摘要：Objectives. We examined disparities in age-related patterns of marijuana initiation in 2 culturally distinct American Indian reservation communities (from the Northern Plains and the Southwest) compared with a national sample. Methods. We used discrete-time survival models to estimate age-related risk for initiation with data from 2 population-based studies: the American Indian Service Utilization, Psychiatric Epidemiology, Risk and Protective Factors Project and the baseline National Comorbidity Survey. Results. Among respondents who were born before 1960, peak risk for marijuana initiation in all samples was at age 18 years, and risk was greatest in the national sample. Among those who were born later than 1960, risk peaked at age 16 years and was highest in the American Indian samples. Males were at increased risk compared with females, especially in the older cohort and the Southwest tribal sample. Conclusions. Findings of disproportionate risk for marijuana initiation among younger members of the tribal samples raise concerns that American Indian reservation youths may be increasingly vulnerable to drug use and its concomitants, which suggests a need for more aggressive prevention efforts in these communities. Disproportionately high rates of alcohol use among American Indians have engendered calls for intervention efforts that target American Indian communities. 1 – 4 At the same time, considerable gaps exist in our understanding of other substance-related disparities in these communities, especially among adults. In particular, it is unclear whether observed disparities in alcohol use reflect cultural patterns that are unique to alcohol or whether they simply represent one manifestation of a broader spectrum of heightened substance use in these communities. Including marijuana—particularly the early use of marijuana—in the investigation of disparities is a critical next step. Marijuana is the most commonly used illegal drug among both adolescents and adults across the United States 5 , 6 and among American Indians in particular. 7 – 12 Findings indicate that rates of marijuana use among American Indian youths may be higher than US rates as a whole and may be higher compared with rates among most other racial/ethnic groups. 7 , 9 – 12 In addition to alcohol, marijuana is an important initiating substance in reservation populations, especially on reservations where the possession and consumption of alcohol at any age is illegal. 13 Early initiation of marijuana use has been associated with a variety of negative outcomes, including increased risk for misuse ( Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, definition of substance abuse or dependence) 14 and increased risk for use of additional substances. 13 , 15 – 23 Other problem behaviors also have been associated specifically with early marijuana use, including proximal outcomes (e.g., psychopathology, antisocial behavior, and sexual risk taking) and distal outcomes (e.g., divorce and unemployment during adulthood). 15 , 17 , 24 – 27 Identifying populations at particular risk for early marijuana use can help direct intervention efforts to communities where interventions are most needed. Identifying periods of development when youths within these high-risk populations are at greatest risk can prompt interventions that are likely to be preventive rather than corrective. We know relatively little about age-related initiation patterns for marijuana use in American Indian communities. Some studies have found that American Indian youths have an increased risk for early marijuana use compared with other groups, 9 , 28 – 30 whereas other studies have found no differences. 31 Most studies of substance use among American Indians have depended on school-based samples, which often are unrepresentative because of substantial school dropout rates in these communities. 32 Furthermore, such samples often preclude comparisons of initiation patterns with national samples and do not allow for direct comparison of patterns across birth cohorts. 7 , 12 , 33 Studies that have reported population-based estimates using national data sources have typically included only small percentages of American Indians—percentages that are generally reflective of overall population distributions but not large enough to provide reliable estimates for American Indians. 34 , 35 Even when sufficient numbers of American Indians have been included, multiple tribal groups have routinely been combined, and underlying differences among various cultures have been minimized or disregarded. 36 Additionally, most large surveys that have included American Indians have had only small numbers of reservation-based participants; however, the unique ecologies of reservation communities are likely to shape attitudes about substance use and to contribute to disparities. For instance, many reservations are dry (possession of alcohol is illegal) or damp (possession of alcohol is not illegal, but no liquor is sold locally). Drugs, particularly marijuana, may be as readily available as alcohol in such settings. Moreover, studies that have examined substance use across American Indian groups have consistently reported tribal variation in prevalence of both use 37 – 39 and initiation patterns, 13 which makes the investigation of tribal differences imperative. Our study adds to the base of comparative information on the initiation of marijuana use in American Indian reservation communities. We used data from 2 large-scale population-based epidemiological studies: the American Indian Service Utilization, Psychiatric Epidemiology, Risk and Protective Factors Project (AI-SUPERPFP) 40 and the baseline National Comorbidity Survey (NCS). 41 Discrete-time survival analysis techniques examined marijuana initiation patterns and explored differences in both level and timing of risk for initiation across the tribal and national samples. The placement of these patterns within a national context provides a fuller picture of differences in marijuana use than has heretofore been available. One concern with estimates of marijuana use, particularly across different datasets collected at different time points, is that marijuana use rates fluctuate. Rates of use vary—sometimes dramatically across historical time periods—with older generations less likely to report lifetime marijuana use compared with younger generations. 42 Thus, we considered generational patterns in reports of marijuana initiation by comparing 2 birth cohorts. Additionally, because males have an elevated risk for initiating marijuana use compared with females, 15 , 43 – 47 we explored gender effects across samples and birth cohorts.
