Impulsivity Moderates the Relation between Depressive Symptoms and Substance Use across Adolescence

Abstract

Adolescence is characterized by significant increases in substance use problems. Although early depressive symptoms have been linked to adolescent substance use, recent literature reviews highlight important inconsistencies in these relations that may be accounted for by behavioral disinhibition factors. The current study aims to examine impulsivity as a specific moderator of the relation between early depressive symptoms and trajectories of substance use using a 6-year, prospective longitudinal design. Participants included 247 male and female adolescents (52.5% White, 45% female) who were, on average, 13 years of age at baseline (SD_age = 0.90). They completed self-report inventories of depressive symptoms, impulsivity, substance use, and externalizing symptoms. Using a latent growth modeling approach, we found main effects for depressive symptoms and impulsivity, such that youth with lower initial levels of depressive symptoms and higher levels of impulsivity at baseline evidenced greater increases in substance use. Moreover, the interaction between impulsivity and depressive symptoms significantly predicted substance use, indicating that depressive symptoms were more positively related to substance use for youth reporting higher levels of impulsivity. Findings suggest that impulsivity is an important moderator of the relation between depressive symptoms and the development of substance use during adolescence and highlight the need for interventions focusing on both impulsivity and depression in minimizing risk of substance use.

Notes

¹ We also examined an identical model that did not control for externalizing symptomology. This model also fit the data well, χ²(45) = 65.13, p = .026, CFI = .95, TLI = .94, RMSEA = 0.05, 90% CI [0.02, 0.07], and the pattern of results was nearly identical. Specifically, the interaction between impulsivity and depressive symptoms predicted only the slope of substance use (std. est. = .82, p = .022).

Additional information

Funding

This research was supported in part by a grant from the National Institute on Drug Abuse Grant R01DA018647 (primary investigator, C. W. Lejuez).