Brain networks alterations in cocaine use and gambling disorders during emotion regulation MARIA PICÓ-PÉREZ1,2,3,4p , VÍCTOR COSTUMERO4, JUAN VERDEJO-ROMÁN5,6, NATALIA ALBEIN-URIOS7, Journal of Behavioral JOSÉ MIGUEL MARTÍNEZ-GONZÁLEZ8, Addictions CARLES SORIANO-MAS9,10,11, 11 (2022) 2, 373–385 ALFONSO BARRÓS-LOSCERTALES4pp and DOI: ANTONIO VERDEJO-GARCIA12 10.1556/2006.2022.00018 © 2022 The Author(s) 1 Life and Health Sciences Research Institute (ICVS), University of Minho, Braga, Portugal 2 ICVS/3B's, PT Government Associate Laboratory, Braga/Guimarães, Portugal 3 Clinical Academic Center - Braga, Braga, Portugal 4 Departamento de Psicología Básica, Clínica y Psicobiología, Universitat Jaume I, Castelló de la Plana, Spain FULL-LENGTH REPORT 5 Mind, Brain and Behavior Research Center (CIMCYC), University of Granada, Granada, Spain 6 Department of Personality, Assessment and Clinical Treatment, University of Granada, Granada, Spain 7 Cognitive Neuroscience Unit, School of Psychology, Deakin University, Geelong, Victoria, Australia 8 Clinical Psychologist, Centro Provincial de Drogodependencias, Diputación de Granada, Spain 9 Mental Health Research Networking Center (CIBERSAM), Barcelona, Spain 10 Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet de Llobregat, Barcelona, Spain 11 Department of Psychobiology and Methodology of Health Sciences, Universitat Autònoma de Barcelona, Spain 12 School of Psychology, Monash Institute of Cognitive and Clinical Neurosciences, Monash University, Victoria, Australia Received: October 19, 2021 • Revised manuscript received: January 25, 2022; March 14, 2022 • Accepted: March 22, 2022 Published online: April 22, 2022 ABSTRACT Background: Cocaine use disorder (CUD) and gambling disorder (GD) share clinical features and neural alterations, including emotion regulation deficits and dysfunctional activation in related networks. However, they also exhibit differential aspects, such as the neuroadaptive effects of long-term drug consumption in CUD as compared to GD. Neuroimaging research aimed at disentangling their shared and specific alterations can contribute to improve understanding of both disorders. Methods: We compared CUD (N 5 15), GD (N 5 16) and healthy comparison (HC; N 5 17) groups using a network-based approach for studying temporally coherent functional networks during functional magnetic resonance imaging (fMRI) of an emotion regulation task. We focused our analysis in limbic, ventral frontostriatal, dorsal attentional (DAN) and executive networks (FPN), given their involvement in emotion regulation and their alteration in CUD and GD. Correlations with measures of emotional experience and impulsivity (UPPS-P) were also performed. Results: The limbic network was signifi- cantly decreased during emotional processing both for CUD and GD individuals compared to the HC p Corresponding author. group. Furthermore, GD participants compared to HC showed an increased activation in the ventral E-mail: b8467@med.uminho.pt frontostriatal network during emotion regulation. Finally, networks' activation patterns were modulated by impulsivity traits. Conclusions: Functional network analyses revealed both overlapping and unique pp effects of stimulant and gambling addictions on neural networks underpinning emotion regulation. Corresponding author. E-mail: barros@uji.es KEYWORDS addiction, cocaine, emotion regulation, fMRI, gambling, independent-component analysis 374 Journal of Behavioral Addictions 11 (2022) 2, 373–385 INTRODUCTION ventrolateral prefrontal cortex, dorsal anterior cingulate cortex, premotor and parietal cortices (Buhle et al., 2013). These regions overlap with typical resting-state cognitive Cocaine use and gambling disorders share common clinical control networks such as the dorsal attentional network features, such as unsuccessful efforts to control use despite (DAN) and the frontoparietal networks (left and right FPN) negative consequences, craving and withdrawal effects (Lee, (Buhle et al., 2013), which are also associated with addiction Hoppenbrouwers, & Franken, 2019). Moreover, elevated (Goldstein & Volkow, 2011; Zilverstand, Parvaz, & Gold- impulsivity and altered reward processing, decision making stein, 2017). Zilverstand et al. (2017) highlighted that several and emotion regulation feature both disorders (Lee et al., studies in participants with substance use disorders, 2019; Michalczuk, Bowden-Jones, Verdejo-Garcia, & Clark, including CUD, have observed reduced activation in frontal 2011; Zilverstand, Parvaz, Moeller, & Goldstein, 2016). and parietal regions during downregulation of negative Additionally, overlapping neurobiological alterations emotions. Moreover, Buhle et al. (2013) pointed to the (involving ventromedial and ventrolateral prefrontal modulatory role of the amygdala and parahippocampus in cortices) underlie psychopathological behaviors in both emotion regulation, usually included within the limbic disorders (Kober et al., 2016; Verdejo-Garcia et al., 2015). network, in non-clinical populations. However, alterations However, cocaine use disorder (CUD) critically differs from in the limbic network have not been consistently observed in gambling disorder (GD) in terms of exposure to the neu- addiction studies (Zilverstand et al., 2016). Instead, ventral roadaptive effects of long-term drug consumption (Albein- frontostriatal circuitry alterations have been consistently Urios, Martinez-González, Lozano, Clark, & Verdejo-García, reported in both CUD (Bustamante et al., 2014) and GD 2012; Torres et al., 2013). Thus, comparing neural func- (Reuter et al., 2005). Therefore, emotion regulation involves tioning in people with CUD versus GD may help disentangle several brain networks related to cognitive and emotional the neuroadaptive effects of cocaine and its contribution to processes which have shown to be affected in addiction. addiction neurobiology. That is, to what extent addiction is These neural network alterations may compromise cognitive produced by predisposition, neuroadaptive or learning ef- reappraisal in people with CUD or GD. fects during the long-term exposure to drugs that lead to The use of adaptive emotion regulation strategies such as addiction. cognitive reappraisal could reduce stress reactivity to nega- Current research shows that alterations related to CUD tive emotional experiences in individuals with substance and may be understood in terms of large-scale brain networks behavioral addictions (Goldstein & Volkow, 2011). In this dysfunction (Zilverstand, Huang, Alia-Klein, & Goldstein, line, emotional states or cues linked to drug-related affective 2018). Among network analysis techniques, independent states are important contributors to relapse (Sinha, 2009). component analysis (ICA) is becoming increasingly used to During emotional regulation tasks, addiction-related stimuli investigate how functional networks are modulated by tend to produce upregulation of brain networks and regional