Methods: Data is from the nationally representative 2014 12th grade Monitoring the Future Survey (MTF) (n=8,696). Youth self-reports of substance use relating to e-cigarette, regular combustible cigarette, binge drinking, marijuana, and other illicit drugs were used to derive patterns of PSU. Posterior probability-based multiple imputations were used to determine mean differences across patterns of PSU in school context (school extracurricular participation, perceptions of school norms against drug use, current academic grades, suspension or expulsion), family (parental involvement), and individual variables (sensation seeking, life satisfaction, perception of drug use risk). A three-step multinominal logistic regression was used to compare how school, family, and individual variables relate to the probabilities of class membership across patterns of PSU.
Results: A 3-class model was chosen over the 4-class solution as it represented distinct PSU configurations. The “low level” users (75.1%) represented students who mostly abstained from substance use. The “pre-dominantly e-cigarette” users (8.4%) characterized youth who mainly use e-cigarettes with some amount of regular cigarette, marijuana, alcohol, and other drug use. The “poly-users” (16.4%) represented students who exhibited high amounts and frequencies of all substances. A number of school contextual factors differentiated the three patterns of PSU. Students with higher academic grades had lower odds of membership in the predominately e-cigarette group versus the low-level use group (OR = 0.83 [0.77-0.88]). Occurrences of suspension or expulsion were associated with higher odds of membership in the e-cigarette use group as compared to the low-level use (OR = 2.99 [1.29-6.98]). Pro-social school involvement was associated with lower odds of engagement among poly-users as compared to the e-cigarette use group (OR = 0.60 [0.42 – 0.80]). Important family and individual factors include parental involvement and sensation seeking. Higher parental involvement was associated with lower odds of engagement among poly-users as compared to the predominantly e-cigarette use group (OR = 0.49 [0.30 – 0.81]). Higher levels of sensation seeking were associated with higher odds of membership when comparing the e-cigarette use pattern and low-level users (OR: 1.28 [1.10 – 1.48]), and when comparing the poly-substance and e-cigarette use pattern (OR: 1.28 [1.04 – 1.56]).
Conclusions and Implications: The school context, among other family factors and individual characteristics, has important implications in understanding student involvement with PSU. Results indicate the importance of academic grades, serious school disciplinary problems (occurrence of suspension or expulsion), and school extracurricular activities to differentiate the different patterns of PSU. Findings illustrate the importance of screening for pertinent school, family, and individual risk factors to develop targeted interventions against youth substance use.