Reduction in Valence-Differentiating Neural Activities in Binge Drinking and Its Comorbidities: An Exploratory Study of the Human Connectome Project Data

• Shared and distinct responses to win and loss stimuli 

Reward and punishment are both salient. However, reward reinforces whereas punishment deters the repetition of the same behavior [1-4]. Reward and punishment elicit both shared and distinct neural activities [5-7], as we demonstrated in a meta-analysis of win and loss processing during the monetary incentive delay task or MIDT [8]. Win and loss anticipation engaged a shared network of bilateral anterior insula, striatum, thalamus, supplementary motor area, and precentral gyrus. Win and loss outcome engaged higher activity in Medial Orbitofrontal Cortex (mOFC) and dorsal Anterior Cingulate Cortex (ACC), respectively. Findings of higher dACC, anterior insula, and striatum activation during loss vs. nil anticipation were also noted in an earlier meta-analysis of the MIDT [9]. Win vs. loss in a card guessing game engaged the Medial Prefrontal Cortex (mPFC), including the mOFC, and bilateral nucleus accumbens [10]. Another study of delay discounting found that Intertemporal decision-making recruited the fronto-parietal network, whilst reward and loss anticipation were related to activation in the salience network [11]. Together, although the findings varied across behavioral paradigms and not all studies contrasted win and loss trials, it seems that both shared and distinct regional activities support win and loss processing. 

• Binge drinking and valence-differentiating neural processes 

Binge drinking is associated with a host of neurocognitive effects, including disruption of memory [12] and hippocampal neurogenesis [13] as well as impairment in executive control and regional activities and connectivities underlying executive control [14-18]. Some have specifically investigated reward and punishment processing in binge and/or heavy drinkers. An earlier study reported a positive and negative correlation, respectively, between reward and punishment sensitivity as evaluated with the Sensitivity to Punishment and Sensitivity to Reward Questionnaire with alcohol use severity in college students [19]. Another work similarly showed the contribution of hyperactive reward and hypoactive punishment response along with weakened impulse regulation to alcohol misuse [20]. Other studies highlighted the roles of elevated reward sensitivity and social isolation stress in adolescent binge drinking [21] and associated the severity of Alcohol Use Disorder (AUD) to behavioral impairment and reduction in physiological arousal, as inflected in Skin Conductance Response (SCR), in the Iowa Gambling Task [22]. Further, in the latter study, although SCRs did not differ in relation to AUD severity for reward outcomes, behavioral activation system trait was associated with higher anticipatory arousal to disadvantageous deck choices among those with lower AUD severity. Alcohol dependent and binge drinkers showed lower performance in implicit and explicit reward learning than low risk drinkers, supporting chronic reward deficit as a mechanism in the development of AUD [23]. A more recent work showed deficits, including enhanced negative reinforcement as well as diminished reward learning and choice consistency during probabilistic contingencies in people with AUD [24]. Thus, both behavioral and physiological data suggest altered reward and punishment processing in relation to alcohol misuse. Many human imaging studies focused on dysfunctional neural processing of reward and punishment in heavy and problem drinkers [see [25-28] for reviews and meta-analyses]. For instance, in an earlier study of the “Doors” task (guessing to win), binge relative to non-binge drinkers exhibited greater ventral striatal activation during reward receipt relative to loss and less ventral striatal dACC connectivity, perhaps reflecting deficient regulation of activation to rewards as compared to losses [29]. Our recent work investigated whether binge and non-bingers drinkers differed in reward and/or loss processing in a gambling task [30]. Compared with non-bingers, bingers showed significantly faster response during post-loss than during post-win trials. Further, loss and fronto-striatal reactivity characterized the relationship between rule-breaking behavior and the severity of alcohol use. In a recent work, neural responses to neither reward nor loss anticipation in the MIDT distinguished participants with AUD, healthy participants with family history of AUD and those without AUD [31]. Another study of adolescents showed that higher Alcohol Use Disorder Identification Test scores were associated with reduced neural differentiation between reward vs. punishment feedback in the striatum, posterior cingulate cortex, and parietal cortex during a probabilistic association task [32]. To our knowledge, the latter and Crane et al. are the only studies to date that focused on regional activities differentiating reward and punishment in relation to alcohol misuse. On the other hand, the findings vary and may be paradigm specific and neither study distinguished the behavioral consequences following reward vs. punishment. Thus, it remains to be seen how regional activities distinguishing reward and punishment are altered in link with alcohol misuse. 

• The present study 

                Extant studies have largely focused on reward or punishment responses in elucidating the pathophysiology of AUD and comorbidities. Reward and punishment are both behaviorally salient but to guide behaviors in opposite directions. Deficiency in distinguishing reward and punishment may reflect neural dysfunction central to mal-adaptive behavior, as observed in many clinical conditions. For instance, while a meta-analysis showed punishment relative to reward sensitivity a better predictor of depression and anxiety across neurotypical individuals and those with an acute and/or remitted clinical condition [33], patients with depression found both rewarding and punishing stimuli less salient than healthy participants [34]. The latter study was relevant particularly because of the high comorbidities of alcohol misuse and emotional disorders, as has been shown in numerous earlier studies (see Centanni et al., 2019; Tietbohl-Santos et al., 2019 for a review). Emotional dysfunction perpetuates drinking which, in turn, aggravates anxiety and depression. Is it possible that diminished behavioral and neural differentiation of reward and punishment represents a marker of alcohol misuse and other clinical conditions?. We tested this hypothesis using the Human Connectome Project (HCP) data collected of a gambling task in young adults. We identified binge drinkers and non-binge drinkers to examine group differences in behavioral and neural responses to reward vs. punishment and investigated how they relate to drinking severity and characteristics of negative affect, which is frequently comorbid with alcohol misuse.

• Dataset and Demographics 

With permission from the HCP [35], as in our previous work [36-38], we employed the 1200 Subjects Release (S1200). Binge drinking was defined as having ≥ 4/5 drinks for women/men on a single occasion [39]. The binge drinking group comprised 181 adults who reported binge drinking at least once a week for the last 12 months (132 men, 72.9%) [40]. A total of 314 adults reported no binge drinking in the prior year. However, 26 of the 314 met criteria for life-time alcohol abuse or dependence and were excluded, leaving 288 adults (97 men, 33.7%) in the non-binge drinking group. Thus, the data of a total of 469 adults (229 men; mean ± SD = 27.9 ± 3.6 years; 240 women, 29.8 ± 3.7 years) were included in this study, with more men than women in the binge drinking group (x² = 68.52, p < 0.001, chi-square test). In male (n=132) and female (n=49) bingers, 75 and 20 each met DSM-IV criteria for alcohol abuse or dependence. Analysis of variance (ANOVA) showed a significant group (F = 5.76, p = 0.017) and sex (F = 13.63, p < 0.001) main as well as group × sex interaction (F = 5.12, p = 0.024) effect in age. Age and sex were included as covariates in the analyses of all subjects and age alone was included as a covariate in the analyses of men and women separately. All subjects were physically healthy with no severe neurodevelopmental, neuropsychiatric or neurological disorders. The HCP study was approved by the Washington University Institutional Review Board (IRB #201204036). 

