Table 2.
Summary of 28 studies from systematic review, with sample characteristics, measures used, neuroimaging tasks, and findings.
| Authors (Year) |
Sample size, age (SD), sex, ethnicity | Measure of substance use | Timing of substance use | Type of substance | Task used in fMRI | Contrast of interest | Significant brain regions | Summary of findings |
|---|---|---|---|---|---|---|---|---|
| Social Reward Category | ||||||||
| Jarcho et al. (2022) |
N = 33 Mage = 21.88 (4.35) 64 % Female |
Self-report via the Externalizing Spectrum Inventory-brief form (ESI-bf) | Concurrent substance use | Polydrug | Peer Acceptance and Rejection Task | Positive rewarding feedback > Non-rewarding feedback | VS, substantia nigra, crus cerebri | Decreased right VS response to social reward (positive rewarding feedback from a peer) was related to greater substance abuse behavior. |
| Zimmermann et al. (2019) |
N = 47 (n = 23 cannabis) Mage = 23.76 (3.12) 100 % Male |
Self-report | Effects of substance use on neural function | Cannabis | Interpersonal Touch Paradigm | Touch > Close Female > Male |
Striatum, insula | Cannabis users showed less activity in the dorsal striatum with female experimenter touch, while non-users showed more striatal activity during interpersonal touch. |
| Li et al. (2021) |
N = 460 (Human Connectome Project) Mage = 28.75 (3.6) 51.7 % Female |
Self-report, perceived friendship | Effects of substance use on neural function | Alcohol | Social Cognition Task (shapes interacting either socially or randomly) | Social Interaction of abstract shapes > Non-Social | Right posterior insula, mOFC, left ventral precuneus | Binge drinkers, especially female, had less response in the mOFC and precuneus to social interactions of abstract shapes compared to random movement. Greater posterior insula activity was correlated with lower self-reported scores on perceived friendships. |
| Emotional Sensitivity Category | ||||||||
| Aloi et al. (2018) |
N = 82 (n = 47 from a residential treatment facility) Mage = 16.1 (1.32) 62.2 % Male |
Self-report via AUDIT, CUDIT | Concurrent substance use | Polydrug | Affective Stroop Task (Emotional Faces) | Negative Emotion > Neutral Face Positive Emotion > Neutral Face |
Amygdala, precuneus, PCC, iPL | AUDIT scores were positively associated with amygdala response to emotional stimuli (both negative and positive), but negatively with dACC, dlPFC, and precuneus. CUDIT scores were positively related to activity in the PCC, precuneus, and iPL, but not amygdala. |
| Blair et al. (2019) |
N = 87 Mage = 16.48 (1.17) 50.6 % Female |
Self-report via AUDIT, CUDIT | Concurrent substance use | Polydrug | Looming Task (Emotional Faces) | Looming Stimulus > Receding Stimulus Angry Face > Neutral Face |
Rostral medial frontal cortex, left fusiform gyrus, right cerebellum | CUDIT scores were negatively correlated with response to looming stimuli in the rmPFC, cerebellum, and fusiform gyrus. The pattern was similar for the amygdala, but not significant after correcting for multiple comparisons. Threatening stimuli were not related to CUDIT or AUDIT scores. |
| Chaplin et al. (2019) |
N = 66 Mage = 12.59 (0.70) 48.5 % Female, 71.2 % White |
YRBS, Teen Addiction Severity Index, urine screens, breathalyzer |
Concurrent substance use | Polydrug | Emotional Photos (International Affective Picture System) | Negative Emotion > Neutral | Amygdala, ACC, AI | In girls, greater insula response to negative stimuli was associated with more substance use. Boys did not show an association between neural response to emotional stimuli and substance use. |
| Cohen-Gilbert et al. (2022) |
N = 60 Mage = 18.9 (0.4) 55.5 % Female |
AUDIT, Counseling Center Assessment of Psychological Symptoms (C- CAPS), Young Adult Alcohol Consequences Questionnaire (YAACQ) | Concurrent substance use & effects on neural function | Alcohol | Emotional Go/No-Go with emotional photos (International Affective Picture System) | Negative NoGo > Neutral NoGo | Lateral frontoparietal networks (rL- FPN; lL- FPN), dorsal attention network (DAN), salience network (SN) | Alcohol use and negative consequences of drinking were negatively associated with DAN recruitment to negative Go trials. This pattern suggests that in young adults with more problematic drinking, negative emotional information interferes more with engagement of neural networks involved in top- down attentional control, compared to those with less alcohol misuse. |
