Patterns of Substance Use Initiation among Young Adolescents in a Northern Plains American Indian Tribe

Abstract

Background:

Substantial evidence documents problematic substance use in Northern Plains American Indian communities. Studies suggest that disparities can be traced to disproportionate rates of early substance use, but most evidence comes from the retrospective reports of adults or older adolescents.

Objective:

To use a prospective longitudinal design to examine substance use initiation patterns as they emerge among young American Indian adolescents.

Methods:

Four waves of data were collected across three consecutive school years from middle school students on a Northern Plains reservation (N = 450). Discrete-time survival analyses were used to estimate risks of initiation of cigarettes, alcohol, and marijuana from age 10 to 13.

Results:

Risk for cigarette initiation was relatively high at age 10 and stable until age 13. Marijuana risk was low at age 10 but increased sharply by age 12. Alcohol initiation lagged, not surpassing risk for cigarette initiation until age 13 and remaining below risk for marijuana initiation throughout middle school. Hazards for girls trended higher than those for boys across all substances, but differences did not reach significance.

Conclusion:

Initiation patterns among these American Indian adolescents differed from patterns reported in other US groups, particularly with respect to deviation from the sequence characterized the initiation of marijuana before alcohol that is predicted by the gateway theory.

Scientific Significance:

Findings suggest that prevention efforts with youth in this community should begin early with a primary focus on marijuana use. They also suggest the importance of examining sequences of substance initiation among youth in other American Indian communities.

INTRODUCTION

Problematic substance use is widespread in many American Indian communities and disparities in rates of DSM-defined substance use disorders are well documented (1–3). Research into the origins of substance use problems in national studies increasingly points to early adolescence as a critical period (4). Close to one-third of adolescents begin drinking by age 13 and 10%, by age 10 (5); tobacco use often begins even earlier. Early use is associated with increased risk for problematic substance use and substance use disorder later in life (1,6) and with a variety of concurrent problems, for example, suicide ideation and attempts (7), risky sexual behavior (early initiation of sexual activity, unprotected sex, and multiple sexual partners), driving while intoxicated and motor vehicle accidents, school failure, cognitive deficits and alterations in brain morphology and activity, psychopathology, and antisocial behavior (often resulting in criminal activity and incarceration) (5,6). In addition, the typical episodic/opportunistic pattern of substance use in early adolescence is associated with multiple adverse effects on health, including biological, cognitive, and social problems (6). Indeed, substance use among adolescents has long been seen as part of a constellation of problem behaviors that threaten successful developmental outcomes and, in the extreme, put adolescents at risk of not surviving to adulthood (8–10).

Reducing substance abuse disparities in American Indian communities may ultimately depend on understanding and averting early use among American Indian youth (11–13). Use appears to start earlier (14,15), to include earlier use of marijuana (14,16), and frequently to involve problematic patterns such as bingeing (17). Age of first substance use among American Indian youth has been decreasing in recent years (18). Links between early initiation, greater and prolonged use, and risk of disorder documented in other populations have also been found among American Indians (19,20).

These findings, however, come from retrospective accounts of older adolescents and adults, biased by filters of time and memory. Prospective data from early adolescence can provide a clearer picture of patterns and timing of emergent substance use. Such data are critical to effective substance abuse prevention efforts, providing information on what substances youth are experimenting with, when they are initiating use, and how the course of their use proceeds. This study documented initiation patterns as they unfolded in early adolescence among American Indian youth and explored differential patterns of initiation across substances and gender.

Of particular interest was situating patterns of early substance use within the literature on initiation sequences, notably “gateway” sequences (21). Substantial evidence suggests that “normative” initiation proceeds from tobacco and alcohol to illicit substances, beginning with marijuana and progressing to other illicit drugs (22,23). This sequence has implications for prevention – intervening earlier in the sequence is hypothesized to provide a more powerful lever for the prevention of substance disorder than is intervening later in the sequence. Studies both in the United States and internationally have shown the sequence to be largely invariant (22,23); but significant variations from the standard sequence have been noted, particularly among ethnic minority populations and in areas of concentrated disadvantage (24,25). These findings suggest that the sequence is contextual rather than universal, related to availability and local norms around substances rather than either to inherent properties of particular substances or to deviant patterns of use (22,24,26). Sequence violations also appear to be increasingly common (25). Given the unique contexts of American Indian reservations (e.g., laws prohibiting alcohol) and previous research suggesting divergence from gateway sequences in this population (16), we looked closely at the sequence of initiation among young adolescents in this study.

METHODS

Sample and Procedures

Both the original study and this secondary analysis project were reviewed and approved by the Colorado Multiple Institutional Review Board and the Research Review Board (RRB) overseeing research for the tribe within which these data were collected. Our agreement with the RRB stipulates that we protect community confidentially and the privacy of the tribal community. The agreement also requires that all reports emanating from this study undergo RRB review before being submitted for publication or for public presentation.

