Time-Varying Effects of Family Bonding on Marijuana Use: The Effects of Gender, Behavioral Disinhibition, and Parental Marijuana Use

Prevention studies have documented that children from families characterized by warmth, cohesion, and high levels of bonding among family members report less drug use than children from less bonded families.  However, little is known about whether the protective effect of family bonding varies over development and whether the relationship between family bonding and drug use persists beyond the adolescent years. The current study examines the relationship between family bonding and marijuana use from at ages 10-25. Additionally, we tested for individual differences in the strength of the relationship between family bonding and marijuana use by gender and behavioral disinhibition. Finally, we investigated whether the relationship between family bonding and marijuana use differed for families where parents were engaged in high levels of drug use compared to non-drug using parents.

Participants were drawn from the Raising Healthy Children study (n = 1040), a community study of risk and protective factors related to substance use and delinquency. Analyses were performed using time-varying effects modeling (TVEM), which is an extension of spline regression that estimates continuous longitudinal effects between predictors and outcomes. Using TVEM allowed us to test the stability of the relationship between family bonding and marijuana use from ages 10 to 25 and compare the strength of the association at different points in development.

Results indicated a curvilinear relationship between family bonding and marijuana use; the strongest protective association was evident at ages 14-16, then the association gradually weakened and became nonsignificant between the ages of 21-23. After age 23, however, the protective relationship between family bonding and marijuana use was again evident. These patterns did not differ by gender or when comparing bonding to marijuana using parents (n = 148) and bonding to non-using parents. However, among youth with more behavioral disinhibition, the relationship between family bonding and marijuana use was less protective between ages 14-18 than for youth with lower levels of behavioral disinhibition. Further analyses will examine differences for parents who binge drink or use other drugs and test for interactions between parental drug use, gender, and behavioral disinhibition.

Findings provide information about the optimal timing for delivery of family-centered preventive interventions for marijuana use. For example, promoting family cohesion, particularly during the ages where the most protective effect is shown, could help prevent or delay marijuana initiation among adolescents.