| Cohen-Gilbert et al. (2017) |
N = 23 Mage = 18.8 (0.4) 69.6 % Female |
Self-report via AUDIT, Baratt Impulsiveness Scale | Concurrent substance use & effects on neural function | Alcohol | Emotional Go/No-Go with emotional photos (International Affective Picture System) | Negative NoGo > Neutral NoGo | OFC, amygdala, MFG, frontal pole, iTG, occipital pole, precuneus, cerebellum | Binge drinking was negatively associated with activity in the dlPFC, dmPFC, and AI. This pattern emerged for negative emotion only, and not positive emotion. |
| Elsayed et al. (2018) |
N = 330 (n = 32 early initiators) Mage = 13.37 (1.08) 56 % White, 44 % Other Race 44 % Girls (early initiators) 50 % Girls (late initiators) |
Self-report via Substance Use Questionnaire (SUQ) and Alcohol Expectancy Questionnaire (AEQ) | Effects of neural function on substance use initiation | Polydrug | Face-Matching Task (Angry and Fearful Faces), Card-Guessing Task (Monetary Reward) | Angry and Fearful Faces > Neutral Shapes Monetary Reward > Baseline |
Amygdala, VS | Adolescents who were classified as early initiators showed greater amygdala activation to fearful faces, compared to those classified as late initiators. Activity in the VS to monetary reward did not differ between groups. |
| Gorka et al. (2013) |
N = 12 Mage = 23.2 (1.8) 83.3 % Male, 66.7 % Caucasian |
In-person screening; ingestion of alcohol (0.8 g/kg; 16% volume) or placebo (dextrose with 0.0 g/kg; 1% volume ethanol as a taste mask) | Effects of substance use on neural function | Alcohol | Emotional Face Assessment Task | Fearful Faces > Neutral Faces Happy Faces > Neutral Faces |
Amygdala, OFC, MTG, insula, precuneus, L supplementary motor area | Alcohol ingestion led to reduced coupling between the amygdala and right OFC to faces, regardless of emotion (both threatening and happy faces). Alcohol also led to reduced coupling between amygdala and left OFC to happy faces. |
| Gorka et al. (2015) |
N = 16 Mage = 20.8 (2.6) 50 % Female 57.3 % Black, 31.8 % White, 6.4 % Hispanic, 5.5 % Asian |
In-person screening; ingestion of THC (Marinol; 7.5 mg) or placebo (dextrose) | Effects of substance use on neural function | Cannabis | Emotional Face Assessment Task | Fearful Faces > Neutral Faces Happy Faces > Neutral Faces |
Amygdala, OFC, MTG, insula, precuneus, L supplementary motor area | THC ingestion led to increased coupling between amygdala and subregions of the mPFC and rostral ACC while viewing threatening faces, compared to happy faces and neutral shapes. |
| Heitzeg et al. (2015) |
N = 40 Mage = 20.17 (1.38) 100 % Female |
Interview and self-report | Effects of substance use on neural function | Cannabis | Emotional Word task | Negative Emotion > Neutral Positive Emotion > Neutral |
Insula, iPL, dlPFC, superior frontal gyrus, right calcarine fissure, frontal gyrus | Heavy cannabis users had less activation to negative words in the right insula, PFC, and occipital cortex; less activation to positive words in the right iPL; less activation of amygdala to emotion (positive or negative); and higher dlPFC to positive words. PFC to negative words mediated adolescent cannabis use and young-adult negative emotionality. |
| Herman et al. (2019) |
N = 30 Mage = 23.40 (5.01) 70 % Female |
Self-report via Alcohol Use Questionnaire | Concurrent substance use | Alcohol | Affective Stop-Signal Task (Faces), Affective Delay Discounting Task (Faces) |
Fearful Faces > Neutral Faces | Lateral OFC, angular gyrus, left frontal pole, superior parietal lobule, postcentral gyrus | More binge drinking was linked to more activity during successful inhibition in the fearful context within frontal and parietal regions. More binge drinking was also related with a steeper decrease in frontal pole activity while making delayed decisions in the fearful context. |
| Leiker et al. (2019) |
N = 123 Mage = 15.95 (1.23) 63 % Male |
Self-report via AUDIT and CUDIT | Concurrent substance use | Polydrug | Face-Identifying Task (Fearful and Happy Faces) | Fearful Faces >Neutral Happy Faces > Neutral Emotional Faces (combined) > Neutral |
ACC, vmPFC, lingual gyrus, medial temporal pole, iPL | Adolescents’ AUD scores were negatively correlated with vmPFC and lingual gyrus responses to emotional faces (both fearful and happy), and CUD scores negatively correlated with rostromedial PFC (including the ACC). Greater alcohol use was linked to higher iPL response to fearful faces. |
| Nikolova et al. (2016) |
