Risk calculation circuit abnormalities plus psychosocial risk variables predict problematic substance use in youth...

Area of Science:

• Neuroscience
• Developmental Psychology
• Addiction Research

Background:

• Problematic substance use in youth is a significant concern.
• Existing research links brain regions to substance use but lacks insight into decision-making processes.
• Integration of psychosocial and environmental risk factors into predictive models is limited.

Purpose of the Study:

• To investigate if brain activation during risky decision-making in drug-naïve, high-risk children predicts adolescent problematic substance use.
• To assess the predictive value of neural activity, psychosocial factors, and their combination.

Main Methods:

• Functional magnetic resonance imaging (fMRI) and the Balloon Analogue Risk Task (BART) were used in 95 high-risk youth (mean age 11.7 years).
• Cost-sensitive logistic regression models incorporated brain activation from key regions and psychosocial variables (family history, parental monitoring, violence exposure).
• Models were adjusted for demographic and clinical factors.

Main Results:

• Psychosocial factors alone showed moderate predictive accuracy (AUC=0.76).
• Neural activation alone had poor predictive accuracy (AUCs=0.60-0.67).
• Combining psychosocial and neural factors significantly improved prediction (AUCs=0.83-0.86), demonstrating good specificity and fair sensitivity.

Conclusions:

• Neural activity in risk evaluation, reward response, and sensory integration regions, alongside psychosocial factors, predicts later problematic substance use.
• Multidimensional models integrating neural and psychosocial data are valuable for early identification of youth at elevated risk for substance use.