• Clinical Measures 

The HCP data comprised 15 inter-related drinking metrics to assess the severity of alcohol use. We performed a Principal Component Analysis (PCA) on the 15 measures and identified one Principal Component (PC1) with an eigenvalue > 1 and accounting for 60.73% of the variance. Supplementary Table S1 shows the mean ± SD of the drinking measures and PC1 (“drinking severity PC1,” henceforth) and the statistics of group × sex ANOVA. All participants were assessed with the NIH-Toolbox Emotion Measures – 18+ (i.e., > 18 years old) battery – which consists of 4 sub-domains: negative affect, psychological well-being, stress and self-efficacy, and social relationships. We focused on the negative affect domain, which contains six scales - anger-affect, anger-hostility, anger-physical aggression, fear-affect, fear-somatic arousal, and sadness. There are 6 items in fear somatic subscale (Supplementary Methods) each scored from 1 (Not at all) to 5 (Extremely), with a higher score indicating severe fear somatic problem. All analyses are to be conducted with T-scores, which are standard scores in which a T-score of 50 represents the mean of the US general population (based on the 2010 Census) and 10 T-score units represents one standard deviation. The fear somatic T score ranged from 40.1 to 79.4 for the current sample. We performed a group by sex ANOVA of the score with age as a covariate and followed with simple effects analyses to examine sex differences in each drinking group (e.g., male vs. female bingers) and group differences in each sex (e.g., male bingers vs. non-bingers). 

• Imaging protocol and gambling task for fMRI 

Imaging protocols are described in Supplementary Methods [36]. Participants completed two runs of a gambling task each with 4 blocks (~3 m and 12 s each run) – 2 each of reward and punishment, each with more win and loss trials – and a fixation period (baseline, 15 s) between blocks [41]. In the gambling task, participants guessed whether the number of a “mystery” card (1 to 9) was larger or smaller than 5 by pressing a corresponding button. Correct and incorrect guess led to $1 win and $0.5 loss, with a wash for mystery card number = 5. Reaction Time (RT) was recorded throughout the experiment. Please see the Supplement for details. We performed a two-way (group ´ sex) repeated measures ANOVA of the Reaction Time (RT) of trials following loss (post-loss), win (post-win) and neutral (post-neu) to assess loss and win reactivity. 

• Imaging data modeling and statistics 

We followed published routines [42,43] in image data preprocessing and modeled the BOLD signals to identify regional responses to reward and punishment blocks, relative to the baseline (Supplementary Methods). In group analyses, we performed a one-sample t test of the contrasts and evaluated the results with voxel p < 0.05, FWE-corrected.Functional regions of interest (ROIs) were defined based on clusters obtained from whole-brain analyses. We used MarsBar (http://marsbar.sourceforge.net/) to derive for individual subjects the activity (β contrast averaged across voxels) of the ROIs. 

• Mediation Analyses 

Mediation analyses were performed to evaluate the relationships between contrast “reward – punishment” activity (left SMG β), drinking PC1 and fear-somatic score (see Results). Please note the results of mediation analyses did not imply causality but served to clarify the inter-relationships of multiple, correlating variables. Details are described in Supplementary Methods.

• Clinical and Cognitive Measures 

As described earlier, Supplementary Table S1 shows the mean ± SD of the drinking measures and PC1 identified of principal component analysis of these measures. Table 1 shows demographics and clinical measures of bingers and non-bingers separately. Age and severity of alcohol use, as indexed by the PC1, showed significant group difference (p’s < 0.001). The results of a two-way (group ´ sex) ANOVA on demographics and clinical measures are shown in Supplementary Table S2.

Table 1: Demographics and clinical measures of the participants. Values are mean ± SD; *two-sample t test (^with age and sex as covariates). Drinking PC1: the first principal component obtained of principal component analyses of all drinking measures; Physi-Aggr: Physical Aggression.

• Behavioral measures of the gambling task

To characterize how individuals reacted to wins and losses, we computed individual RT of trials following loss (post-loss RT) and win (post-win RT) for individuals, and calculated the difference, i.e., post-win RT minus post-loss RT [30]. The results of a two-way (group ´ sex) ANOVA on behavioral measures are shown in Supplementary Table S3. Briefly, bingers vs. non-bingers showed longer “post-win minus post-loss” RT (bingers: 27.1 ± 46.5 ms, non-bingers: 12.5 ± 51.4 ms, t = 2.54, p = 0.012) with age and sex as covariates. In linear regressions with age and sex as covariates, the RT difference was significantly correlated with drinking severity PC1 across all subjects (r = 0.13, p = 0.006) but not in bingers (r = 0.11, p = 0.162) or in non-bingers (r = -0.03, p = 0.630) alone. Thus, more severe drinking is associated with greater difference in post-win and post-loss RT.

• Reward- and punishment-related regional activations

One-sample t test of the contrasts “reward – baseline”, “punishment – baseline” and “reward – punishment” revealed significant clusters for bingers (Figure 1) and non-bingers (Figure 2) at voxel p < 0.05, FWE-corrected. The clusters for “reward – punishment” are summarized in Table 2. The two-way (group ´ sex) ANOVA of beta estimate of clusters are shown in Supplementary Table S4. Notably, although reward and punishment are more salient than baseline, these contrasts revealed regional activities in both directions, with many, including the left supramarginal gyrus (SMG, z=24), showing diminished responses during reward/punishment vs. baseline both in bingers and non-bingers.

Figure 1: Regional responses in bingers. One-sample t-test of contrast (A): “reward - baseline”; (B) “punishment - baseline”; (C) “reward - punishment”. L: left; R: right; v: ventral; s: sulcus. Warm and cool colors show voxels for reward > baseline and reward < baseline, respectively, etc.

Figure 2: Regional responses in non-bingers. One-sample t-test of contrast (A): “reward - baseline”; (B) “punishment - baseline”; (C) “reward - punishment”. L: left; R: right; v: ventral; s: sulcus. Warm and cool colors show voxels for reward > baseline and reward < baseline, respectively, etc. MFG: middle frontal gyrus; mOFG: medial orbitofrontal gyrus; s: sulcus; MOG: middle occipital gyrus; PCC: posterior cingulate cortex; PCu: precuneus; PCG: precentral gyrus; SPC: superior parietal cortex; MCC mid-cingulum.

Table 2: Regional responses of reward vs. punishment in one-sample t-test of bingers and of non-bingers. Brain regions were identified by reference to the Automated Anatomic Labeling or AAL Atlas [44]. v: ventral; a: anterior; L: left; R: right: CS: calcarine sulcus; PCC: posterior cingualate cortex; PCu: precuneus; SPC: superior parietal cortex; MOG: middle occipital gyrus; mOFG: medial orbitofrontal gyrus; PCG: precentral gyrus MFG: middle frontal gyrus; MCC mid-cingulum.