N = 170 Mage = 19.55 (1.26) 61.2 % Female 45 % Caucasian, 26 % Asian, 18 % African-American, 6 % Bi/Multi-racial, 5% other |
Self-report | Effects of neural function on substance use initiation | Alcohol | Face-Matching Task (Emotion), Number-Guessing Task (Reward) | Fearful Faces > Neutral Faces Positive Reward Feedback > Negative Reward Feedback |
VS, NAcc, amygdala | Higher risk for stress-related problem drinking was associated with (1) high VS response to reward and low amygdala response to threat, and (2) low VS response to reward and high amygdala response to threat. |
| Spechler et al. (2015) |
N = 140 (IMAGEN) Mage = 14.69 (0.53) 71.4 % Male (Cannabis users), 58.6 % Male (Non-using controls) |
In-person screening | Effects of substance use on neural function | Cannabis | Emotion Video facial task (video clips of faces turning angry) | Negative Emotional Stimuli > Neutral | PFC, amygdala, hippocampus, striatum | Adolescents who tried cannabis showed greater activity to angry faces in the amygdala, and those who had never tried cannabis showed lower TPJ to angry faces as well as higher activity to neutral faces in the bilateral dlPFC and right TPJ. |
| Spechler et al. (2020) |
N = 1119 (IMAGEN) Mage = 14.41 (0.4) 54.6 % Female |
Self-report via Alcohol and Other Drugs questionnaire | Effects of neural function on substance use initiation | Cannabis | Face Processing task | Angry Faces > Neutral Faces | Amygdala | Heightened amygdala response to angry faces at age 14 predicted greater use of cannabis by age 19, and the lowest amygdala activation was seen in adolescents who remained abstinent by age 19 in a dose-dependent pattern. Amygdala activation at age 19 did not differ between cannabis users and non-users. |
| Sullivan et al. (2022) |
N = 66 (34 Cannabis users with 50 + lifetime occasions or 40 + in last year) Mage = 21.3 (2.27) 47 % Female 64 % Caucasian |
Self-report via Timeline Follow Back (TLFB), Customary Drinking and Drug Use Record (CDDR) | Effects of substance use on neural function | Cannabis | Emotional Go/No-Go | Fearful Faces > Calm Faces | Rostral ACC | Decreased left and right rACC activation was found during successful inhibition with fearful faces, in cannabis users (abstinent at time of study) compared to non-users. Greater connectivity between right rACC and right cerebellum was found during successful inhibition with calm faces for male cannabis users, compared to female cannabis users. |
| Swartz et al. (2017) |
N = 377 Mage = 19.8 (1.3) 59 % Female 48 % Caucasian, 32 % Asian, 9 % African-American, 7 % Bi/Multi-racial, 4 % other |
Self-report via AUDIT | Effects of neural function on substance use initiation | Alcohol | Emotional Face Assessment Task | Fearful Faces > Neutral Faces | Basolateral and centromedial amygdala | Greater amygdala activity to fearful faces was linked to peer-reported (but not self-reported) lower extraversion and higher conscientiousness. Conscientiousness was in turn related to problem drinking in men, but not women. Amygdala response indirectly predicted males’ drinking via conscientiousness. |
| Social Sensitivity Category | ||||||||
| Blair et al. (2021) |
N = 102 Mage = 16.54 (1.26) 66 % Male |
Self-report via AUDIT, CUDIT | Concurrent substance use | Polydrug | Retaliation Task (modified Ultimatum Game) | Retaliation Phase > Baseline | vmPFC, dmPFC, AI, caudate, periaqueductal gray, middle frontal gyrus, superior temporal gyrus | AUD scores were positively associated with brain activity during retaliation in the dmPFC, AI, and caudate, representing an exaggerated retaliation response. CUD scores were not significantly associated with neural response during the retaliation task. |
| Chung et al. (2020) |
N = 78 (n = 46 substance-naïve) Mage = 15.92 (0.78) 79 % White 54 % Male for substance-naïve 38% Male for substance-exposed |
Self-report via YRBS | Effects of substance use on neural function | Polydrug | Gambling Task with risky and safe choices, completed alone or after viewing peers’ choices | Safe choices by peer > Solo Risky choices by peer > Solo |
vmPFC, dmPFC, amygdala, temporal pole, right STS, left TPJ, precuneus, PCC | Substance-naïveté is linked to increased valuation of peers’ safer choices, as adolescents who never used substances showed stronger neural response while observing safe choices. Further, vmPFC activity while viewing peers’ safe choices was negatively related with substance exposure. |
| Gilman et al. (2016) |
N = 40 Mage = 20.9 (2.1) 100 % Female |
