Propensity score modeling of adolescent e-cigarette use and cognitive performance: One-year follow-up study

TL;DR

This study finds that adolescent e-cigarette use is linked to lower cognitive performance in reading recognition after one year.

Contribution

The study uses propensity score modeling to assess cognitive effects of e-cigarette use in adolescents.

Findings

Exclusive e-cigarette use was associated with lower Oral Reading Recognition scores.

Propensity score matching improved covariate balance between users and non-users.

Findings suggest potential neurocognitive risks of e-cigarette use in adolescents.

Abstract

Adolescent e-cigarette use remains an important public health challenge, and evidence on its neurocognitive effects at an early age is limited. This study examined associations between exclusive e-cigarette use and cognitive performance in adolescents.

This study is a propensity score modeling cohort study. This study performed a secondary data analysis on data collected from 21 US sites in the Adolescent Brain Cognitive Development (ABCD) Study between 1 October 2016 and 31 October 2018, with one year of follow-up data. Adolescents reported substance use at ages 12–13 years and completed National Institute of Health (NIH) Toolbox Cognition Batteries one year later, which consists of the following scores: Picture Vocabulary, Flanker Inhibitory Control and Attention, List Sorting Working Memory, Dimensional Change Card Sort, Pattern Comparison Processing Speed, Picture Sequence Memory, and Oral Reading Recognition. Propensity score models incorporating 12 confounders assessed associations between baseline exclusive e-cigarette use (vs no tobacco use) and follow-up cognitive performance at one year.

The cohort included 4574 participants (55.7% White, 25.3% Hispanic, 8.9% Black). Propensity score matching substantially improved covariate balance between e-cigarette users and non-users. Adolescents who exclusively used e-cigarettes in the past six months at baseline (n=35) had lower scores in Oral Reading Recognition (b[Standard Error (SE)]= -4.5 [1.3]; 95% CI: -7.3 – -1.7; p=0.003) and similar associations were observed in multivariable regression models adjusting for demographic, socioeconomic, and behavioral covariates.

Exclusive e-cigarette use among early adolescents was associated with poorer performance in specific cognitive domains. These preliminary findings raise concerns about potential neurocognitive implications of e-cigarette use and warrant confirmation in larger, longitudinal studies with longer follow-up.

Figures