experimental factors (Barrós-Loscertales et al., 2020; activation, while non-drug related emotional stimuli tend to Costumero et al., 2015; Kim et al., 2009; Worhunsky, produce downregulation (Barrós-Loscertales, 2016; Zilver- Potenza, & Rogers, 2017). ICA is a data-driven approach stand et al., 2018). These effects are particularly related with that assumes that functional magnetic resonance imaging the reward, limbic, attentional and executive networks, (fMRI) data are linear mixtures of independent source sig- although others such as the salience and default mode net- nals, and it attempts to extract maximally independent sig- works have also been implicated (Zilverstand et al., 2018). nals and their mixing coefficients (Calhoun, Adali, Pearlson, Identifying the neural underpinnings of impaired emotion & Pekar, 2001). The driving principle behind ICA is that regulation and their association to addiction-related traits these independent source signals represent coherent (e.g., stress-related impulsivity) may contribute to improve groupings of MRI activations, which implies the represen- understanding of both CUD and GD. tation of a functionally connected network. When applied to Brain regions implicated in impulsivity are largely task-fMRI data, the ICA methodology allows to first identify overlapping with addiction related abnormalities in meso- the intrinsic functional connectivity networks, and then to corticolimbic regions (Mackey & Paulus, 2013). Thus, it has study how the estimated time courses associated with these been suggested that inadequate regulation of emotion fuels networks are modulated during a task. The study of tem- addictive behaviors in impulsive individuals (Torres et al., poral signals estimated by ICA reveals how specific func- 2013). Also, trait impulsivity has been related to poor tional networks relate with experimental conditions, which cognitive control of emotions and inappropriate decisions may provide new insights into the functional activity hidden (Torres et al., 2013; Verdejo-García, Alcázar-Córcoles, & from conventional voxel-wise GLM analyses (Xu, Potenza, & Albein-Urios, 2019). The UPPS-P model has been applied to Calhoun, 2013). CUD (Albein-Urios, Pilatti, Lozano, Martínez-González, & The fMRI cognitive reappraisal task can be studied using Verdejo-García, 2014; Fernández-Serrano, Perales, Moreno- this approach. During this task, participants have to rein- López, Pérez-García, & Verdejo-García, 2011) and GD terpret the meaning of affective stimuli in order to modify (Navas et al., 2017; Torres et al., 2013). In the context of GD, their emotional intensity (Ochsner, Bunge, Gross, & impulsivity seems to play a global role (Mestre-Bach et al., Gabrieli, 2002). Reappraisal-evoked activation typically in- 2020), but a specific role of negative urgency in the context volves a distributed network, including the dorsolateral and of emotion regulation has also been shown (Navas et al., Journal of Behavioral Addictions 11 (2022) 2, 373–385 375 2017). Conversely, people with CUD compared with healthy task consists of the presentation of neutral or negative picture controls (HC) score higher on several UPPS-P dimensions stimuli that participants must (1) Observe (to passively such as lack of premeditation, positive and negative urgency observe neutral pictures); (2) Maintain (to actively focus on (Fernández-Serrano et al., 2011). Therefore, different di- the emotions elicited by negative emotional pictures); or (3) mensions of impulsivity may subserve emotion regulation Regulate (to reappraise the emotions induced by the negative effects related to CUD and GD. emotional pictures by virtue of cognitive reappraisal tech- In this study, we compared CUD, GD and healthy niques previously trained). Moreover, the intensity of the comparison groups during an fMRI negative emotion negative emotion experienced in the different conditions was regulation paradigm. Our aim was to analyze the parallel- self-rated on a 1–5 number scale. Further information isms and differences between these groups during regulation regarding the task is described in Appendix. of negative emotions using a data-driven, network-based approach for studying temporally coherent functional net- Outside-scanner behavioral measures. Trait Impulsivity works (Kim et al., 2009). Given the scarcity of studies was measured with the Spanish version of the UPPS-P parsing which neural alterations result from prolonged impulsive behavior scale (Whiteside & Lynam, 2001), which exposure to cocaine use or as a hallmark of both substance has shown sound reliability and internal and construct val- and behavioral addictive disorders, we leveraged compari- idity (Verdejo-García, Lozano, Moya, & Pérez-García, 2010). sons between CUD and GD groups to address this question. The scale contains 59 items that comprehensively assess Furthermore, we sought to explore if the neural un- different personality pathways leading to impulsive derpinnings of emotion regulation correlate with clinical and behavior: Sensation seeking (12 items), Lack of perseverance trait features related to addiction (e.g., severity, impulsivity). (10 items), Lack of premeditation (11 items), Negative ur- We focused our analysis in limbic, ventral frontostriatal, gency (12 items), and Positive urgency (14 items). DAN and executive networks (FPN) given their involvement in emotion regulation and their alteration in CUD and GD. Imaging data acquisition and preprocessing. We used a We hypothesized a downregulation of these networks during 3.0 T Intera Achieva Philips MRI scanner, equipped with an emotion regulation in both CUD and GD. We also explored eight-channel phased-array head coil. Details on acquisition differential patterns on CUD when compared with GD and parameters and preprocessing are presented in Appendix. HC, which might reflect neuroadaptive effects of prolonged cocaine use. Statistical analysis Behavioral data. Behavioral data were analyzed with Sta- METHODS tistical Package for the Social Sciences software version 24 (SPSS; Chicago, IL, USA). We conducted one-way Analysis Participants of Variance (ANOVA) to compare the three groups on relevant variables (Table 1). Interactions between in-scanner We recruited 17 participants with CUD and 16 with GD negative emotion intensity ratings for each condition from outpatient clinics, and 18 HC from the general com- (Observe, Maintain and Regulate) and subject groups were munity. Inclusion criteria were: DSM-IV-TR criteria for evaluated using a 333 mixed ANOVA analysis. Moreover, cocaine dependence or pathological gambling as measured participants' self-reported success in lowering their in- by the Structured Clinical Interview for DSM-IV Disorders- scanner negative emotion intensity was calculated by sub- Clinical Version (SCID-IV-CV) (First, Williams, Benjamin, tracting Regulate ratings from Maintain ratings (Success 5 & Spitzer, 1997) and ≥15 day abstinence tested via urine