We conducted a two-sample t test to compare binge and non-binge drinkers in the contrast “reward – punishment”. At a threshold of voxel p < 0.005, uncorrected, in combination with cluster p < 0.05, FWE-corrected, two clusters showed higher activities in non-bingers vs. bingers, each involving the left posterior insula and putamen (Ins/Put; X=-42, Y=-16, Z=6, Z score=3.70, volumes=1280 mm³) and the left supramarginal gyrus (SMG; X=-46, Y=-54, Z=36, Z score=3.73, volumes=1000 mm³). These clusters are shown in Figure 3A. We extracted the β’s of the left Ins/Put and SMG clusters for post-hoc analyses (Figures 3B & 3C). In a repeated measures group x block analysis of variance (ANOVA), the left Ins/Put showed significant block main (F = 20.69, p < 0.001) and interaction (F = 30.39, p < 0.001) but not group main (F = 0.53, p = 0.468) effect. Left SMG also showed significant block main (F = 6.84, p = 0.009) and interaction (F = 18.90, p < 0.001) but not group main (F = 0.27, p = 0.603) effect. In additional planned comparisons with paired t tests, the left Ins/Put β’s showed significant difference between reward and punishment blocks in non-bingers (t = 5.82, p < 0.001) but not in bingers (t = -0.86, p = 0.392). The left SMG β’s too showed significant difference between blocks in non-bingers (t = 5.47, p < 0.001) but not in bingers (t = -0.30, p = 0.764).

Figure 3: (A) Two-sample t-test of bingers vs. non-bingers in the contrast of “reward - punishment” for the whole-brain; no significant findings for the reverse contrast; (B) Bar plot of mean ± SE of “reward - baseline” and “punishment - baseline” of left insula/putamen (Ins/Put) for non-bingers and bingers; and (C) Bar plot of mean ± SE of “reward - baseline” and “punishment - baseline” of left SMG for non-bingers and bingers. ** p < 0.001, paired-t tests.

• Correlation of regional activations with behavioral metrics 

The difference in post-win vs. post-loss RT (post-win RT minus post-loss RT) was significantly and negatively correlated with left SMG β (r = -0.11, p = 0.024) and Ins/Put β (r = -0.09, p = 0.043) across all subjects but not in bingers (both p’s > 0.190) or in non-bingers (both p’s > 0.070) alone. Thus, the higher these regional activities differentiating reward and punishment, the lower the difference between post-win and post-loss RT was across all subjects.

• Correlation of regional activations with clinical metrics and mediation analyses

With age and sex as covariates, the left SMG β showed a significant correlation with fear-somatic arousal score across all subjects (r= -0.10, p=0.026) and in bingers (r= -0.21, p=0.006) but not in non-bingers (r= -0.01, p=0.913). Further, slope tests confirmed the group difference (Z=-2.11, p=0.0349). The scatter plots and linear regressions are shown in Figure 4A. The left SMG β did not show a significant correlation with any other negative affect measures (all p’s > 0.05). None of the negative affect measures showed correlation with the β’s of Ins/Put. Fear-somatic arousal score was correlated with PC1 across all subjects (r= 0.14, p=0.003) and in bingers (r= 0.15, p=0.044), but not non-bingers (r= 0.04, p=0.515), although bingers and non-bingers did not differ significantly in the slope of the regressions (Z=1.18, p=0.238, Figure 4B). The left SMG β was significantly correlated with PC1 across all subjects (r= -0.19, p < 0.001) but not in bingers (r= -0.05, p=0.495) or non-bingers (r= 0.01, p=0.819) alone. The left Ins/Put β was significantly correlated with PC1 across all subjects (r= -0.22, p < 0.001) but not in bingers (r= -0.03, p=0.741) or in non-bingers (r= 0.01, p=0.852) alone, either. 

Thus, we have a three-way correlation, pairwise, amongst left SMG β, PC1, and fear-somatic score across all subjects. We performed a mediation analysis to determine the inter-relationships amongst these metrics. The results showed that PC1 completely mediated the relationship between left SMG β and fear-somatic arousal score bidirectionally (Figure 4C). The statistics of all six models are shown in Supplementary Table S5.

Figure 4: (A) The linear correlation between BMI and left SMG beta of “reward-punishment”; and (B) linear correlation between fear somatic score and left SMG beta of “reward-punishment”, please note age and sex have been controlled in beta value. Solid red star: significant correlation. p < 0.01, hollow red star: no significant correlation. (C) Mediation models between left SMG beta, PC1 and fear somatic score.

Behaviorally, bingers as compared to non-bingers showed significantly faster RT following loss than win trials, i.e., a larger “post-win minus post-loss RT,” likely reflecting bingers’ negative urgency. This finding thus contradicts the hypothesis that binge drinking is associated with diminished differentiation of reward and punishment. Bingers as compared to non-bingers showed lower responses differentiating reward and punishment in a cluster that encompassed the left insula and putamen (Ins/Put) and another cluster of the left supramarginal gyrus (SMG). The β’s of the SMG were negatively correlated with fear-somatic scores in binge but not non-binge drinkers, and the group differences in the correlations were confirmed by slope tests. Further, mediation analyses showed drinking severity as a factor inter-linking diminished SMG β and fear-somatic score. Together, the findings highlight the effects of alcohol misuse on blunting the neural responses differentiating reward and punishment and how they may relate to other potentially comorbid, clinical characteristics. We discussed the main findings below. 

• Alcohol misuse and reduced activities differentiating reward and punishment 

Bingers relative to non-bingers showed reduced activities in the left posterior Ins/Put and left SMG during reward relative to punishment blocks. In contrast to the extant studies that have focused on reward or punishment [19-21], these regional activities were specific to differences between reward and punishment processing. Further, the higher these regional activities differentiating reward and punishment, the lower the difference between post-win and post-loss RT was across all subjects. Faster post-loss as compared to post-win RT may reflect negative urgency – a tendency of impulsive responses following negative events. Thus, these regional activities may mitigate negative urgency. 

The posterior insula is known to be involved in processing somatosensory and interoceptive signals. Both reward and punishment are salient, evoking physiological arousal and subjective “visceral” experiences during decision making [45-48]. The posterior putamen is most studied as a motor structure [49,50], and motor action too is associated with heightened physiological arousal [51-53]. The SMG is involved in saliency processing as during the detection of an odd-ball stimulus [54,55]. For instance, haptic mismatch – a saliency signal – for shape or texture led to higher activation in the left SMG and connectivity of the SMG with primary somatosensory cortex during an object manipulation task [56]. On the other hand, the left SMG is involved in the suppression of somatosensory saliency during reaching movements of the limb [57]. Thus, the left SMG may show higher activation in response to environmental saliency and to suppression of saliency during internally generated actions, perhaps via an efference copy of the motor signals. These flexible roles of the left SMG contribute to modulation of saliency as needed for behavioral control. 

