UC San Diego

Objective: Much research has investigated the effect of adolescent substance use on neurocognitive abilities but the influence of neurocognition on later use behaviors has been relatively less studied, especially among youth who are already engaged in substance use. Hence, the goal of this dissertation project was to explore predictive associations of neurocognitive markers with substance use behaviors via the following 3 aims (Chapters).Methods: Chapter 1- A scoping review was conducted on the existing literature to synthesize current research on minimally reviewed neurocognitive domains and their predictive associations with substance use (hence papers on well-reviewed impulsivity facets were excluded). Secondary analyses were conducted in Chapters 2 and 3 with data from a longitudinal study for adolescents (Youth At Risk study) to investigate neurocognitive performance in late adolescence during maximum substance use, as predictors of changes in later use. Chapter 2- Hierarchical linear regression models with 4 neurocognitive abilities (inhibition/cognitive flexibility, visuospatial ability, verbal memory, working memory) predicting change in follow-up alcohol use (drinking days, average drinks per drinking day, peak drinks, binge episodes) were estimated while covarying for baseline age, follow-up duration, and sex. Chapter 3- To investigate relationships between neurocognition and cannabis and nicotine use, in addition to alcohol, hierarchical linear regression models predicting change in follow-up overall substance use frequency index scores and individual substance (alcohol, cannabis, nicotine) use outcomes were estimated. Follow-up analyses exploring this relationship within 3 groups of one-, co-, and tri-substance users were also conducted independently. Results: Our scoping review revealed a common theme where cognitive processing in constructs of working memory, attention, emotion regulation, and elevated reward circuit activity during decision-making in childhood and early adolescence predicted earlier onset, greater use escalation, and even development of substance use disorders in some instances. Findings from our 2 analytical aims suggest working memory, verbal memory, visuospatial ability, and inhibition/cognitive flexibility at maximum substance (alcohol, cannabis, and/or nicotine) use in late adolescence are useful as predictors of changes in later use behaviors. Conclusion: Our results have the potential to inform policies and intervention research on cognitive vulnerability markers in youth that predict substance use into adulthood.