Neurocognitive Latent Factors Associate With Early Tobacco and Alcohol Use Among Adolescent Brain Cognitive Development Study Youth

Stephanie K Jones,Mary Lauren Benton,Bethany J Wolf,Jackson Barth,ReJoyce Green,Sara L Dolan

DOI: https://doi.org/10.1016/j.jadohealth.2024.06.017

2024-08-13

Abstract:Purpose: Prospective associations between preadolescent neurocognitive structure and onset of substance use in adolescence have not been examined. This study investigated associations between cognitive structure among youth aged 9 - 10 years and the likelihood of experimentation with tobacco and alcohol by ages 13-14 years. Methods: A principal component (PC) analysis of nine neurocognitive assessments was used to identify the cognitive structure of unrelated adolescent brain cognitive development study participants (n = 9,655). We modeled associations between neurocognitive PCs and odds of tobacco or alcohol use by ages 13-14 years using generalized linear mixed models with a logit link and random intercept for recruitment sites. Demographics, family conflict, neighborhood safety, and externalizing and internalizing behavior were considered covariates. Results: Four neurocognitive PCs were identified and labeled general ability, executive function, learning and memory, and mental rotation. Mental rotation [odds ratio [OR] = 0.88, p-value = .013] was associated with lower odds of youth tobacco use; the association was stronger among female youth. General ability [OR = 1.20, p-value < .0001] among both males and females, and learning and memory [OR = 1.11, p-value = .024] among females, were associated with increased odds of youth alcohol use. Discussion: Among youth, higher neurocognitive performance was protective for tobacco use but increased the likelihood of alcohol use. Potential sex differences were identified. The role of cognition in processing the social contexts surrounding tobacco and alcohol use in the United States may contribute to the formation of disparate youth expectancies for tobacco and alcohol use.