The current findings extend this literature by showing that the left posterior Ins/Put and left SMG are both involved in differentiating the saliency of reward and punishment in relation to alcohol misuse. Specifically, these regional activities are diminished in association with the severity of alcohol misuse and manifestation of negative urgency following loss as compared to following win trials. In addition to saliency processing, the left SMG is also involved in action planning, re-programming, and decision making [58-60]. These findings suggest the left SMG as a higher-order cognitive brain region and its potential role in valence-related decision making. Here, in the card guessing task, there were more wins than losses in the reward and vice versa in the punishment blocks, and, as the outcomes have been dictated, participants would have no reason to experience higher saliency during wins vs. losses or vice versa. Across subjects, higher left SMG activity may serve to suppress the difference in saliency signaling between reward and punishment blocks, and this activity is compromised in individuals with more severe drinking. To the best of our knowledge, the current finding is the first to implicate altered saliency processing of reward relative to punishment in relation to problem drinking, adding to the literature of neural markers of alcohol misuse (e.g., Morris et al., 2019; Zhao et al., 2021; Zhu et al., 2022; Zhu et al., 2024). 

• Supramarginal gyrus and fear-related activities 

These findings also add to the literature of altered saliency processing in many comorbidities of alcohol misuse, including depression [61], mal-adaptive stress reactivity [62-65], and chronic pain [66-69]. The left SMG, amongst other cortical and subcortical regions, responds to fear-conditioning stimuli, as shown in a meta-analysis [70], and to physiological arousal during fear learning [71]. Nonverbal human sounds representing fears relative to those of neutral emotions evoked higher activity in bilateral SMG and right superior temporal gyrus in neurotypical individuals [72]. When imagining daily activities as illustrated in pictures, patients with chronic pain relative to control participants demonstrated higher activity in the inferior parietal cortex, including the left SMG, and other cortical regions [73]. These earlier findings together implicate the left SMG in processing fear-related information. However, it remains unclear why valence-differentiating activities of the SMG were negatively related to fear-somatic arousal scores across individuals. Perhaps lower activities differentiating reward and punishment suggest indiscriminate responses to saliency, including those generated during e.g., general, anticipatory anxiety?. 

• Broader considerations 

The importance of differentiating reward and punishment sensitivity have been illustrated in other contexts. Neural responses to win and loss may distinguish individual personality traits [74,75]. These neural responses may also help in disambiguating the effects of psychoactive substances. For instance, brain reward signals were greater in individuals who reported higher euphoric response to d-amphetamine and in those showing greater motivation (i.e., wanting) to drink more after a single dose of alcohol [76-78]. However, methamphetamine appeared to enhance Ventral Striatal (VS) activation to monetary loss but did not alter VS responses to win, thus suppressing responses to win vs. loss during the MIDT in healthy young adults [79]. In an earlier study of the MIDT, Knutson et al. also found that dextroamphetamine increased VS activation during anticipation of losses but not gains, and this effect was associated with higher positive arousal ratings for loss incentive cues [79]. Thus, the roles of the VS in individual differences in drug and alcohol craving appear best clarified by contrasting its responses to reward vs. punishment, rather than to reward or punishment alone. 

Distinguishing reward and punishment activities may also facilitate biotyping of alcohol cue reactivities in drinkers. In a recent work we employed meta-analyses of drug cue-elicited reactivity and win and loss processing in the MIDT to identify distinct neural correlates of appetitive and aversive responses to drug cues [80]. We then characterized the appetitive and aversive cue responses in alcohol drinkers performing a cue craving task during fMRI and observed that individuals varied in appetitive and aversive cue responses. In clinical characteristics, the avoidance subtype showed higher sensitivity to punishment, and, in contrast, the approach subtype showed higher levels of sensation seeking and alcohol expectancy for social and physical pressure. The findings highlighted distinct neural underpinnings of appetitive and aversive components of cue-elicited reactivity and provided evidence for potential subtypes of alcohol misuse.

In conclusion, binge drinking is associated with reduced neural responses to reward vs. punishment. Higher left SMG activity may help in suppressing differences in saliency signaling of reward vs. punishment, which is to no avail in a card guessing task. Alcohol misuse compromises left SMG activities and disposes individuals to negative urgency, as reflected in faster post-loss as compared to post-win RT, and greater fear-somatic concerns in a neurotypical population.

We considered a few limitations of the study. First, more men than women were binge drinkers; although we have considered sex as a covariate in data analyses, we could not entirely rule out the effects of sex on the current findings. As importantly, men and women are known to show differences in the pathophysiology and behavioral manifestation of alcohol misuse (Ide et al., 2018; Li et al., 2020c; Wilsnack et al., 2018). More work is needed to explore sex differences in the impact of alcohol misuse on valence-differentiating behaviors and their neural bases. Second, neural responses to monetary win and loss may be influenced by many contextual factors, including uncertainty [81] and subjective emotional states [82]. Therefore, the current findings need to be considered as specific to the card guess paradigm. Further, participants were engaged in the gambling task with a fixed order of winning and losing blocks. It remains unclear how this would impact the findings. Third, some but not all of the “significant” findings on somatic fear score were significant with correction for multiple comparisons. Thus, these as well as the lack of significant correlations for other measures of negative affect need to be replicated. Further, the current findings on the posterior insula are specific to a contrast between reward and punishment and may be considered along with many previous reports of the roles of anterior insula dysfunction in alcohol misuse (see Flook et al., 2021; Sommer et al., 2022 for a review). Fourth, while “over-differentiating” wins and losses in behavioral response, bingers relative to non-bingers demonstrated less left SMG activities differentiating reward and punishment. Thus, how the left SMG partakes in valence-differentiating behavior and whether other regional activities and connectivities may be involved remain to be investigated. Finally, limited neural differentiation of reward and punishment may impact instrumental learning and other motivated behavior, compromising motivation-focused cognitive behavioral therapy in people with substance misuse. Although we demonstrated the inter-related influences of drinking severity and diminished left SMG activity on negative urgency, more studies are warranted to investigate whether and how diminished valence differentiation may impact learning and decision making.

Guangfei Li: Writing – review & editing, Writing – original draft, Visualization, Software, Methodology, Formal analysis, Data curation, Conceptualization. Yun Dong: Writing – review & editing, Writing – original draft. Yu Chen: Writing – review & editing, Formal analysis, Data curation. Shanmei Wang: Writing – review & editing, Writing – original draft. Yanan Su: Writing – review & editing. Shufang Li: Writing – review & editing. Ziyan Sun: Writing – review & editing. Thang M. Le: Writing – review & editing, Writing – original draft, Funding acquisition, Conceptualization. Chiang-Shan R. Li: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding acquisition, Conceptualization.