Maintain Regulate), while participants' reactivity during toxicological tests run twice weekly in those with cocaine emotional processing was computed as Reactivity 5 Main- dependence, or cross-validated reports from individuals with tain Observe. pathological gambling and their spouse/relatives. Exclusion criteria were meeting any other Axis I disorder and under- Independent component analysis. Group ICA (Calhoun going any addiction treatments <2 years before the study. et al., 2001) was performed with the Gift toolbox (v3.0 b, Controls were recruited from local employment agencies http://icatb.sourceforge.net) using the Infomax algorithm and were screened for any DSM-IV-TR criteria of substance (Bell & Sejnowski, 1995). Before ICA, voxel intensity was use disorder – but nicotine dependence—or pathological normalized, and data from all participants were pooled into gambling. All groups were required to be ≥18 years old, have a single dataset through a two-step data reduction approach an IQ score ≥80 (Kaufman & Kaufman, 1990), and no using principal component analysis to enable the analysis of history of head injury and neurological or nervous system large data sets. Thirty-six independent components based on diseases. the minimum description length criteria were selected (Li, Adalı, & Calhoun, 2007). Fifty ICA iterations were per- Measures and procedure formed by ICASSO (Himberg, Hyvärinen, & Esposito, 2004) to ensure stability of the estimated components. Finally, fMRI cognitive reappraisal task. We used a cognitive individual component maps and time courses were esti- reappraisal task described in Albein-Urios et al. (2013). The mated using a group ICA 3 back-reconstruction approach. 376 Table 1. Sociodemographic and clinical information CUD (N 5 15) GD (N 5 16) HC (N 5 17) Statistic, P Significant post-hoc test (pholm, Cohen's d) Age, Mean (SD) 37.6 (5.03) 33.13 (7.98) 32.35 (4.02) F(2,45) 5 3.59, 0.036 CUD>HC: 0.047, 1.162 Sex, N males (%) 14 (93.3) 14 (87.5) 16 (94.1) c2(2) 5 0.55, 0.759 - Years of education, Mean (SD) 9.73 (1.83) 9.94 (1.95) 10.35 (2.03) F(2,45) 5 0.43, 0.655 - Verbal IQ, Mean (SD) 101 (8.63) 101.06 (6.03) 106.06 (9.52) F(2,45) 5 2.06, 0.14 - Manipulative IQ, Mean (SD) 94.6 (10.19) 96.06 (11.11) 103.12 (11.19) F(2,45) 5 2.88, 0.067 - Age of onset, Mean (SD) 23.27 (8.2) 22.07 (9.18)* - t(28) 5 0.38, 0.709 - Monthly amount (CUD grams, GD 17.27 (26.81) 41.27 (40.92) - - - hours), Mean (SD) Duration of regular use (months), 59.97 (61.92) 23.2 (20.66) - - - Mean (SD) Severity, Mean (SD) 1000 (2403.37) 1174.4 (1896.24) - t(28) 5 0.25¶, 0.804 - Journal of Behavioral Addictions 11 (2022) 2, 373–385 UPPS, Mean (SD) N 5 14 Negative urgency 33.43 (7.07) 28.69 (4.38) 24.47 (4.68) F(2,44) 5 10.54, <0.001 CUD>HC: <0.001, 1.526 CUD>GD: 0.042, 0.82 GD>HC: 0.042, 0.93 Positive urgency 32.57 (10.62) 29.88 (6.7) 24.18 (6.91) F(2,44) 5 4.39, 0.018 CUD>HC: 0.019, 0.957 Sensation seeking 28.36 (7.73) 30.38 (3.5) 31.24 (8.21) F(2,44) 5 0.708, 0.498 - Lack of premeditation 23.43 (4.59) 26 (3.85) 22.76 (4.1) F(2,44) 5 2.71, 0.077 - Lack of perseverance 21.07 (4.92) 22.25 (3.07) 19.88 (3.94) F(2,44) 5 1.45, 0.247 - CUD, patients with cocaine use disorder; GD, patients with gambling disorder; HC, healthy controls; SD, standard deviation. *Missing data from one participant. ¶Between-group comparison done with a standardized severity measure. Journal of Behavioral Addictions 11 (2022) 2, 373–385 377 Analysis of the component spatial maps. One-sample t-tests corrected for multiple comparisons. Thus, any significant of the spatial maps generated by ICA were performed using results should be considered preliminary. the full sample at the second-level of analyses with SPM12 at P < 0.05 family-wise error (FWE)-corrected, to determine the Ethics brain regions that were significantly related to each compo- The study procedures were carried out in accordance with nent. Given the study hypothesis, the networks of interest the Declaration of Helsinki. The Human Research Ethics were the limbic, the ventral frontostriatal, the bilateral FPNs, Committee of the University of Granada approved the study. and the DAN, and were visually identified taking into ac- All subjects were informed about the study and all provided count the presence of core regions of these networks. informed consent. Data collection took place between the beginning of 2010 and the summer of 2012, and neuro- Analysis of the component time courses. General Linear imaging data from the participants included in this study Model (GLM) was applied on the component time courses using different analytical approaches have been previously for each subject using a design matrix representing the task. reported (Albein-Urios et al., 2014; Contreras-Rodríguez et This yielded a set of beta-weights representing the modu- al., 2015; Irizar, Albein-Urios, Martínez-González, Verdejo- lation of component time courses by the GLM regressors. Garcia, & Lorenzetti, 2020; Navas et al., 2017). In this study, The GLM design matrix used in these analyses included we applied ICA and compared for the first time neural separate regressors to model each of the conditions activation from the three groups (CUD, GD and HC) in the (Observe, Maintain and Regulate), which were convolved cognitive reappraisal task. with the hemodynamic response function. Time derivatives and parameters that modeled residual motion were also included. Then, we performed separate second-level group RESULTS analyses for the contrasts Maintain > Observe and Regulate > Maintain using the estimated beta-weights. Behavioral results Group comparisons were performed in SPSS by means of one-way ANOVAs with group as the between-subjects The demographical data of participants included in the final factor with three levels (CUD, GD, and HC). To study sample are summarized in Table 1. The CUD group had group differences during the experience of negative emo- significantly higher scores in the negative and positive ur- tions, reactivity in the limbic network was explored for the gency subscales of the UPPS-P compared to HC, and Maintain > Observe contrast, while the modulation of the significantly higher scores than the GD group only for the FPNs and the DAN was analyzed for the Regulate > negative urgency subscale. On the other hand, the GD group Maintain contrast in order to study group differences showed significantly higher scores than the HC group also in during the regulation of emotions. Finally, since the ventral the negative urgency subscale of the UPPS-P (Table 1). frontostriatal network was of clinical relevance, between- Regarding the in-scanner ratings, a main effect of con- group differences in the modulation of this network were dition was found (F(2,90) 5 78.85, P < 0.001), and planned studied during both contrasts. Benjamini-Hochberg FDR comparisons showed that Maintain differed from Observe, correction was used to correct for the comparison of indicating successful negative