Participants were students attending middle schools on an American Indian reservation in the Northern Plains. In order to protect community confidentiality, as noted above, we do not name the specific reservation here but instead include broad descriptive information about this community to provide context. The reservation is located in one of the poorest regions in the United States, with striking gaps in both income and education compared with the rest of the country. It is relatively remote, isolated from population centers, and characterized by small villages separated by great distances. Rates of substance use disorder among older adolescents and adults are elevated (3). It is also important to note that this reservation is “dry” – possession of alcohol on the reservation is illegal, placing it in the category of illicit drugs, such as marijuana.

Data reported here were from a larger study of students attending all middle schools across the reservation (27). Participants completed surveys at their schools each semester from the spring of 2006 through the spring of 2009. The overall sample constituted 71% of the middle-school population of the reservation at Wave 1 (W1). A total of 81% of parents/guardians were located and asked to allow their children to participate; 98% of them provided consent; 90% of these youth then assented to participate. The subsample reported on here included students attending seven of the participating schools that were randomly selected for this component of the study; data from the first four waves were analyzed (spring, 2006, through fall, 2007). Age, gender, and grade distributions for the study sample are shown in Table 1.

TABLE 1.

Participant demographics at Wave 1 (N = 430).

	Number	Percentage of sample
Gender
Male	229	53.3
Female	201	46.7
Grade in school
6th	263	61.2
7th	167	38.8
Age (years)1
11	48	11.2
12	119	27.8
13	162	37.9
14 or older	99	23.1

Note:

¹Age data missing for two participants.

Measures

Students were asked “Do you smoke cigarettes?” and those who responded positively were asked, “How old were you when you first tried cigarettes?” Similarly, those who responded “yes” to the question “Have you ever had a drink (more than a sip) of alcohol, like beer, wine, or hard liquor?” were asked “Not counting small tastes, how old were you when you started drinking alcohol?” For marijuana use, adolescents who endorsed the statement “At least once in my life, I have used marijuana (also called weed or grass)” were asked to fill in the blank in the statement “I was ____ years old when I used marijuana the first time.”

To settle occasional inconsistencies in adolescents’ reports across waves, we accepted their first reports as most accurate (i.e., likely to be least influenced by recall bias). If an adolescent reported having already used a particular substance at W1 and reported an age of first use, we used that age as the age of initiation. If substance use was reported at W1 but no age of first use was reported, we used the adolescent’s age at W1 as the best available estimate of age of initiation. If an adolescent reported no history of substance use at W1 or if data were missing for W1, we followed the same process of determining age of first use at W2; this process was repeated for each subsequent wave of data. Adolescents who reported no history of substance use at any wave of data collection were coded as missing on age of use (right-censored for survival analyses).

Analyses

Variable construction and descriptive analyses were completed with SPSS. Discrete-time survival analyses (DTSAs) (28) were conducted with Mplus, using “type=complex” to adjust for clustering by school. DTSAs began with fitting a baseline hazard model for initiation of each substance, across 1-year periods from age 10 to 13. Youth who reported use of a substance before age 10 were left-censored for the analysis of that substance (i.e., they had initiated before the focal period, 10–13 years of age). Adolescents who did not report use of a particular substance by age 13 were right-censored for the corresponding analysis (i.e., they did not report use before the end of the time period of interest). Hazard estimates from these models represent the probability that youth who had not used each substance before each particular age would initiate use at that age. For example, a hazard of .05 for marijuana initiation at age 11 would mean that one in 20 children who had not used marijuana before their 11th birthday would be expected to initiate use before they turned 12.

After fitting the baseline model for each substance, we estimated a second series of models in which gender was included as a covariate, to determine whether the risk of initiation of each substance differed for boys and girls. Significant positive covariate coefficients would indicate greater risk for girls relative to boys; significant negative coefficients, greater relative risk for boys. We began with an assumption of proportional odds (i.e., that the effect of gender would be constant across ages 10–13); however, we also tested a nonproportional odds model to determine whether the effect of gender varied depending on age. We compared the Bayesian Information Criteria, sample-size adjusted (BIC_SSA), across models. The BIC_SSA is a measure of model fit that adjusts for both the number of model parameters and the size of the sample; lower BIC_SSA values represent better overall fit. For all three substances, these comparisons favored the proportional models (lower BIC_SSA values) and these models were retained.

Community Review

Findings were reviewed with the community advisory group that worked with us throughout this project, the EAST Council. (EAST was the local name for the research project, an acronym for Early Adolescent Substance use Trajectories.) The Council was made up of community members involved with youth on the reservation, including school counselors and teachers, leaders of community youth programs, physicians, and substance abuse counselors. We met with the Council periodically throughout the project to review research findings, solicit guidance on interpretation within the cultural context, and discuss next steps toward prevention efforts. The Council reviewed findings from the analyses presented here and suggested explanations for the deviation from the gateway sequence we observed; their input is reflected in the Discussion section.

In addition to review by the EAST Council, both the overall project and this report were reviewed by the tribal RRB; members of the review board also contributed to the interpretation of results presented in the Discussion section.

RESULTS

Table 2 shows the percentage of participants who reported having initiated use of each substance (lifetime use) at each age, from 10 to 13 years. Cigarette use was more common than marijuana use, which was, in turn, more common than alcohol use. Reports of lifetime use of each substance trended up over time, with marijuana showing the greatest jump in rates, between 11 and 12 years. At ages 12 and 13, girls reported having initiated more than boys did (particularly for cigarettes).