1. Holmes NM, Marchand AR, Coutureau E (2010) Pavlovian to instrumental transfer: a neurobehavioural perspective. Neuroscience and biobehavioral reviews 34: 1277-1295.
2. Pessiglione M, Petrovic P, Daunizeau J, Palminteri S, Dolan RJ, et al. (2008) Subliminal instrumental conditioning demonstrated in the human brain. Neuron 59: 561-567.
3. Watson P, Wiers RW, Hommel B, Ridderinkhof KR, de Wit S (2016) An associative account of how the obesogenic environment biases adolescents' food choices. Appetite 96: 560-571.
4. Lo Gerfo E, De Angelis J, Vergallito A, Bossi F, Romero Lauro LJ, et al. (2018) Can Monetary Reward Modulate Social Attention? Frontiers in psychology 9: 2213.
5. Hawes DR, DeYoung CG, Gray JR, Rustichini A (2014) Intelligence moderates neural responses to monetary reward and punishment. Journal of neurophysiology 111: 1823-1832.
6. Dong G, Lin X, Hu Y, Lu Q (2013) Brain activity in advantageous and disadvantageous situations: implications for reward/punishment sensitivity in different situations. PloS one 8: e80232.
7. Delgado MR, Nystrom LE, Fissell C, Noll DC, Fiez JA (2000) Tracking the hemodynamic responses to reward and punishment in the striatum. Journal of neurophysiology 84: 3072-3077.
8. Chen Y, Chaudhary S, Li CR (2022) Shared and distinct neural activity during anticipation and outcome of win and loss: A meta-analysis of the monetary incentive delay task. Neuroimage 264: 119764.
9. Dugré JR, Dumais A, Bitar N, Potvin S (2018) Loss anticipation and outcome during the Monetary Incentive Delay Task: a neuroimaging systematic review and meta-analysis. PeerJ 6: e4749.
10. Wang C, Bao C, Gao J, Gu Y, Dong XW (2020) Pain modulates neural responses to reward in the medial prefrontal cortex. Human brain mapping 41(5): 1372-1381.
11. Pinger M, Thome J, Halli P, Sommer WH, Koppe G, et al. (2022) Comparing Discounting of Potentially Real Rewards and Losses by Means of Functional Magnetic Resonance Imaging. Frontiers in systems neuroscience 16: 867202.
12. Arendt T (1994) Impairment in memory function and neurodegenerative changes in the cholinergic basal forebrain system induced by chronic intake of ethanol. Journal of neural transmission Supplementum 44: 173-187.
13. Taffe MA, Kotzebue RW, Crean RD, Crawford EF, Edwards S, et al. (2010) Long-lasting reduction in hippocampal neurogenesis by alcohol consumption in adolescent nonhuman primates. Proceedings of the National Academy of Sciences of the United States of America 107: 11104-11109.
14. Li G, Le TM, Wang W, Zhornitsky S, Chen Y, et al. (2021) Perceived stress, self-efficacy, and the cerebral morphometric markers in binge-drinking young adults. Neuroimage Clin 32: 102866.
15. Le TM, Potvin S, Zhornitsky S, Li CR (2021) Distinct patterns of prefrontal cortical disengagement during inhibitory control in addiction: A meta-analysis based on population characteristics. Neuroscience and biobehavioral reviews 127: 255-269.
16. Ide JS, Zhornitsky S, Chao HH, Zhang S, Hu S, et al. (2018) Thalamic Cortical Error-Related Responses in Adult Social Drinkers: Sex Differences and Problem Alcohol Use. Biological psychiatry Cognitive neuroscience and neuroimaging 3: 868-877.
17. Le TM, Zhornitsky S, Wang W, Ide J, Zhang S, et al. (2019) Posterior Cingulate Cortical Response to Active Avoidance Mediates the Relationship between Punishment Sensitivity and Problem Drinking. The Journal of neuroscience: the official journal of the Society for Neuroscience 39: 6354-6364.
18. Lees B, Mewton L, Stapinski LA, Squeglia LM, Rae CD, et al. (2019) Neurobiological and Cognitive Profile of Young Binge Drinkers: a Systematic Review and Meta-Analysis. Neuropsychology review 29: 357-385.
19. Jonker NC, Ostafin BD, Glashouwer KA, van Hemel-Ruiter ME, de Jong PJ (2014) Reward and punishment sensitivity and alcohol use: the moderating role of executive control. Addictive behaviors 39: 945-948.
20. Noël X (2014) Why adolescents are at risk of misusing alcohol and gambling. Alcohol and alcoholism (Oxford, Oxfordshire) 49: 165-172.
21. Lodha J, Brocato E, Wolstenholme JT (2022) Areas of Convergence and Divergence in Adolescent Social Isolation and Binge Drinking: A Review. Frontiers in behavioral neuroscience 16: 859239.
22. Logge WB, Morley KC, Haber PS, Baillie AJ (2023) Impaired Decision-Making and Skin Conductance Responses Are Associated with Reward and Punishment Sensitivity in Individuals with Severe Alcohol Use Disorder. Neuropsychobiology 82: 117-129.
23. Paelecke-Habermann Y, Paelecke M, Mauth J, Tschisgale J, Lindenmeyer J, et al. (2019) A comparison of implicit and explicit reward learning in low risk alcohol users versus people who binge drink and people with alcohol dependence. Addictive behaviors reports 9: 100178.
24. Bagci B, Düsmez S, Zorlu N, Bahtiyar G, Isikli S, et al. (2022) Computational analysis of probabilistic reversal learning deficits in male subjects with alcohol use disorder. Frontiers in psychiatry 13: 960238.
25. Zeng J, You L, Sheng H, Luo Y, Yang X (2023) The differential neural substrates for reward choice under gain-loss contexts and risk in alcohol use disorder: Evidence from a voxel-based meta-analysis. Drug and alcohol dependence 248: 109912.
26. Voon V, Grodin E, Mandali A, Morris L, Doñamayor N, et al. (2020) Addictions NeuroImaging Assessment (ANIA): Towards an integrative framework for alcohol use disorder. Neuroscience and biobehavioral reviews 113: 492-506.
27. Tanabe J, Regner M, Sakai J, Martinez D, Gowin J (2019) Neuroimaging reward, craving, learning, and cognitive control in substance use disorders: review and implications for treatment. The British journal of radiology 92: 20180942.
28. Galandra C, Basso G, Cappa S, Canessa N (2018) The alcoholic brain: neural bases of impaired reward-based decision-making in alcohol use disorders. Neurological Sciences: official journal of the Italian Neurolog Soc Italian Soc Clin Neurophysiol 39: 423-435.
29. Crane NA, Gorka SM, Weafer J, Langenecker SA, de Wit H, et al. (2017) Preliminary Evidence for Disrupted Nucleus Accumbens Reactivity and Connectivity to Reward in Binge Drinkers. Alcohol and alcoholism (Oxford, Oxfordshire) 52: 647-654.
30. Li G, Chen Y, Chaudhary S, Tang X, Li CR (2022) Loss and Frontal Striatal Reactivities Characterize Alcohol Use Severity and Rule-Breaking Behavior in Young Adult Drinkers. Biological psychiatry Cognitive neuroscience and neuroimaging 7: 1007-1016.
31. Musial MPM, Beck A, Rosenthal A, Charlet K, Bach P, et al. (2023) Reward Processing in Alcohol-Dependent Patients and First-Degree Relatives: Functional Brain Activity During Anticipation of Monetary Gains and Losses. Biol Psychiatry 93: 546-57.
32. Aloi J, Blair KS, Crum KI, Bashford-Largo J, Zhang R, et al. (2020) Alcohol Use Disorder, But Not Cannabis Use Disorder, Symptomatology in Adolescents Is Associated With Reduced Differential Responsiveness to Reward Versus Punishment Feedback During Instrumental Learning. Biological psychiatry Cognitive neuroscience and neuroimaging 5: 610-618.
33. Katz BA, Matanky K, Aviram G, Yovel I (2020) Reinforcement sensitivity, depression and anxiety: A meta-analysis and meta-analytic structural equation model. Clinical Psychology Review 77: 101842.
34. Li Y, Cheng X, Li Y, Sui X (2021) Neural Responses to Reward and Punishment Stimuli in Depressed Status Individuals and Their Effects on Cognitive Activities. Frontiers in psychology 12: 808341.
35. Van Essen DC, Ugurbil K, Auerbach E, Barch D, Behrens TEJ, et al. (2012) The Human Connectome Project: A data acquisition perspective. NeuroImage 62: 2222-31.
36. Li G, Zhang S, Le TM, Tang X, Li C-SR (2020) Neural Responses to Reward in a Gambling Task: Sex Differences and Individual Variation in Reward-Driven Impulsivity. Cereb Cortex Commun 1: tgaa025-tgaa25.
37. Li G, Zhang S, Le TM, Tang X, Li C-SR (2020) Neural responses to negative facial emotions: Sex differences in the correlates of individual anger and fear traits. Neuroimage 221: 117171-71.
38. Li G, Chen Y, Wang W, Dhingra I, Zhornitsky S, et al. (2020) Sex Differences in Neural Responses to the Perception of Social Interactions. Front Hum Neurosci 14: 565132-32.
39. Wechsler, Davenport, Dowdall, Moeykens, Castillo (1994) Health and behavioral consequences of binge drinking in college. A national survey of students at 140 campuses. Jama J Am Med Ass.
40. Gowin JL, Manza P, Ramchandani VA, Volkow ND (2020) Neuropsychosocial markers of binge drinking in young adults. Mol Psychiatry.