emotion induction during this multiple networks and contrasts, while Holm's correction condition for all groups (Maintain > Observe: F(1,45) 5 was used to correct for pair-wise post-hoc tests within each 137.93, P < 0.001), and Regulate differed from Maintain ANOVA. (Maintain > Regulate: F(1,45) 5 11.04, P 5 0.002), indi- We investigated how individual differences in the cating successful emotion regulation. No significant main modulation of these functional networks related to clinical effect of group (F(2,45) 5 2.28, P > 0.05) nor interaction variables, by means of Pearson's correlation analyses with effect between condition and group (F(4,90) 5 1.39, P > age of onset and severity. Severity of cocaine use and 0.05) were found (see Table 2 for the ratings mean and gambling exposure was measured via the Interview for standard deviation for each group). Research on Addictive Behaviour (Verdejo-García, López- Torrecillas, Aguilar De Arcos, & Pérez-García, 2005), that provides information on monthly amounts and total duration of regular use (more information regarding this Table 2. In-scanner ratings during the cognitive reappraisal task instrument can be found in Appendix). These two variables were multiplied in order to compute a composite measure CUD GD HC of cumulative use. Additional correlation analyses were Observe, Mean (SD) 1.78 (0.9) 1.69 (0.69) 1.73 (0.76) performed with the Reactivity and Success variables, and Maintain, Mean (SD) 3.77 (0.99) 3.59 (0.88) 3.07 (1) with the UPPS-P subscales. These correlation analyses were Regulate, Mean (SD) 3.29 (0.88) 3.16 (0.8) 2.59 (1) performed for each group separately. When there was a Reactivity, Mean (SD) 1.99 (1.1) 1.9 (0.96) 1.35 (1.02) significant correlation for one group, the correlation co- Success, Mean (SD) 0.48 (1.24) 0.44 (0.79) 0.49 (0.86) efficients were compared by means of the Fisher r-to-z CUD, patients with cocaine use disorder; GD, patients with transformation. Importantly, correlation analyses were gambling disorder; HC, healthy controls; SD, standard deviation. considered exploratory secondary analyses and were not Reactivity 5 Maintain – Observe; Success 5 Maintain – Regulate. 378 Journal of Behavioral Addictions 11 (2022) 2, 373–385 ICA results showed an increased activation of this network when having to regulate negative images compared to controls (Fig. 1A). The networks of interest were visually identified from the 36 Moreover, there was a trending group effect for the DAN components. The limbic network mainly included the (F(2,45) 5 3.652, P 5 0.034, pFDR-corr 5 0.068), due to a bilateral amygdala, putamen and hippocampi; the ventral decreased activation in participants with CUD compared to frontostriatal network included the ventral striatum and the HC while regulating negative emotions (t 5 2.697, pholm 5 medial prefrontal cortex (mPFC), including the rostral 0.029). On the other hand, participants with GD did not anterior cingulate cortex (rACC); the left and right FPN differ from those with CUD (t 5 1.271, pholm 5 0.318) or comprised either the left or the right middle frontal gyrus HC (t 5 1.432, pholm 5 0.318) (Fig. 1A). Finally, we found (MFG) and angular gyrus; and the DAN comprised, bilat- no significant effects in the FPNs (lFPN: F(2,45) 5 0.251, P erally, the inferior (IOG) and middle occipital gyri (MOG) 5 0.779, pFDR-corr 5 0.779; rFPN F(2,45) 5 0.6, P 5 0.553, extending to the superior parietal lobule, the fusiform gyri, pFDR-corr 5 0.663). the inferior frontal gyri (IFG), the lateral orbitofrontal cortex (OFC), and the midcingulate cortex (MCC) (see Fig. S1). Brain-behavior correlations We found significant group differences in the limbic There were no significant correlations with age of onset or network (F(2,45) 5 5.83, P 5 0.006, pFDR-corr 5 0.036), and severity measures. post-hoc tests showed that both participants with CUD (t 5 Regarding the in-scanner ratings, correlations with the 3.072, pholm 5 0.011) and GD (t 5 2.775, pholm 5 0.016) Reactivity variable were tested for those networks showing displayed decreased network activation in response to significant differences in the ANOVA for Maintain > negative images relative to HC (Fig. 1A). Participants with Observe, and with the Success variable for Regulate > CUD and GD did not differ between them (t 5 0.339, Maintain results. There was a significant positive correlation pholm 5 0.736). Regarding the ventral frontostriatal network, for the CUD group between Reactivity and the limbic groups differed only during the contrast Regulate > Main- network (r(13) 5 0.602, P 5 0.018, Fig. 1B). This correlation tain (F(2,45) 5 4.531, P 5 0.016, pFDR-corr 5 0.048), and was significantly higher for the CUD group when compared post-hoc tests showed that participants with GD signifi- with the HC group (Fisher Z 5 2.36, one-tailed P 5 0.009). cantly differed from HC (t 5 2.804, pholm 5 0.022) while the Moreover, there was a significant positive correlation for the difference between participants with CUD and HC trended HC group between Success and the DAN (r(15) 5 0.569, P towards significance (t 5 2.3, pholm 5 0.052). The clinical 5 0.017, Fig. 1B). groups did not differ between them (t 5 0.45, pholm 5 Finally, correlations were also performed with the UPPS-P 0.655). In this case, both participants with CUD and GD subscales. There was a negative correlation in the CUD Fig. 1. A) Boxplots representing the activation in each group (CUD, GD and HC) for each component and contrast with significant results. B) Scatterplots showing the correlation between activity in those networks showing significant between-group differences and the corre- sponding ratings reported for that contrast, separately for each group. CUD, patients with cocaine use disorder; GD, patients with gambling disorder; HC, healthy controls; DAN, dorsal attention network Journal of Behavioral Addictions 11 (2022) 2, 373–385 379 group between the limbic network and the sensation seeking frontostriatal networks may be considered a general effect of (r(12) 5 0.557, P 5 0.039) and positive urgency (r(12) 5 addiction. 