Initiation and Progression of Substance Use Among Young American Indian Adolescents: Patterns and Predictors

Introduction:  Substance use disorders are disproportionately prevalent in many American Indian communities, and the consequences of these disparities are far-reaching, both in terms of physical and mental health.  Early initiation of substance use is also well documented among American Indian youth, and links between early use and later disorder are clear.  Reducing disparities in substance use disorder, therefore, is likely to hinge at least in part on delaying initiation of use.  Developing interventions that successfully delay initiation will depend on understanding patterns of initiation and progression to multiple substance use, and identifying factors associated with elevated risk for both early initiation and rapid progression.

Methods:  This study used longitudinal data (4 waves) from middle school students on a Northern Plains American Indian Reservation (N=465) to examine initiation of cigarette, alcohol, and marijuana use from age 9 to 13.  Using discrete-time survival analysis (DTSA), we modeled age-related risk of initiation (hazard) of first substance use and, among those who had initiated use of one substance, age-related risk of progression to a second substance.  Hazard ratios were then modeled as a function of putative risk and protective factors, including early puberty, peer affiliation (deviant and prosocial), parent-child relationship factors (warmth, communication, shared activities), exposure to stressful experiences, and cultural identity and engagement.

Results:  Risk of initiation (first substance) increased steadily from age 10 to age 12, and tapered off slightly by age 13.  Youth who had not tried any substances before age 9 were not very likely to initiate at that age (hazard = .08); but more than a quarter of those who had not initiated use by age 12 were likely to do so at that age (hazard = .31) and the risk remained high at age 13 (hazard = .26).  The best-fitting DTSA model of first substance initiation did not show differential risk by gender.

Risk of progression to a second substance also increased steadily from age 9 to age 12, and then remained steady to age 13.   Unlike initiation of first substance use, however, the best fitting DTSA model for progression allowed risk to vary both across gender and across time (nonproportional odds model).  Girls were at significantly greater risk than boys, and their risk increased disproportionately across early adolescence.  At ages 9 and 10, risk was comparable for boys and girls (difference between hazards .03 at age 9 and .05 at age 10).  At age 11, however, risk of starting use of a second substance began to diverge, with hazards for girls .10 to .20 higher than for boys across ages 11 to 13.

Next, we examined the extent to which both initiation and progression were related to risk and protective factors and found several significant predictors of both initiation and progression.  The strength of associations varied across risk and protective factors, across initiation and progression, and across gender.

Conclusion:  Patterns of risk and protection for both initiation and progression will be discussed in terms of their implications for efforts to delay onset of substance use and prevent the emergence of substance use disorder among American Indian youth.