41. Barch DM, Burgess GC, Harms MP, Petersen SE, Schlaggar BL, et al. (2013) Function in the human connectome: task-fMRI and individual differences in behavior. Neuroimage 80: 169-89.
42. Zhornitsky S, Zhang S, Ide JS, Chao HH, Wang W, et al. (2019) Alcohol Expectancy and Cerebral Responses to Cue-Elicited Craving in Adult Nondependent Drinkers. Biol Psychiatry Cogn Neurosci Neuroimaging 4: 493-504.
43. Zhang S, Zhornitsky S, Le TM, Li CR (2019) Hypothalamic Responses to Cocaine and Food Cues in Individuals with Cocaine Dependence. Int J Neuropsychopharmacol 22: 754-64.
44. Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, et al. (2002) Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage 15: 273-89.
45. North NT, O'Carroll RE (2001) Decision making in patients with spinal cord damage: afferent feedback and the somatic marker hypothesis. Neuropsychologia 39: 521-4.
46. Bechara A n(2001) Neurobiology of decision-making: risk and reward. Seminars in clinical neuropsychiatry 6: 205-16.
47. Crone EA, Somsen RJ, Van Beek B, Van Der Molen MW (2004) Heart rate and skin conductance analysis of antecendents and consequences of decision making. Psychophysiology 41: 531-40.
48. Maia TV, McClelland JL (2004) A reexamination of the evidence for the somatic marker hypothesis: what participants really know in the Iowa gambling task. Proceedings of the National Academy of Sciences of the USA 101: 16075-80.
49. Mao Q, Wang X, Chen B, Fan L, Wang S, et al. (2020) KTN1 Variants Underlying Putamen Gray Matter Volumes and Parkinson's Disease. Frontiers in neuroscience 14: 651.
50. Luo X, Mao Q, Shi J, Wang X, Li CR (2019) Putamen gray matter volumes in neuropsychiatric and neurodegenerative disorders. World journal of psychiatry and mental health research 3: 1020.
51. Kazmierczak M, Nicola SM (2022) The Arousal-motor Hypothesis of Dopamine Function: Evidence that Dopamine Facilitates Reward Seeking in Part by Maintaining Arousal. Neuroscience 499: 64-103.
52. Liu D, Dan Y (2019) A Motor Theory of Sleep-Wake Control: Arousal-Action Circuit. Annual review of neuroscience 42: 27-46.
53. Cao G, Liu P (2021) Arousal modulates the motor interference effect stimulated by pictures of threatening animals. Peer J 9: e10876.
54. Torii T, Sato A, Iwahashi M, Itoh Y, Iramina K (2012) Time-dependent effects of low-frequency repetitive transcranial magnetic stimulation of the supramarginal gyrus. Annual International Conference of the IEEE Engineering in Medicine and Biology Society IEEE Engineering in Medicine and Biology Society Annual International Conference 2012: 3372-5.
55. Schiff S, Bardi L, Basso D, Mapelli D (2011) Timing spatial conflict within the parietal cortex: a TMS study. Journal of cognitive neuroscience 23: 3998-4007.
56. Gurtubay-Antolin A, León-Cabrera P, Rodríguez-Fornells A. Neural Evidence of Hierarchical Cognitive Control during Haptic Processing: An fMRI Study.eNeuro 5: 6.
57. Arikan BE, Voudouris D, Voudouri-Gertz H, Sommer J, Fiehler K(2021) Reach-relevant somatosensory signals modulate activity in the tactile suppression network. NeuroImage 236: 118000.
58. Randerath J, Valyear KF, Philip BA, Frey SH (2017) Contributions of the parietal cortex to increased efficiency of planning-based action selection. Neuropsychologia 105: 135-43.
59. Hartwigsen G, Siebner HR (2015) Joint Contribution of Left Dorsal Premotor Cortex and Supramarginal Gyrus to Rapid Action Reprogramming. Brain stimulation 8: 945-52.
60. Hartwigsen G, Bestmann S, Ward NS, Woerbel S, Mastroeni C, et al. (2012) Left dorsal premotor cortex and supramarginal gyrus complement each other during rapid action reprogramming. The Journal of neuroscience : the official journal of the Society for Neuroscience 32: 16162-71a.
61. Gaffrey MS, Barch DM, Bogdan R, Farris K, Petersen SE, et al. (2018) Amygdala Reward Reactivity Mediates the Association Between Preschool Stress Response and Depression Severity. Biol Psychiatry. 83: 128-36.
62. Koob GF, Le Moal M (2008) Review. Neurobiological mechanisms for opponent motivational processes in addiction. Philosophical transactions of the Royal Society of London Series B, Biological sciences. 363: 3113-23.
63. Pleil KE, Rinker JA, Lowery-Gionta EG, Mazzone CM, McCall NM, et al. (2015) NPY signaling inhibits extended amygdala CRF neurons to suppress binge alcohol drinking. Nature neuroscience. 18: 545-52.
64. Rinker JA, Marshall SA, Mazzone CM, Lowery-Gionta EG, Gulati V, et al. (2017) Extended Amygdala to Ventral Tegmental Area Corticotropin-Releasing Factor Circuit Controls Binge Ethanol Intake. Biol Psychiatry 81: 930-40.
65. Aroni S, Marino RAM, Girven KS, Irving JM, Cheer JF, (2021) Repeated binge ethanol drinking enhances electrical activity of central amygdala corticotropin releasing factor neurons in vivo. Neuropharmacology 189: 108527.
66. Hu B, Yu Y, Dai YJ, Feng JH, Yan LF, et al. (2019) Multi-modal MRI Reveals the Neurovascular Coupling Dysfunction in Chronic Migraine. Neuroscience 419: 72-82.
67. Baumbach P, Meißner W, Reichenbach JR, Gussew A (2022) Functional connectivity and neurotransmitter impairments of the salience brain network in chronic low back pain patients: a combined resting-state functional magnetic resonance imaging and 1 H-MRS study. Pain 163: 2337-47.
68. Xu H, Seminowicz DA, Krimmel SR, Zhang M, Gao L, et al. (2022) Altered Structural and Functional Connectivity of Salience Network in Patients with Classic Trigeminal Neuralgia. The journal of pain. 23: 1389-99.
69. Ferraro S, Klugah-Brown B, Tench CR, Yao S, Nigri A, et al. (2022) Dysregulated anterior insula reactivity as robust functional biomarker for chronic pain-Meta-analytic evidence from neuroimaging studies. Human brain mapping 43: 998-1010.
70. Biggs EE, Timmers I, Meulders A, Vlaeyen JWS, Goebel R, et al. (2020) The neural correlates of pain-related fear: A meta-analysis comparing fear conditioning studies using painful and non-painful stimuli. Neuroscience and biobehavioral reviews 119: 52-65.
71. Leuchs L, Schneider M, Czisch M, Spoormaker VI (2017) Neural correlates of pupil dilation during human fear learning. NeuroImage 147: 186-97.
72. Köchel A, Schöngassner F, Schienle A (2013) Cortical activation during auditory elicitation of fear and disgust: a near-infrared spectroscopy (NIRS) study. Neuroscience letters 549: 197-200.
73. Taylor AM, Harris AD, Varnava A, Phillips R, Taylor JO, et al. (2015) A Functional Magnetic Resonance Imaging Study to Investigate the Utility of a Picture Imagination Task in Investigating Neural Responses in Patients with Chronic Musculoskeletal Pain to Daily Physical Activity Photographs. PloS one 10: e0141133.
74. Sun S, Wang Y, Bai X (2023) Outcome Evaluation in Social Comparison: When You Deviate from Others. Brain sciences 13: 925.
75. Li G, Li Y, Zhang Z, Chen Y, Li B, et al. (2023) Sex differences in externalizing and internalizing traits and ventral striatal responses to monetary loss. J Psychiatr Res 162: 11-20.
76. Crane NA, Gorka SM, Weafer J, Langenecker SA, de Wit H, et al. (2018) Neural activation to monetary reward is associated with amphetamine reward sensitivity. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 43: 1738-44.
77. Langenecker SA, Kling LR, Crane NA, Gorka SM, Nusslock R, et al. (2020) Anticipation of monetary reward in amygdala, insula, caudate are predictors of pleasure sensitivity to d-Amphetamine administration. Drug and alcohol dependence 206: 107725.
78. Radoman M, Crane NA, Gorka SM, Weafer J, Langenecker SA, et al. (2021) Striatal activation to monetary reward is associated with alcohol reward sensitivity. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 46: 343-50.
79. Crane NA, Molla H, de Wit H (2023) Methamphetamine alters nucleus accumbens neural activation to monetary loss in healthy young adults. Psychopharmacology 240: 1891-900.
80. Chen Y, Li CR (2023) Appetitive and aversive cue reactivities differentiate neural subtypes of alcohol drinkers. Addiction neuroscience 7: 100089.
81. Zhang K, Li Y, Chen X, Xu Y, Zhang C, et al. (2022) Money priming enhances sensitivity to the outcome feedback of decision-making under uncertainty: Evidence from an ERP study. Neuropsychologia 176: 108390.
82. Gaillard C, Guillod M, Ernst M, Torrisi S, Federspiel A, et al. (2019) Striatal responsiveness to reward under threat-of-shock and working memory load: A preliminary study. Brain and behavior 9: e01397.