0.574, P 5 0.032) subscales of the UPPS-P (Fig. 2). For The downregulation of the limbic network agrees with sensation seeking, the correlation coefficient of the CUD the pivotal role suggested for limbic regions in emotion group was significantly lower than that of HC (Fisher Z 5 regulation in neuropsychiatric disorders (Zilverstand et al., 2.28, one-tailed P 5 0.011). For positive urgency, the cor- 2017). In healthy participants, exposure to negative affective relation coefficient of the CUD group was significantly lower stimuli is associated with increased activation in limbic than that of GD (Fisher Z 5 1.95, one-tailed P 5 0.026). network structures such as the amygdala, the hippocampus Regarding the regulation of negative images, the DAN was or the striatum (García-García et al., 2016). In our sample, significantly correlated with negative urgency in the CUD this limbic response seems to be blunted, in line with pre- group (r(12) 5 0.614, P 5 0.019). This correlation was vious reports of attenuated neural response to emotional significantly higher than that of HC (Fisher Z 5 2.6, one- cues in individuals with CUD (Canterberry, Peltier, Brady, & tailed P 5 0.005) (Fig. 2). Hanlon, 2016). Moreover, in our exploratory correlations, we found an association between the downregulation of the limbic network and the Sensation Seeking and Positive Ur- DISCUSSION gency scores in the CUD group. This suggests a potential link to impulsivity variables related to exploration and Emotion regulation elicited differential modulation of limbic reward-related traits in addiction (Meade, Bell, Towe, & and ventral frontostriatal networks in people with CUD and Hall, 2020; Moreno-López et al., 2012; Yip, Worhunsky, GD relative to healthy controls. Both clinical groups showed et al., 2018), although these results were not corrected for downregulation of the limbic network during emotional multiple comparisons and should be considered preliminary. experience, and upregulation of the ventral frontostriatal On the other hand, we observed a higher upregulation of network during emotion regulation. These results suggest the ventral frontostriatal network in individuals with CUD that altered emotional reactivity in the limbic and ventral and GD compared to controls. This result may be interpreted Fig. 2. Scatterplots showing the correlation between activity in those networks showing significant between-group differences and the UPPS-P subscales, separately for each group. CUD, patients with cocaine use disorder; GD, patients with gambling disorder; HC, healthy controls; DAN, dorsal attention network 380 Journal of Behavioral Addictions 11 (2022) 2, 373–385 as a general signature of addiction, and suggests that the studies involving healthy controls (Buhle et al., 2013) and neuroadaptive effects of cocaine may target the same regions clinical samples (Picó-Pérez, Radua, Steward, Menchón, & that contribute to general addiction vulnerability. Regions of Soriano-Mas, 2017; Zilverstand et al., 2017). the ventral frontostriatal network (medial OFC and rACC) Our study has some limitations that need to be addressed have previously shown increased activation to highly in future studies. First, as in most studies with human clinical arousing stimuli in participants with elevated sensation groups, it was not possible to address the etiology of these seeking traits (Joseph, Liu, Jiang, Lynam, & Kelly, 2009). In activity pattern differences. We suggest that the DAN is the this regard, compromised regulation of drug wanting has only network that seems to be specifically impacted by been suggested to explain the loss of control over drug intake cocaine use instead of a general addiction trait. Nonetheless, in addiction (Goldstein & Volkow, 2011), and an increased this result is only trending towards significance after multiple frontostriatal activation when participants are expected to comparison correction, and is limited by the absence of exert regulation might underlie this deficit. significant differences with the GD group. Similarly, the We also observed a trend towards a significant down- sample size was modest, due to the difficulty of recruiting regulation of the DAN in the CUD group compared to HC, homogeneous groups fulfilling inclusion and exclusion which was in agreement with previous studies reporting the criteria in clinical groups within the time of the project. Thus, alteration of this network in addiction (Zilverstand et al., the results of this study should be considered preliminary 2016, 2018). DAN downregulation in GD was observed but and replication is needed before drawing firm conclusions. to a lower extent, which might suggest a neuroadaptive effect Also, we did not acquire objective psychophysiological of cocaine in brain function rather than a general addiction measures such as electroencephalography (EEG) or heart- effect. Nonetheless, the pattern of activation in GD seems to rate variability, which would allow for a more robust verifi- overlap to a greater extent with CUD than HC participants, cation of reappraisal success, complementing the in-scanner so this interpretation should be taken with caution. More- reported scores. Future studies should consider including this over, other factors may also influence differences between type of measures in their project design to overcome the self- clinical groups, such as differential predispositions (impul- report issues abovementioned. Finally, the cross-sectional sivity traits) or comorbid disorders (although this was design precludes a causal interpretation of substance effects controlled by including only participants with no current given that other variables associated with substance use DSM-IV-TR Axis I comorbidities). Yip, Gross, et al. (2018) behavior and contextual factors, as compared to gaming also observed a downregulation of the cingulo-parietal associated ones, may be affecting cocaine effects beyond network during emotion regulation using ICA in a negative substance intake. The lack of effects in the FPNs in CUD emotion regulation task. However, there were no significant contrast with previous reports (Barrós-Loscertales et al., effects in the other IC identified in their study, such as the 2020; Costumero et al., 2015, 2018). However, these previous dorsolateral and dorsomedial inhibitory networks. These IC studies had shown FPN alterations in CUD related to reac- partially overlap with ours and suggest a pattern of down- tivity to external emotional stimuli involving drug and regulation in posterior portions of the DAN. Interestingly, reward related effects, but not general negative stimuli. our study shows that HC's upregulation of the DAN was Future studies may clarify the role of different frontoparietal associated with reduced affective scores to negative stimuli networks (DAN vs FPN) in addiction regarding emotion during emotion regulation, suggesting that subjective expe- reactivity and regulation. rience during regulation aligns with appropriate DAN acti- vation. On the other hand, in participants with CUD, Negative Urgency was related to increased DAN upregula- CONCLUSIONS tion, which suggests an opposite effect of cocaine in DAN- impulsivity related effects. Nonetheless, these interpretations Cocaine use and gambling disorders have overlapping pat- should be taken with caution and require further replication. terns of brain network's activation during negative emotion These results contrast with a lack of between-group processing and regulation involving downregulation of differences in the subjective scoring of stimulus. This is limbic and upregulation of ventral frontostriatal networks similar to previous reports (Albein-Urios et al., 2013; Yip, compared to HC. Moreover, DAN activation during Gross, et al., 2018). Zilverstand et al. (2017) suggest in their emotion regulation was