Independent and Interactive Effects of Impulse Control and Sensation Seeking in Predicting Alcohol Use in Middle Adolescents

Introduction: Early onset of alcohol use is associated with subsequent abuse and dependence (Grant & Dawson, 1997). Developmental imbalance models (Casey et al., 2008) attribute the rise in adolescent risk-taking to increasing reward sensitivity and lagging impulse control. Although most studies have reported independent effects of sensation seeking (SS) and impulsivity on adolescent alcohol use (Khurana et al., 2015), some have found an interaction, suggesting that heightened SS predicts substance use only in the context of low impulse control (IC; Kim-Spoon et al., 2017). We used both self-report and behavioral measures of IC to examine potential moderating effects on the association between SS and alcohol use in mid-adolescents. We hypothesized that acting-without-thinking (AWT; self-report measure of impulsivity) would independently predict alcohol use and not moderate the effect of SS, whereas the behavioral measure of IC would be a significant moderator. This is because lab-tasks like SSRT are pure measures of top-down inhibitory control whereas AWT reflects the tendency to act impulsively in rewarding contexts (e.g., do you often buy things without thinking?), hence indirectly measuring lack of impulse control over rewarding urges (Khurana et al., 2018).

Method: Data from a community sample of adolescents (N=325; Mage (SD)=15.8±0.95 years; 52% female) were analyzed using logistic regression models. Alcohol use was assessed by self-reports of ever and recent (past 30 days) drinking. SS was measured using the Brief Sensation Seeking Scale (Hoyle et al., 2002). AWT was assessed using 9 items from the Junior Eysenck Impulsivity Scale (Eysenck et al., 1984). SSRT was used as the behavioral assessment of IC (Aron & Poldrack, 2006).

Results: Both AWT (B(SE)=1.26 (0.46), p<0.01) and SS (B(SE)=0.67(0.17), p<0.001) were significantly associated with alcohol use. Although SSRT and AWT were correlated at 0.19, SSRT did not have a significant direct effect on alcohol use. Examination of moderation effects revealed a significant interaction between SSRT and SS (B(SE)= -0.01(0.003), p<0.05, such that high levels of SS were associated with greater odds of alcohol use only at low levels of SSRT. These effects were significant controlling for age, gender, race-ethnicity, and socioeconomic status. AWT did not moderate the effect of SS on alcohol use.

Discussion: SS is associated with alcohol use in the context of low IC. AWT, which characterizes poor impulse control over rewarding urges, has a strong independent effect on alcohol use regardless of SS differences. Adolescents with poor impulse control are especially prone to early alcohol use onset and should be targeted in prevention efforts.