Dong et al., Reduction in Valence-Differentiating Neural Activities in Binge Drinking and its Comorbidities: an Exploratory Study of the Human Connectome Project Data

 1. Supplementary Methods

 A. Six items of the fear-somatic subscale of the NIH Emotion Toolbox

1. I was short of breath.
2. I felt nauseous.
3. I felt dizzy or lightheaded.
4. My muscles twitched or trembled.
5. My heart was racing or pounding.
6. My muscles were tense or sore.

B. Imaging protocol, gambling task and data preprocessing 

MRI was done using a customized 3 T Siemens Connectome Skyra with a standard 32-channel Siemens receiver head coil and a body transmission coil. T1-weighted high-resolution structural images were acquired using a 3D MPRAGE sequence with 0.7 mm isotropic resolution (FOV = 224 × 224 mm, matrix = 320 × 320, 256 sagittal slices, TR = 2400 ms, TE = 2.14 ms, TI = 1000 ms, FA = 8°) and used to register functional MRI data to a standard brain space. FMRI data were collected using gradient-echo Echo-Planar Imaging (EPI) with 2.0 mm isotropic resolution (FOV = 208 × 180 mm, matrix = 104 × 90, 72 slices, TR = 720 ms, TE = 33.1 ms, FA = 52°, multi-band factor = 8).

Participants completed two runs of a gambling task each with 4 blocks (~3 m and 12 s each run) – 2 of punishment and 2 of reward – in a fixed order (first run: reward – punishment – punishment – reward; and second run: punishment – reward – punishment – reward) with a fixation period (15 s) between blocks. The participants guessed whether the number of a mystery card (represented by a ‘?’ and ranging from 1 to 9) was larger or smaller than 5 by pressing a corresponding button [1]. The feedbacks comprised a green up-pointing arrow for correct guess and $1 win, a red down-pointing arrow for $0.5 loss; or a gray double-headed arrow for a wash (mystery card number = 5). The mystery number was controlled by the program and shown for 1.5 s, followed by the feedback for 1.0 s. There was a 1.0 s inter-trial interval with a “+” shown on the screen. Each block contained 8 trials. In reward blocks, 6 win trials were pseudo-randomly interleaved with either 1 neutral and 1 loss trial, 2 neutral trials, or 2 loss trials. In punishment blocks, 6 loss trials were interleaved with either 1 neutral and 1 win trial, 2 neutral trials, or 2 win trials. Thus, the amount of money won was the same across subjects.