reduced in participants with CUD review that it may be attributed to sensitivity of the proce- compared to HC. Finally, the patterns of activation in these dure employed for subjective ratings during fMRI. This may networks were further modulated by impulsivity traits. be affecting our results, but the fMRI paradigm was the same These findings illustrate both overlapping and unique effects in every group, which guarantees that the effects are com- of substance versus behavioral addictions on neural net- parable across groups. On the other hand, it might reflect works underpinning emotion regulation. limited introspection and self-awareness in individuals with addiction as was suggested by Goldstein and Volkow (2011) and Moeller et al. (2010). Nonetheless, the interpretation of Funding sources: The study was funded by the grant functional alterations of brain networks during emotion COPERNICO from the Spanish Ministry of Health/National regulation in the absence of between-group differences in Drugs Plan, and by the Norte Portugal Regional Operational subjective experience aligns with those provided across Programme (NORTE 2020) under the Portugal 2020 Journal of Behavioral Addictions 11 (2022) 2, 373–385 381 Partnership Agreement (NORTE-01-0145-FEDER-000013 Academic Press. https://doi.org/10.1016/B978-0-12-800212- and NORTE-01-0145-FEDER-000023). VC was supported 4.00020-0. by the grant PID2019-105077RJ-I00/AEI/10.13039/ Barrós-Loscertales, A., Costumero, V., Rosell-Negre, P., Fuentes- 501100011033. JVR was supported by the Spanish Ministry Claramonte, P., Llopis-Llacer, J.-J., & Bustamante, J. C. (2020). of Science and Innovation (IJC2019-041916-I). AVG was Motivational factors modulate left frontoparietal network dur- supported by a Medical Research Future Fund fellowship ing cognitive control in cocaine addiction. Addiction Biology, (MRF1141214). ABL has received research support from the 25(4), e12820. https://doi.org/10.1111/ADB.12820. Universitat Jaume I (UJI-B2020-23). MPP was supported by Bell, A. J., & Sejnowski, T. J. (1995). An information-maximization the Spanish Ministry of Universities, with funds from the approach to blind separation and blind deconvolution. Neural European Union - NextGenerationEU (MAZ/2021/11). Computation, 7(6), 1129–1159. https://doi.org/10.1162/NECO. 1995.7.6.1129. Authors' contribution: JVR, NAU, JMMG, and AVG were Buhle, J. T., Silvers, J. a, Wager, T. D., Lopez, R., Onyemekwu, C., responsible for the study concept and design. JVR and NAU Kober, H., . . . Ochsner, K. N. (2013). Cognitive reappraisal of contributed to the data acquisition. MPP and VC performed emotion: A meta-analysis of human neuroimaging studies. the statistical analysis. CSM and ABL assisted with data Cerebral Cortex, 24(11), 2981–2990. https://doi.org/10.1093/ analysis and interpretation of findings. MPP and ABL cercor/bht154. drafted the manuscript. CSM, VC and AVG provided critical Bustamante, J.-C., Barrós-Loscertales, A., Costumero, V., Fuentes- revision of the manuscript for important intellectual con- Claramonte, P., Rosell-Negre, P., Ventura-Campos, N., . . . tent. All authors critically reviewed content and approved Ávila, C. (2014). Abstinence duration modulates striatal func- final version for publication. tioning during monetary reward processing in cocaine patients. Addiction Biology, 19(5), 885–894. https://doi.org/10.1111/ Conflict of interest: The authors declare no conflict of ADB.12041. interest. Calhoun, V. D., Adali, T., Pearlson, G. D., & Pekar, J. J. (2001). A method for making group inferences from functional MRI data using independent component analysis. Human Brain Map- ping, 14(3), 140. https://doi.org/10.1002/HBM.1048. REFERENCES Canterberry, M., Peltier, M. R., Brady, K. T., & Hanlon, C. A. (2016). Attenuated neural response to emotional cues in Albein-Urios, N., Martinez-González, J. M., Lozano, Ó., Clark, L., & cocaine-dependence: A preliminary analysis of gender differ- Verdejo-García, A. (2012). Comparison of impulsivity and ences. The American Journal of Drug and Alcohol Abuse, 42(5), working memory in cocaine addiction and pathological 577. https://doi.org/10.1080/00952990.2016.1192183. gambling: Implications for cocaine-induced neurotoxicity. Drug Contreras-Rodríguez, O., Albein-Urios, N., Vilar-López, R., Perales, and Alcohol Dependence, 126(1–2), 1–6. https://doi.org/10. J. C., Martínez-Gonzalez, J. M., Fernández-Serrano, M. J., . . . 1016/j.drugalcdep.2012.03.008. Verdejo-García, A. (2015). Increased corticolimbic connectivity Albein-Urios, N., Pilatti, A., Lozano, Ó., Martínez-González, J. M., in cocaine dependence versus pathological gambling is associ- & Verdejo-García, A. (2014). The value of impulsivity to define ated with drug severity and emotion-related impulsivity. subgroups of addicted individuals differing in personality Addiction Biology, Kalivas 2007, n/a-n/a. https://doi.org/10. dysfunction, craving, psychosocial adjustment, and wellbeing: A 1111/adb.12242. latent class analysis. Archives of Clinical Neuropsychology, 29(1), Costumero, V., Bustamante, J. C., Rosell-Negre, P., Fuentes, P., 38–46. https://doi.org/10.1093/arclin/act072. Llopis, J. J., Ávila, C., & Barrós-Loscertales, A. (2015). Reduced Albein-Urios, N., Verdejo-Román, J., Asensio, S., Soriano-Mas, C., activity in functional networks during reward processing is Martínez-González, J. M., & Verdejo-García, A. (2014). Re- modulated by abstinence in cocaine addicts. Addiction Biology, appraisal of negative emotions in cocaine dependence: 22(2), 479–489. https://doi.org/10.1111/adb.12329. Dysfunctional corticolimbic activation and connectivity. Costumero, V., Rosell-Negre, P., Bustamante, J. C., Fuentes-Clar- Addiction Biology, 19(3), 415–426. https://doi.org/10.1111/j. amonte, P., Llopis, J. J., Ávila, C., & Barrós-Loscertales, A. 1369-1600.2012.00497.x. (2018). Left frontoparietal network activity is modulated by Albein-Urios, N., Verdejo-Román, J., Soriano-Mas, C., Asensio, S., drug stimuli in cocaine addiction. Brain Imaging and Behavior, Martínez-González, J. M., & Verdejo-García, A. (2013). Cocaine 12(5), 1259–1270. https://doi.org/10.1007/s11682-017-9799-3. users with comorbid Cluster B personality disorders show Fernández-Serrano, M. J., Perales, J. C., Moreno-López, L., Pérez- dysfunctional brain activation and connectivity in the emotional García, M., & Verdejo-García, A. (2011). Neuropsychological regulation networks during negative emotion maintenance and profiling of impulsivity and compulsivity in cocaine dependent reappraisal. European Neuropsychopharmacology, 23(12), 1698– individuals. Psychopharmacology 2011 219:2, 219(2), 673–683. 1707. https://doi.org/10.1016/j.euroneuro.2013.04.012. https://doi.org/10.1007/S00213-011-2485-Z. Barrós-Loscertales, A. (2016). Structural and functional aspects First, M. B., Williams, J. B. W., Benjamin, L. S., & Spitzer, R. L. of stimulant misuse and addiction. Neuropathology of drug (1997). User’s guide for the structured clinical interview for addictions and substance misuse (Vol. 2, pp. 209–219). DSM-IV Axis II per- sonality disorders: SCID-II. 40. 