Session: Using Time-Varying Effect Models to Understand Predictors of Substance Use and Depression within-Days and across Developmental Periods (Society for Prevention Research 27th Annual Meeting)

In this symposium, we present three unique papers that demonstrate the utility of using time-varying effect models (TVEM) to generate a better understanding of predictors of substance use (SU) and mental health across time. TVEM represents a unique opportunity to capture both changes within a day and across multi-year developmental periods as coefficients are allowed to vary in both strength and direction across time.

Paper 1, “Dynamic Associations between Negative Affect and Craving Within- and Across-Day,” explores the time-varying association between negative affect and craving for illicit substance use across 12 days. The authors found that within a given day, negative affect had the strongest association with craving until afternoon, then decreased until evening. When looking across days, the association between affect and craving declined over the first week, but then reversed on day 8, where the association between affect and craving increased continually through day 12.

Paper 2, “Age-Varying Association between Childhood Maltreatment and Depression and Substance Use,” examines the impact of child maltreatment on SU and depression across adolescence and into adulthood. TVEM results revealed that participants who experienced CM before 6^th grade were significantly more likely to be depressed throughout adolescence and into adulthood. CM was associated with both heavy episodic drinking (HED) and cigarette use in mid-adolescence and adulthood, but not during late adolescence. Additionally, the association between CM and SU was stronger for participants who did not attend college.

Paper 3, “Age-Varying Effects of Parental Warmth and Closeness on Adolescent and Young Adult Substance Use and Depression,” presents findings on the time-varying effect of parental warmth and closeness on SU and depression across adolescence and into young adulthood. Additionally, the authors found that the time-varying effects were moderated by participant gender. For men, parental warmth and closeness were most strongly associated with SU in early/mid-adolescence, but for some SU outcomes, warmth and closeness were no longer significant in early adulthood. For women, the association between warmth and closeness and SU typically remained significant across the transition into adulthood.

After the three papers have been presented, the discussant will compare and contrast the three papers, provide additional information on TVEM and how it can be applied, and moderate a discussion between presenters and symposium attendees. We feel these three presentations in concert will provide SPR attendees with an example of different ways TVEM can be applied, research questions TVEM can answer, and how TVEM can be used to inform appropriately timed interventions.