BOLD data were analyzed with Statistical Parametric Mapping (SPM8, Welcome Department of Imaging Neuroscience, University College London, U.K.), following our published routines [4-6]. Images of each individual subject were first realigned (motion corrected). A mean functional image volume was constructed for each subject per run from the realigned image volumes. These mean images were co-registered with the high-resolution structural MPRAGE image and then segmented for normalization with affine registration followed by nonlinear transformation. The normalization parameters determined for the structural volume were then applied to the corresponding functional image volumes for each subject. The voxel is of 2x2x2 mm³ after spatial normalization. Finally, the images were smoothed with a Gaussian kernel of 4 mm at Full Width at Half Maximum.

C.  The GLM and 2^nd-level analyses 

Briefly, a statistical analytical block design was constructed for each individual subject, using a general linear model (GLM) by convolving the canonical Hemodynamic Response Function (HRF) with a boxcar function in SPM. Realignment parameters in all six dimensions were entered in the model as covariates. We constructed for each individual subject the statistical contrast “reward vs. baseline” and “punishment vs. baseline” and “reward vs. punishment”, with baseline = 15-s fixation period between blocks. In group analyses, we conducted a one-sample t test of the contrasts. We evaluated the results at voxel p < 0.05, corrected for family-Wise Error (FWE) of multiple comparisons, on the basis of Gaussian random field theory, as implemented in SPM. We identified brain regions using the Data Processing & Analysis of Brain Imaging toolbox (DPABI) [2] and an atlas (Duvernoy, 2009) [3], if the peak was not identified by the DPABI.

D. Mediation analysis 

In a mediation analysis, the relation between the independent variable X and dependent variable Y, i.e. X→Y, is tested to see if the relation is significantly mediated by a variable M. The mediation test is performed by employing three regression equations:where a represents X→M, b represents M→Y (controlling for X), c’ represents X→Y (controlling for M), and c represents X→Y. The constants i1, i2, i3 are the intercepts, and e1, e2, e3 are the residual errors. In the literature, a, b, c and c’ were referred as path coefficients or simply paths, and we followed this notation. Variable M is a mediator of the correlation X→Y if (c –c’), which is mathematically equivalent to the product of the paths a*b, is significantly different from zero (MacKinnon et al., 2007). If the product a*b and the paths a and b are significant, one concludes that X→Y is mediated by M. In addition, if path c’ is not significant, there is no direct connection from X to Y and that X→Y is completely mediated by M. Note that path b is the relation between Y and M, controlling for X, and should not be confused with the correlation coefficient between Y and M.

Supplementary Table S1: ANOVA of alcohol drinking measures with age as a covariate. Note: Age/years, RTREW (RT of reward blocks, ms), RTPUN (RT of punish blocks, ms), Dp_Sx (Number of DSM4 Alcohol Dependence Criteria Endorsed), Ab_Dx (DSM4 Alcohol Abuse Criteria Met), Ab_Sx (DSM4 Alcohol Abuse number of symptoms), Dp_Dx (DSM4 Alcohol Dependence Criteria Met), Daily drinks (Drinks per drinking day in past 12 months), Frq (Frequency of any alcohol use in past 12 months), Frq_5plus (Frequency of drinking 5+ drinks in past 12 months), Frq_Drk (Frequency drunk in past 12 months), Max_Drk (Max drinks in a single day in past 12 months), Age_Use (Age at first alcohol use), Hvy_Daily (Drinks per day in heaviest 12-month  period), Hvy_Frq (Frequency of any alcohol use, heaviest 12-month period), Hvy_5plus (Frequency of drinking 5+ drinks,  heaviest 12-month period), Hvy_Drk (Frequency drunk in heaviest 12-month period), Hvy_Max (Lifetime max drinks in single day), PC1 (Severity of alcohol use as quantified by the weight of the first principal component (PC1) of PCA of all 15 drinking measures). Note that Frq, Frq_5plus, Frq_Drk, Hvy_Frq, Hvy_5plus and Hvy_Drk were flipped. Values are mean ± SD. Note that some entries are negative in value because the original scales needed to be reversed in scoring so that across metrics, a higher value reflects more severe alcohol misuse, to be consistent. 

Supplementary Table S2:  ANOVA of demographics and clinical measures with age as a covariate. Note: Age/years, Education/years. InstruSupp: instrumental support. Values are mean ± SD. AA: Anger Affect: AH: Anger Hostility; APH: Anger physical aggressiveness; Mean Purp: Mean Purpose; Perceiv H: Perceived Hostility; Perceiv R: Perceived Rejection; Emot Supp: Emotional Suppression; Instru Supp: Instrumental Support; Perceiv Str: Perceived Stress; Self Effic: Self Efficacy. 

Supplementary Table S3: ANOVA of behavioral measures with age as a covariate. Note: Values are mean ± SD. 

Supplementary Table S4: ANOVA of beta estimate of clusters identified by bingers and non-bingers with age as a covariate. Note: Values are mean ± SD. v: ventral; a: anterior; L: left; R: right: CS: calcarine sulcus; PCC: posterior cingualate cortex; PCu: precuneus; SPC: superior parietal cortex; MOG: middle occipital gyrus; mOFG: medial orbitofrontal gyrus; PCG: precentral gyrus MFG: middle frontal gyrus; MCC mid-cingulum. 

Supplementary Table S5: Mediation analyses with age and sex as covariates. Note: Significant models are bolded.

1. Barch DM, Burgess GC, Harms MP, Petersen SE, Schlaggar BL, et al. (2013) Function in the human connectome: task-fMRI and individual differences in behavior. NeuroImage 80: 169-189.
2. Yan CG, Wang XD, Zuo XN, Zang YF (2016) DPABI: Data Processing & Analysis for (Resting-State) Brain Imaging. Neuroinformatics 14: 339-351.
3. Duvernoy HM (2009) The Human Brain. Second Edition. Springer-Verlag, Wien/New York AJNR Am J Neuroradiol 30: e75.
4. Wang W, Zhornitsky S, Le TM, Zhang S, Li CSR (2020) Heart Rate Variability, Cue-Evoked Ventromedial Prefrontal Cortical Response, and Problem Alcohol Use in Adult Drinkers. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 5: 619-628.
5. Zhang S, Zhornitsky S, Le TM, Li CR (2019) Hypothalamic Responses to Cocaine and Food Cues in Individuals with Cocaine Dependence. Int J Neuropsychopharmacol 22: 754-764.
6. Zhornitsky S, Zhang S, Ide JS, Chao HH, Wang W, et al. (2019) Alcohol Expectancy and Cerebral Responses to Cue-Elicited Craving in Adult Nondependent Drinkers. Biol Psychiatry Cogn Neurosci Neuroimaging 4: 493-504.