382 Journal of Behavioral Addictions 11 (2022) 2, 373–385 García-García, I., Kube, J., Gaebler, M., Horstmann, A., Villringer, (2020). Dimensions of impulsivity in gambling disorder. Sci- A., & Neumann, J. (2016). Neural processing of negative entific Reports, 10(1). https://doi.org/10.1038/S41598-019- emotional stimuli and the influence of age, sex and task-related 57117-Z. characteristics. Neuroscience and Biobehavioral Reviews, 68, Michalczuk, R., Bowden-Jones, H., Verdejo-Garcia, A., & Clark, L. 773–793. https://doi.org/10.1016/J.NEUBIOREV.2016.04.020. (2011). Impulsivity and cognitive distortions in pathological Goldstein, R. Z., & Volkow, N. D. (2011). Dysfunction of the gamblers attending the UK national problem gambling clinic: A prefrontal cortex in addiction: Neuroimaging findings and preliminary report. Psychological Medicine, 41(12), 2625. clinical implications. Nature Reviews Neuroscience 2011 12:11, https://doi.org/10.1017/S003329171100095X. 12(11), 652–669. https://doi.org/10.1038/nrn3119. Moeller, S. J., Maloney, T., Parvaz, M. A., Alia-Klein, N., Woicik, P. Himberg, J., Hyvärinen, A., & Esposito, F. (2004). Validating the A., Telang, F., . . . Goldstein, R. Z. (2010). Impaired insight in independent components of neuroimaging time series via cocaine addiction: Laboratory evidence and effects on cocaine- clustering and visualization. Neuroimage, 22(3), 1214–1222. seeking behaviour. Brain, 133(5), 1484. https://doi.org/10.1093/ https://doi.org/10.1016/J.NEUROIMAGE.2004.03.027. BRAIN/AWQ066. Irizar, P., Albein-Urios, N., Martínez-González, J. M., Verdejo-Garcia, Moltó, J., Montañés, S., Poy, R., Segarra, P., Pastor, M. C., & A., & Lorenzetti, V. (2020). Unpacking common and distinct Tormo, M. (1999). Un método para el estudio experimental de neuroanatomical alterations in cocaine dependent versus patho- las emociones: el International affective picture system (IAPS). logical gambling. European Neuropsychopharmacology, 33, 81–88. Adaptación española. Rev. Psicol. Gen. Apl, 52, 55–87. https://doi.org/10.1016/j.euroneuro.2020.01.019. Moreno-López, L., Catena, A., Fernández-Serrano, M. J., Delgado- Joseph, J. E., Liu, X., Jiang, Y., Lynam, D., & Kelly, T. H. (2009). Rico, E., Stamatakis, E. A., Pérez-García, M., & Verdejo-García, Neural correlates of emotional reactivity in sensation seeking. A. (2012). Trait impulsivity and prefrontal gray matter re- Psychological Science, 20(2), 215. https://doi.org/10.1111/J.1467- ductions in cocaine dependent individuals. Drug and Alcohol 9280.2009.02283.X. Dependence, 125(3), 208– 214. https://doi.org/10.1016/J. Kaufman, A., & Kaufman, N. (1990). K-BIT: Kaufman brief intel- DRUGALCDEP.2012.02.012. ligence test. In A. G. Service (Ed), Los Angeles, CA: SAGE Navas, J. F., Contreras-Rodríguez, O., Verdejo-Román, J., Per- PublicationsSage CA. https://journals.sagepub.com/doi/abs/10. andrés-Gómez, A., Albein-Urios, N., Verdejo-García, A., & 1177/153450849902401-412?journalCode5aeia. Perales, J. C. (2017). Trait and neurobiological underpinnings Kim, D. I., Mathalon, D. H., Ford, J. M., Mannell, M., Turner, J. A., of negative emotion regulation in gambling disorder. Addiction, Brown, G. G., . . . Calhoun, V. D. (2009). Auditory oddball 112(6), 1086–1094. https://doi.org/10.1111/add.13751. deficits in schizophrenia: An independent component analysis Ochsner, K. N., Bunge, S. A., Gross, J. J., & Gabrieli, J. D. E. (2002). of the fMRI multisite function BIRN study. Schizophrenia Rethinking feelings: An FMRI study of the cognitive regulation Bulletin, 35(1), 67–81. https://doi.org/10.1093/schbul/sbn133. of emotion. Journal of Cognitive Neuroscience, 14(8), 1215– Kober, H., Lacadie, C. M., Wexler, B. E., Malison, R. T., Sinha, R., & 1229. https://doi.org/10.1162/089892902760807212. Potenza, M. N. (2016). Brain activity during cocaine craving and Picó-Pérez, M., Radua, J., Steward, T., Menchón, J. M., & Soriano- gambling urges: An fMRI study. Neuropsychopharmacology, Mas, C. (2017). Emotion regulation in mood and anxiety 41(2), 628. https://doi.org/10.1038/NPP.2015.193. disorders: A meta-analysis of fMRI cognitive reappraisal Lang, P., Bradley, M., & Cuthbert, B. (2005). International affective studies. Progress in Neuro-Psychopharmacology and Biological picture system (IAPS): Digitized photographs, instruction Psychiatry, 79, 96–104. https://doi.org/10.1016/j.pnpbp.2017. manual and affective ratings. Technical Report A-6. 2005. 06.001. Lee, R. S. C., Hoppenbrouwers, S., & Franken, I. (2019). A sys- Reuter, J., Raedler, T., Rose, M., Hand, I., Gläscher, J., & Büchel, C. tematic meta-review of impulsivity and compulsivity in addic- (2005). Pathological gambling is linked to reduced activation of tive behaviors. Neuropsychology Review, 29(1), 14–26. https:// the mesolimbic reward system. Nature Neuroscience 2005 8:2, doi.org/10.1007/S11065-019-09402-X. 8(2), 147–148. https://doi.org/10.1038/nn1378. Li, Y., Adalı, T., & Calhoun, V. D. (2007). Estimating the number of Sinha, R. (2009). Modeling stress and drug craving in the labora- independent components for functional magnetic resonance tory: Implications for addiction treatment development. imaging data. Human Brain Mapping, 28(11), 1251. https://doi. Addiction Biology, 14(1), 84. https://doi.org/10.1111/J.1369- org/10.1002/HBM.20359. 1600.2008.00134.X. Mackey, S., & Paulus, M. (2013). Are there volumetric brain dif- Torres, A., Catena, A., Megías, A., Maldonado, A., Cándido, A., ferences associated with the use of cocaine and amphetamine- Verdejo-García, A., & Perales, J. C. (2013). Emotional and non- type stimulants? Neuroscience and Biobehavioral Reviews, 37(3), emotional pathways to impulsive behavior and addiction. 300–316. https://doi.org/10.1016/J.NEUBIOREV.2012.12.003. Frontiers in Human Neuroscience, 0(FEB), 43. https://doi.org/ Meade, C. S., Bell, R. P., Towe, S. L., & Hall, S. A. (2020). Cocaine- 10.3389/FNHUM.2013.00043. related alterations in fronto-parietal gray matter volume Verdejo-García, Alcázar-Córcoles, & Albein-Urios (2019). Neuro- correlate with trait and behavioral impulsivity. Drug and psychological interventions for decision-making in addiction: A Alcohol Dependence, 206, 107757. https://doi.org/10.1016/J. systematic review. Neuropsychology Review, 29(1), 79–92. DRUGALCDEP.2019.107757. https://doi.org/10.1007/S11065-018-9384-6. Mestre-Bach, G., Steward, T., Granero, R., Fernández-Aranda, F., Verdejo-Garcia, A., Clark, L., Verdejo-Román, J., Albein-Urios, N., Mena-Moreno, T., Vintró-Alcaraz, C., . . . Jiménez-Murcia, S. Martinez-Gonzalez, J. M., Gutierrez, B., & Soriano-Mas, C. Journal of Behavioral Addictions 11 (2022) 2, 373–385 383 (2015). Neural substrates of cognitive flexibility in cocaine and APPENDIX gambling addictions. The British Journal of Psychiatry, 207(2), 158–164. https://doi.org/10.1192/BJP.BP.114.152223. Verdejo-García, A. J., López-Torrecillas, F., Aguilar