Conclusions

Emerging Findings and Broader Implications

Across paradigms, several patterns emerge regarding neural activation during emotional processing, inhibitory control under emotional conditions, and ER in individuals engaging in SU. During emotional processing contrasts (e.g., Maintain > Observe; Negative > Neutral; Fear-Go > Calm-Go), many studies report heightened activation in limbic and salience-related regions, including the amygdala, insula, hippocampus, and cingulate cortex, particularly in adults with CUD and in adolescents at risk for SU escalation [25, 29–31]. At the same time, other studies show blunted recruitment in frontoparietal and attentional control regions during both emotional processing and inhibitory control under emotional conditions, particularly in youth or adults with more severe SU [13, 16, 17, 28]. Findings from emotional Go/No-Go paradigms further suggest that SU-related differences may be most evident when inhibitory demands occur in negative contexts, rather than during inhibition or emotion processing alone [14].

During explicit ER contrasts (e.g., Regulate > Maintain), prefrontal and parietal control systems are consistently implicated. Some studies report increased recruitment of dorsolateral prefrontal and attentional regions, potentially reflecting greater emotional regulatory effort [31], whereasothers find reduced activation or weakened prefrontal-limbic functional coupling, suggesting less efficient coordination of control systems during attempts to downregulate negative affect [21, 22, 26].

Collectively, this work indicates that SU is associated with altered engagement of neural systems supporting emotional salience detection and goal-directed control, particularly when negative emotion and behavioral demands intersect. These alterations may increase reliance on substances as a means of modulating distress or arousal; however, it is also possible that repeated heavy SU contributes to progressive changes in frontolimbic and control networks over time. Longitudinal and intervention-focused research will be critical for disentangling directionality and clarifying how these neural patterns relate to risk, maintenance, and recovery.

Future Directions

This review suggests several interesting future directions. First, recent work on the neural correlates of ER in SU reflect a growing interest in moving beyond static group differences toward questions of change, vulnerability, and clinical relevance. Although much of the literature remains cross-sectional, findings from intervention-focused work suggest that ER-related neural systems may be modifiable. For example, Froeliger et al. [32] found that changes in neural engagement during ER following a mindfulness-based intervention were associated with reductions in nicotine use, linking within-person neural change to meaningful behavioral outcomes. Together with studies showing altered prefrontal, salience, and limbic engagement across Go/No-Go, reappraisal, and implicit paradigms, this work raises the possibility that these neural markers are not only correlates of SU but potential targets that can be modulated by intervention.

Further, longitudinal and risk-focused designs are beginning to clarify whether ER-related neural differences precede or follow SU. Reviewed findings collectively suggest that ER-related neural function may operate as both a marker of vulnerability and a consequence of SU exposure, underscoring the need for additional longitudinal studies that track neural and behavioral change over time. With respect to implicit ER, the relationship between neural reactivity and SU appears to differ meaningfully across development and sex [25, 28–30]; as such, future longitudinal work should examine developmental and sex-specific pathways to clarify how these associations emerge and change over time. More broadly, a central theme across implicit ER studies is that effects were highly context-dependent, with passive viewing paradigms more consistently revealing heightened limbic and salience-related reactivity in adults with CUD and adolescents at risk for escalation [25, 29, 30], whereas paradigms requiring active management of emotional interference more commonly showed reduced prefrontal and anterior cingulate recruitment [26, 27]. Future research should examine what specific task and contextual demands drive these divergent patterns, which would help identify the particular emotional or behavioral contexts in which implicit ER difficulties confer the greatest SU risk.

Next, methodological advances highlight the importance of studying ER in contexts that more closely reflect real-world emotional challenges. Emerging evidence suggests that SU-related differences are most apparent when emotional salience and cognitive control demands intersect [14], and that alterations may be better characterized at the level of circuit coordination rather than isolated regional activation [26]. Moving forward, research should prioritize more ecologically valid paradigms that capture emotionally meaningful stimuli, such as personalized negative emotional images or videos drawn from participants’ daily lives [e.g., 9, 33], an approach we are currently implementing in our lab. For example, we conducted a study in which adolescents viewed videos of their own parents expressing positive and negative emotion towards them, drawn from a laboratory task conducted two weeks prior [33]. In this study, we found that higher nucleus accumbens activation to videos of adolescents’ own parents showing positive emotions was linked at a trend level with greater adolescent SU. Such approaches may better model the types of emotionally charged situations in which SU risk unfolds.

Future studies may also integrate fMRI with ecological momentary assessment (EMA) to link neural markers of ER with real-time fluctuations in emotion and SU in naturalistic settings. This multimethod approach allows researchers to test whether neural indices of emotional reactivity or regulation moderate momentary associations between emotions and SU in daily life. For example, preliminary work from our lab found that adolescents’ momentary negative emotional experiences in daily life predicted their momentary SU and externalizing symptoms, particularly among those showing blunted amygdala activation to negative emotional images during fMRI [34].

Overall, this review underscores that alterations in neural systems supporting emotional processing, salience detection, and cognitive control, all of which are related to ER, are closely intertwined with SU across development. This work supports the development of neuroscience-informed interventions that focus on emotional processing and regulation to prevent and reduce SU.