Substance Use Profiles of Urban American Indian 8th, 10th and 12th Grade Students: A Latent Class Analysis

PURPOSE:  Although most (70%) American Indians (AI) now live in urban areas, rigorous research is lacking on the social determinants of this population’s vulnerability to and resilience against substance use.  Urban AI youth, like many of their counterparts from tribal communities, report relatively higher rates and earlier initiation of several types of substance use. Using latent class analysis, this paper describes the substance use patterns of urban AI youth and investigates how those patterns related to other risk and protective factors.

METHODS:  Data come from a sub-sample of urban AI youth (N=2,407) who were respondents in the 2012 Arizona Youth Survey, a state-wide self-report survey of 8^th, 10^th, and 12^th grade students. Eight 30-day substance use measures were examined:  Alcohol, Heavy Episodic Drinking, Tobacco, Inhalants, Marijuana, Other illicit drugs, Prescription drug misuse, and Over-the-counter drug misuse.

RESULTS: A latent class analysis incorporating all eight measures of substance use clearly supported a 4-class solution: (1) non-users (73%); (2) non-alcohol substance users (6%); (3) “gateway” users (alcohol/cigarettes/marijuana) (17%); and (4) multi-substance users, including illicit drugs (4%). Equality of means tests indicated significant differences across the classes in other risk behaviors including antisocial behavior, oppositional behavior, number of times bullied, perceived harmfulness of drugs, and antidrug attitudes.  While the non-users class was consistently the least at-risk, the converse was not found -- the multi-substance users class was not consistently the most at-risk. The non-alcohol substance user class had more experiences of being bullied, while the gateway class reported more oppositional behaviors. Although the classes differed in age and grade level composition (classes 1 & 2 were younger than classes 3 & 4), the demographic profile of the classes did not differ significantly by gender, parental education or presence in the home, household size or SES.

CONCLUSION: These findings add to scientific knowledge of the epidemiology of substance use among urban AI adolescents and describe urban AI youth’s vulnerability to substance according to distinctive profiles of use of specific types of substances. Although most of the latent classes of substance use mirrored patterns found in other samples of youth, the urban AI class that used several substances but not alcohol is distinctive, and may reflect social and cultural forces operating in unique ways in urban AI communities.  This knowledge, in a prevention framework, can better allow urban AI families to become the agents of change through which health disparities that impact AI youth can be reduced and eliminated.