De Arcos, F., & METHODS Pérez-García, M. (2005). Differential effects of MDMA, cocaine, and cannabis use severity on distinctive components of the fMRI cognitive reappraisal task executive functions in polysubstance users: A multiple regres- Stimuli were presented through magnetic resonance- sion analysis. Addictive Behaviors, 30(1), 89–101. https://doi. compatible liquid crystal display goggles (Resonance Tech- org/10.1016/J.ADDBEH.2004.04.015. nology, Northridge, CA, USA). We used 24 stimuli that were Verdejo-García, A., Lozano, Ó., Moya, A., & Pérez-García. (2010). extracted from the International Affective Picture System Psychometric properties of a Spanish version of the UPPS-P (Lang, Bradley, & Cuthbert, 2005): eight neutral pictures impulsive behavior scale: Reliability, validity and association (e.g. household objects) which were presented in the with trait and cognitive impulsivity. Journal of Personality Observe condition, and 16 highly unpleasant arousing pic- Assessment, 92(1), 70–77. https://doi.org/10.1080/002238909 tures (e.g. mutilations) that were presented in the Maintain 03382369. and Regulate conditions. The images were selected accord- Whiteside, S. P., & Lynam, D. R. (2001). The five factor model and ing to IAPS Spanish normative values for valence and impulsivity: Using a structural model of personality to un- arousal (Moltó et al., 1999), and their mean values can be derstand impulsivity. Personality and Individual Differences, found in Albein-Urios et al. (2013). 30(4), 669–689. https://doi.org/10.1016/S0191-8869(00) The task consisted of 12 blocks: four blocks for each of 00064-7. the three conditions. Conditions were pseudo-randomized Worhunsky, P. D., Potenza, M. N., & Rogers, R. D. (2017). Alter- along the task to avoid the induction of sustained mood ations in functional brain networks associated with loss-chasing states. Each block began with the instruction prompt in gambling disorder and cocaine-use disorder. Drug and (‘Observe’ or ‘Maintain’ or ‘Regulate’) presented in the Alcohol Dependence, 178, 363. https://doi.org/10.1016/J. middle of the screen during 4s. After the prompt, partici- DRUGALCDEP.2017.05.025. pants viewed two different pictures of equal valence for 10s Xu, J., Potenza, M. N., & Calhoun, V. D. (2013). Spatial ICA reveals each. Immediately after the second picture, the intensity of functional activity hidden from traditional fMRI GLM-based the negative emotion experienced was self-rated on a 1–5 analyses. Frontiers in Neuroscience, 7(7 AUG). https://doi.org/ number scale that appeared for 5s (where 1 is ‘neutral’ and 5 10.3389/FNINS.2013.00154. is ‘extremely negative’), recorded through a five-button box, Yip, S. W., Gross, J. J., Chawla, M., Ma, S.-S., Shi, X.-H., Liu, L., . . . Evoke Response Pad System (Resonance Technology Inc.). Zhang, J. (2018). Is neural processing of negative stimuli altered Each block was followed by 10s of a cross-fixation baseline in addiction independent of drug effects? Findings from drug- to minimize carryover effects. naïve Youth with internet gaming disorder. Neuro- psychopharmacology, 43(6), 1364. https://doi.org/10.1038/NPP. Imaging data acquisition and preprocessing 2017.283. During acquisition, a T2p -weighted echo-planar imaging (EPI) Yip, S. W., Worhunsky, P. D., Xu, J., Morie, K. P., Constable, R. T., was obtained [repetition time (TR) 5 2000 ms, echo time (TE) Malison, R. T., . . . Potenza, M. N. (2018). Gray-matter re- 5 35 ms, field of view (FOV) 5 2303230 mm, 96396 matrix, lationships to diagnostic and transdiagnostic features of drug flip angle 5 908, 21 4-mm axial slices, 1-mm gap, 234 scans]. and behavioral addictions. Addiction Biology, 23(1), 394. A sagittal three-dimensional T1-weighted turbo-gradient-echo https://doi.org/10.1111/ADB.12492. sequence (160 slices, TR 5 8.3 ms, TE 5 3.8 ms, flip angle 5 Zilverstand, A., Huang, A. S., Alia-Klein, N., & Goldstein, R. Z. 88, FOV 5 2403240, 1 mm3 voxels) was obtained in the same (2018). Neuroimaging impaired response inhibition and experimental session for anatomical reference. salience attribution in human drug addiction. A systematic The functional images were analyzed using Statistical review. Neuron, 98(5), 886–903. https://doi.org/10.1016/j. Parametric Mapping (SPM12) software (https://www.fil.ion. neuron.2018.03.048. ucl.ac.uk/spm/), running under Matlab R2009 (MathWorks, Zilverstand, A, Parvaz, M. A., & Goldstein, R. Z. (2017). Neuro- Natick, MA, USA). Preprocessing included slice timing imaging cognitive reappraisal in clinical populations to define correction, head motion correction, normalization to the neural targets for enhancing emotion regulation. A systematic EPI template in the Montreal Neurological Institute (MNI) review. Neuroimage, 151, 105–116. https://doi.org/10.1016/j. space (voxel size of 23232 mm), and spatial smoothing by neuroimage.2016.06.009. convolution with a 3D Gaussian kernel (full width at half Zilverstand, A., Parvaz, M. A., Moeller, S. J., & Goldstein, R. Z. maximum 5 8 mm). The realignment parameters were (2016). Cognitive interventions for addiction medicine: Un- inspected, and participants were excluded when they had derstanding the underlying neurobiological mechanisms. more than 2 mm or degrees of movement in any of the six Progress in Brain Research, 224, 285. https://doi.org/10.1016/BS. directions, leaving a final sample of 15 CUD patients, 16 GD PBR.2015.07.019. patients and 17 HC (Table 1). 384 Journal of Behavioral Addictions 11 (2022) 2, 373–385 Fig. A1. Networks of interest identified through group ICA. FPN 5 frontoparietal network; DAN 5 dorsal attention network. Coordinates are x, y, z MNI coordinates Journal of Behavioral Addictions 11 (2022) 2, 373–385 385 Interview for Research on Addictive Behavior drug ingested/gambling in each episode of use (number of grams for cocaine; and number of hours for gambling); (2) This instrument evaluates, by means of a brief interview, the the frequency of these consumption episodes per month: quantity (dosing), frequency (consumption episodes by daily, between one and three times upon a week, once a month), and chronicity (years of duration) of the use of a week, between one and three times upon a month, or once a series of substances (such as cocaine) that can produce month; and (3) the number of years that elapsed since the physical or psychological dependence, as well as of the onset of the use. Then, composite severity scores can be exposure to gambling. For each of this, the following in- calculated from these variables. formation is requested: (1) the average amount of each target Open Access. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (https:// creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and reproduction in any medium for non-commercial purposes, provided the original author and source are credited, a link to the CC License is provided, and changes – if any – are indicated.