Why and for Whom Is Marijuana Bad? Understanding the Mechanisms through Which Heavy Adolescent Marijuana Use Is Associated with Long-Term Adverse Effects: Evidence from a Chicago Cohort

Background: With changing policies and attitudes surrounding marijuana use in the U.S., it is critical to gain a better understanding of long-term health and social consequences of marijuana use for developing adolescents, and specifically the mechanisms driving outcomes over time. It is particularly important to study marijuana use among African Americans since recent data shows rates disproportionately increasing in this minority group. It is also critical to consider the role of criminal justice system involvement since African Americans are disproportionately arrested for marijuana possession. We further take into account that individuals use marijuana for different reasons (e.g., recreation, self-medication), as those who use marijuana heavily to cope with stress may have different outcomes than those who use for other reasons.

Methods: This study utilizes data from the Woodlawn Study, a prospective community cohort study of urban African American men and women from Chicago who were first assessed at age 6 (N=1,242) and have most recently been interviewed at age 42. We focus on frequency of marijuana use at age 16, short-term outcomes (mechanisms) at age 32, and long-term effects at age 42. Mechanisms explored include: school dropout, drug dependence, escalation to “harder” drugs, involvement with the drug trade, and criminal justice system involvement. We also explore effect modification by anxious mood. Outcomes include midlife mental health, socioeconomic status, social role functioning, crime, and drug problems. Models adjust for a range of childhood and adolescent risk factor.

Results: Results show that after adjusting for confounding, heavy adolescent marijuana users are 1.8 as likely to drop out of school (p=.015), twice as likely to escalate to cocaine or heroin use (p=.001), twice as likely to be involved in the drug trade (p=.014), and 1.85 as likely to have a drug arrest (p=.015) in young adulthood. Early marijuana use did not predict drug dependence. Escalating to cocaine or heroin use and having a drug arrest explained a significant proportion of the effect of marijuana on drug and crime outcomes in midlife while school dropout explained in part the association of adolescent marijuana with social outcomes. None of the proposed mechanisms explained associations of adolescent marijuana use with mental health. Follow-up analyses will test the competing mechanisms in a single structural equation model using Mplus and stratify by anxious mood.

Conclusion: Identifying the pathways through which marijuana leads to adverse effects over time for vulnerable populations can inform the risk and protective factors that should be the target of intervention programs aimed at decreasing the negative impact of early, heavy marijuana use.