Risk Factors for Adolescent Substance Use: Dimensions and Typologies of Temperament

• In Event: Poster Session 10
  In Poster Session: PS 10 Topic - Social, Emotional, Personality

Abstract

Background: As adolescent substance use is a health concern, research has sought to identify factors for early initiation of use. Specific temperamental dimensions, and resulting typologies, are among such factors (Conrod et al., 2008). Though temperament remains relatively stable through the lifespan (Pedlow et al., 1993), behavioral manifestations may evolve with cognitive and social experience and maturation. Nonetheless, prior temperament work has focused on childhood, leaving a gap with respect to assessment of temperament more proximal to the age of typical substance use onset. Objectives: This study sought to: 1) assess temperament during adolescence, 2) examine temperamental dimensions and typologies as predictors of substance use, and 3) extend assessment to include electronic (e-)cigarette use. Methods: We utilized dyadic (N=145) data provided by adolescents (Mage [Time 1] = 14.99, SD = 0.71; 62.07% female, 87.59% White, 82.76% Hispanic/Latinx) and their primary caregiver during the first two timepoints of a longitudinal study. At Time 1, adolescents and caregivers completed the Early Adolescent Temperament Questionnaire – Revised (EATQ-R; Ellis & Rothbart, 2001), which reflected levels of effortful control, negative affect, surgency, and affiliativeness. Adolescents reported on e-cigarette, alcohol, and marijuana use at Times 1 and 2. Data Analytic Plan: Path models were estimated to examine temperamental dimensions as predictors of adolescent substance use. Typologies were assessed through Latent Profile Analyses (LPA). Next, logistic regression models estimated whether the latent typologies differentially predicted adolescent substance use at Time 2. Separate models were estimated from parent- and adolescent-report of temperament to examine if results varied across informer. Substance use at Time 1, age, biological sex, and ethnicity were accounted for in each model. Results: The parent-report path model indicated that higher affiliativeness predicted greater e-cigarette use (b = .103, p < .05). The adolescent-report path model indicated that lower effortful control predicted greater alcohol use (b = -.177, p < .01) and that higher affiliativeness predicted greater marijuana use (b = .103, p < .05). Both parent- and adolescent-report based LPA models revealed an internalizing (low effortful control, high negative affect), a withdrawn (low affiliativeness), and a resilient (high effortful control, low negative affect, moderate surgency and affiliativeness) typology (Figure 1). Logistical regression analyses (Table 1) indicated that youth characterized by the parent-reported withdrawn typology relative to the resilient typology had a decreased likelihood of e-cigarette use (Odds Ratio [OR] = .291, 95% CI [.287, .881], p < .01). Youth characterized by the adolescent-reported internalizing typology relative to the resilient typology had an increased likelihood of alcohol use (OR = 7.05, 95% CI [1.17, 42.49], p < .05). Conclusion: Expanding examinations of temperament to include typologies may provide more nuanced information regarding which adolescents are at greatest risk for substance use initiation. For example, adolescents characterized by parent-reported high spontaneity and sociability are at risk for e-cigarette use, whereas those characterized by adolescent-reported low spontaneity and high negative emotionality are at risk for alcohol use. Findings could help identify youth who may be particularly vulnerable to traditional substances, such as alcohol, as well as more novel products, such as e-cigarettes.

Authors

• Sarah Anne Hartmann, Florida International University
  Presenting Author

• Timothy Hayes, Florida International University
  Non-Presenting Author

• Matthew T Sutherland, Florida International University
  Non-Presenting Author

• Elisa Trucco, Florida International University
  Non-Presenting Author