Self-report | Effects of substance use on neural function | Cannabis | Social Influence Task | Conformity > Deviation | NAcc, VS, caudate, PFC | Cannabis users showed more NAcc activation while following group influence, and greater activity was associated with greater cannabis use. cannabis users showed more dorsal caudate activation during the feedback phase. |
| Gilman et al. (2016) |
N = 43 Mage = 21.1 (2.2) 52.4 % Female |
Self-report, Multidimensional Iowa Suggestibility Scale (MISS) | Effects of substance use on neural function | Cannabis | Social Influence Task | Conformity > Deviation | dlPFC, dmPFC, vlPFC, vmPFC, IFG, PPC | Susceptibility to social influence was positively correlated with caudate response to social influence, and reaction time with activity in frontal and parietal regions. Cannabis users showed more IFG activity, while non-using controls showed more dlPFC activity. |
| Groefsema et al. (2020) |
N = 153 Mage = 22.78 (1.84) 100 % Male |
Online screening; Self-report | Concurrent substance use | Alcohol | Social‐Alcohol Cue‐Exposure (SACE) task, Beer‐Incentive‐Delay (BID) task | Social > Non-Social Alcohol > Soda |
ACC, vmPFC, superior frontal gyrus, VS, iPL, left STS, right TPJ | Alcohol cues (versus soda) elicited more VS and vmPFC activity. Social cues (versus non-social) elicited ACC, VS, and vmPFC. Social alcohol cues elicited STS and L-IPL. |
| Social Stress | ||||||||
| Beard at al. (2021) |
N = 181 Mage = 17.16 (0.44) 49.3 % Female, 100 % Mexican-American |
Self-report | Effects of neural function on substance use initiation | Polydrug | Cyberball (Social Exclusion) | Social Exclusion > Social Inclusion | sgACC, dACC, AI | dACC activity moderated the link between anxiety and substance use, such that adolescents with lower dACC response reported greater substance use when also reporting higher anxiety. |
| Gilman et al. (2016) |
N = 42 Mage = 21.05 (2.2) 52.4 % Female |
Self- report, Multidimensional Iowa Suggestibility Scale (MISS) | Effects of substance use on neural function | Cannabis | Cyberball (Social Exclusion) | Social Exclusion > Social Inclusion | vACC, right insula | Non-cannabis using controls showed insula activity during exclusion, but cannabis users did not. Both groups had vACC response to exclusion. Conformity was positively correlated with vACC in cannabis users (but not controls). |
| Shakra et al. (2018) |
N = 48 (n = 24 high anxiety, 24 high sensation-seeking) Mage = 20.4 (1.9) 48 % Female, 81 % White, 5 % Asian, 12 % Other, 2 % Unknown |
Michigan Alcohol Screening Test, AUD symptoms from Structured Clinical Interview for DSM-III-R: (SCID-NP), Experimental administration of 1 ml/kg of 95% USP alcohol, p.o. | Effects of substance use on neural function | Alcohol | Face Emotion Processing Task (FEPT; Karolinska Directed Emotional Faces set), Montreal Imaging Stress Task (MIST) | Negative Faces (averaged fearful, disgusted, angry, sad) > Neutral Faces Stress > No Stress |
AI, amygdala, mOFC, NAcc, perigenual ACC | Anxious young adults showed less amygdala response to threat after ingesting alcohol, which predicted problem drinking at follow-up. Anxious men (but not women) showed increased mOFC, pgACC and NAcc activity during stress. Sensation seeking men (but not women) showed decreased mOFC during stress after ingesting alcohol, which predicted problem drinking at follow-up. |
| Zhao et al. (2019) |
N = 51 (n = 28 cannabis users) Mage = 25.05 (4.33) 100 % Male |
In-person screening | Effects of substance use on neural function | Cannabis | Montreal Imaging Stress Task (MIST) | Stress > No Stress | Precuneus, dmPFC | Cannabis users had decreased stress-related reactivity in the precuneus, and increased connectivity between the precuneus and dmPFC. Behaviorally, they performed worse in the social stress condition, but not the no-stress condition. |
Note: AUDIT = Alcohol Use Disorders Identification Test; CUDIT = Cannabis Use Disorders Identification Test; YRBS = Youth Risk Behavior Survey; PFC = prefrontal cortex; OFC = orbitofrontal cortex; ACC = anterior cingulate cortex; PCC = posterior cingulate cortex; PPC = posterior parietal cortex; VS = ventral striatum; NAcc = nucleus accumbens; AI = anterior insula; MTG = medial temporal gyrus; STS = superior temporal sulcus; TPJ = temporoparietal junction; iPL = inferior parietal lobule; iTG = inferior temporal gyrus; v = ventral; d = dorsal; r = rostral; l = lateral; m = medial; sg = subgenual.