risk attitude-related uncertainty coding in human reward systems.
2014.08.016. J. Neurophysiol. 97:1621–1632. http://dx.doi.org/10.1152/jn.00745.2006.
Liu, X., Hairston, J., Schrier, M., Fan, J., 2011. Common and distinct networks underlying Tom, S.M., Fox, C.R., Trepel, C., Poldrack, R.A., 2007. The neural basis of loss aversion in de-
reward valence and processing stages: a meta-analysis of functional neuroimaging cision-making under risk. Science 315:515–518. http://dx.doi.org/10.1126/science.
studies. Neurosci. Biobehav. Rev. 35:1219–1236. http://dx.doi.org/10.1016/j. 1134239.
neubiorev.2010.12.012. van Duijvenvoorde, A.C., Op de Macks, Z.A., Overgaauw, S., Gunther Moor, B., Dahl, R.E.,
Love, T., Laier, C., Brand, M., Hatch, L., Hajela, R., 2015. Neuroscience of internet pornogra- Crone, E.A., 2014. A cross-sectional and longitudinal analysis of reward-related
phy addiction: a review and update. Behav. Sci. 5:388–433. http://dx.doi.org/10. brain activation: effects of age, pubertal stage, and reward sensitivity. Brain Cogn.
3390/bs5030388. 89:3–14. http://dx.doi.org/10.1016/j.bandc.2013.10.005.
McClure, S.M., Berns, G.S., Montague, P.R., 2003. Temporal prediction errors in a passive van Leijenhorst, L., Crone, E.A., Bunge, S.A., 2006. Neural correlates of developmental dif-
learning task activate human striatum. Neuron 38:339–346. http://dx.doi.org/10. ferences in risk estimation and feedback processing. Neuropsychologia 44:
1016/S0896-6273(03)00154-5. 2158–2170. http://dx.doi.org/10.1016/j.neuropsychologia.2006.02.002.
O'Doherty, J., Kringelbach, M.L., Rolls, E.T., Hornak, J., Andrews, C., 2001. Abstract reward Volkow, N.D., Wang, G.J., Fowler, J.S., Tomasi, D., Telang, F., 2011. Addiction: beyond dopa-
and punishment representations in the human orbitofrontal cortex. Nat. Neurosci. mine reward circuitry. Proc. Natl. Acad. Sci. U. S. A. 108:15037–15042. http://dx.doi.
4:95–102. http://dx.doi.org/10.1038/82959. org/10.1073/pnas.1010654108.
O'Doherty, J., Critchley, H., Deichmann, R., Dolan, R.J., 2003. Dissociating valence of out- Worbe, Y., Irvine, M., Lange, I., Kundu, P., Howell, N.A., Harrison, N.A., Bullmore, E.T.,
come from behavioral control in human orbital and ventral prefrontal cortices. Robbins, T.W., Voon, V., 2014. Neuronal correlates of risk-seeking attitudes to antici-
J. Neurosci. 23, 7931–7939. pated losses in binge drinkers. Biol. Psychiatry 76:717–724. http://dx.doi.org/10.
Pawlikowski, M., Brand, M., 2011. Excessive Internet gaming and decision making: do ex- 1016/j.biopsych.2013.11.028.
cessive World of Warcraft players have problems in decision making under risky con- Wu, X., Chen, X., Han, J., Meng, H., Luo, J., Nydegger, L., Wu, H., 2013. Prevalence and fac-
ditions? Psychiatry Res. 188:428–433. http://dx.doi.org/10.1016/j.psychres.2011.05. tors of addictive Internet use among adolescents in Wuhan, China: interactions of pa-
017. rental relationship with age and hyperactivity-impulsivity. PLoS One 8, e61782.
Qi, X., Du, X., Yang, Y., Du, G., Gao, P., Zhang, Y., Qin, W., Li, X., Zhang, Q., 2015. Decreased http://dx.doi.org/10.1371/journal.pone.0061782.
modulation by the risk level on the brain activation during decision making in adoles- Yamamoto, D.J., Woo, C.W., Wager, T.D., Regner, M.F., Tanabe, J., 2015. Influence of dorso-
cents with internet gaming disorder. Front. Behav. Neurosci. 9:296. http://dx.doi.org/ lateral prefrontal cortex and ventral striatum on risk avoidance in addiction: a medi-
10.3389/fnbeh.2015.00296. ation analysis. Drug Alcohol Depend. 149:10–17. http://dx.doi.org/10.1016/j.
Rao, H., Korczykowski, M., Pluta, J., Hoang, A., Detre, J.A., 2008. Neural correlates of volun- drugalcdep.2014.12.026.
tary and involuntary risk taking in the human brain: an fMRI study of the Balloon An- Yao, Y.W., Chen, P.R., Chen, C., Wang, L.J., Zhang, J.T., Xue, G., Deng, L.Y., Liu, Q.X., Yip, S.W.,
alog Risk Task (BART). NeuroImage 42:902–910. http://dx.doi.org/10.1016/j. Fang, X.Y., 2014. Failure to utilize feedback causes decision-making deficits among
neuroimage.2008.05.046. excessive Internet gamers. Psychiatry Res. 219:583–588. http://dx.doi.org/10.1016/j.
Rao, H., Mamikonyan, E., Detre, J.A., Siderowf, A.D., Stern, M.B., Potenza, M.N., Weintraub, psychres.2014.06.033.
D., 2010. Decreased ventral striatal activity with impulse control disorders in Yao, Y.W., Chen, P.R., Li, S., Wang, L.J., Zhang, J.T., Yip, S.W., Chen, G., Deng, L.Y., Liu, Q.X.,
Parkinson's disease. Mov. Disord. 25:1660–1669. http://dx.doi.org/10.1002/mds. Fang, X.Y., 2015. Decision-making for risky gains and losses among college students
23147. with internet gaming disorder. PLoS One 10, e0116471. http://dx.doi.org/10.1371/
Rao, L.L., Zhou, Y., Liang, Z.Y., Rao, H., Zheng, R., Sun, Y., Tan, C., Xiao, Y., Tian, Z.Q., Chen, journal.pone.0116471.
X.P., Wang, C.H., Bai, Y.Q., Chen, S.G., Li, S., 2014. Decreasing ventromedial prefrontal Young, K., 1998. Internet addiction: the emergence of a new clinical disorder.
cortex deactivation in risky decision making after simulated microgravity: effects of CyberPsychology & Behavior 1:237–244. http://dx.doi.org/10.1089/cpb.1998.1.237.
−6 degrees head-down tilt bed rest. Front. Behav. Neurosci. 8:187. http://dx.doi. Young, K., 1999. Internet addiction: symptoms, evaluation, and treatment. Innovations in
org/10.3389/fnbeh.2014.00187. Clinical Practice: A Source Book 17, 19–31.