Adolescent substance use and functional connectivity between the ventral striatum and hippocampus.

Neurodevelopmental explanations for adolescent substance use have focused on heightened sensitivity of mesolimbic circuitry, centered on the ventral striatum (VS). Recent evidence suggests that, relative to adults, adolescents show a stronger link between reinforcement learning and episodic memory for rewarding outcomes and greater functional connectivity between the VS and hippocampus, which may reflect a heightened reward modulation of memory. However, a link between VS-hippocampal circuitry and adolescent substance use has yet to be established. Two separate studies were conducted to evaluate whether variation in VS-hippocampal resting-state functional connectivity (rs-FC) predicts subsequent adolescent substance use exposure. A pilot study (Study 1) was conducted in 19 youth recruited from a high sociodemographic risk population (N = 19; mean age = 13.3 SD = 1.4; 14 females; 47% Black Non-Hispanic, 32% White Non-Hispanic). To replicate results of Study 1, Study 2 utilized data from the National Consortium on Adolescent Neurodevelopment and Alcohol (N = 644; mean age = 16.3 SD = 2.5; 339 females; 11% Black Non-Hispanic, 11% Hispanic/Latino, 66% White Non-Hispanic). Resting-state fMRI data were collected at a baseline time point and lifetime and past year self-reported substance use was collected at a follow up visit. Regression models tested whether baseline VS-hippocampal rs-FC predicted substance use exposure at follow up, as measured by an index score reflecting the number of substance classes (e.g., alcohol, marijuana) tried and overall frequency of use. Across both studies, higher VS-hippocampal rs-FC at baseline predicted greater substance use exposure at follow up (pFWE < 0.05). These data provide the first evidence linking increased VS-hippocampal connectivity with greater adolescent substance use exposure. Results fit with the emerging idea that variation in adolescent substance use may relate to not only individual differences in mesolimbic sensitivity to reward, but also to an individuals' memory sensitivity to reward as measured by connectivity between canonical memory and reward regions.