TABLE 2.

Percentage of adolescents who reported having initiated use of each substance at each age.

	Cigarettes (%)		Alcohol (%)		Marijuana (%)
	Males	Females	Males	Females	Males	Females
Age 10	25.3	25.9	7.0	5.0	17.0	9.5
	20.1–31.0	20.4–31.3	3.9–10.5	2.0–8.0	12.7–21.8	6.0–13.4
Age 11	36.2	42.3	12.7	15.9	25.8	21.4
	30.6–41.9	35.8–49.3	8.7–17.0	10.9–20.9	20.5–31.0	15.9–26.9
Age 12	43.2	58.7	20.5	27.4	42.8	46.3
	37.1–49.8	52.2–65.2	15.7–25.8	21.4–33.3	36.7–48.9	39.8–52.7
Age 13	50.2	67.2	34.1	40.8	52.4	58.2
	44.1–56.8	61.2–73.6	27.9–40.6	34.3–47.3	46.7–58.1	51.7–64.7

Note: Confidence intervals are shown in italics below each estimate.

To explicitly model the risk of initiation of each substance by age, we estimated discrete-time survival models for each substance. We included gender as a covariate and school as a cluster variable to control for nonindependence among adolescents within school. Figure 1 shows the hazard curves for cigarettes, alcohol, and marijuana, separately for boys and girls. The risk for initiation of each substance was substantial in this age period. The pattern of initiation across substances indicated that young adolescents were at earliest risk for cigarette use (already substantial at age 10), followed by marijuana use (escalating sharply by age 12). Risk for initiation of alcohol use was relatively delayed, catching up to risk for cigarette and marijuana initiation only around age 13. This order of initiation is inconsistent with gateway theory and with reports of substance initiation sequences typical among most US samples, which begin with licit substances (first cigarette use, then alcohol use) and then progress to illicit substances (first marijuana, then other illicit drugs) (22). However, the alternative sequence found here – cigarettes, then marijuana, then alcohol – is consistent with earlier reports among American Indian youth (16).

FIGURE 1.

Hazard curves for initiation of cigarettes, alcohol, and marijuana by gender.

As shown in Figure 1, estimates of the effects of gender on the hazards were consistently in the direction of greater risk for girls, but only the effect on risk for initiation of cigarettes was statistically significant (unstandardized estimate = .65, p < .000, standardized estimate = .84); gender effects for alcohol and marijuana were not significant.

DISCUSSION

This study provides information that can help guide substance abuse prevention efforts with American Indian youth. First, the standard gateway sequences of substance use initiation observed in other populations were not evident among young adolescents in this Northern Plains tribal community. These youth were at greater and earlier risk for marijuana initiation than for alcohol initiation, whereas the standard gateway sequence includes alcohol use as a precursor to marijuana use. As in other studies reporting deviations from the gateway sequence, a likely explanation for this pattern is contextual. A myriad of issues about context surround access and use patterns for cigarettes, alcohol and marijuana in this population. For example, in contrast to other communities in the United States, alcohol is an illicit substance by tribal law on this “dry” reservation, putting alcohol in the same category as marijuana vis-á-vis legal access. Access to alcohol is not uncommon, though, through bootlegging and availability in towns that border the reservation. However, it is not as easily accessible as it is in most US communities.

In addition, although the adult population on this reservation has been shown to have higher than national rates of substance use problems, they also have much higher than national rates of abstinence (3), a pattern that reflects the complex culture around alcohol use in this population. Young adolescents growing up in this context are likely exposed to conflicting messages and models regarding alcohol use that may alter their initiation patterns.

Given this context, the finding of relatively early initiation of marijuana use appears to be consistent with previous research that has indicated a “substitution” effect with regard to alcohol. For example, higher rates of marijuana use have been reported in communities where concerted efforts have been made to reduce adolescent alcohol use (26). Alternately, the pattern may be explained more by the relative availability of marijuana than by the relative unavailability of alcohol. Remote reservation locales conceivably make home-grown marijuana more available than alcohol that has to be brought in (surreptitiously) from outside reservation boundaries. As a result, marijuana may be as easy for young adolescents on this reservation to obtain as alcohol, perhaps even easier.

A cultural nuance suggested by the EAST Council may be important here as well. Because marijuana is technically a plant, it may be seen as more “natural” than alcohol and, thus, more congruous with Native culture and traditions. This, coupled with strong negative stereotypes about alcohol use among American Indians, may result in young adolescents within this cultural context perceiving less stigma associated with marijuana use than with alcohol use.

In addition, the culture around cigarette use is somewhat unique in the context of Northern Plains American Indian communities, where tobacco is used for ceremonial and other traditional purposes (e.g., gifts of tobacco as a sign of respect). Although clear differences exist between traditional uses of tobacco and regular cigarette use, the distinctions may be blurred for young adolescents.

The failure to find a gender difference in early initiation patterns is somewhat surprising, given plentiful evidence that males in general have higher rates of substance use problems later in adolescence and in adulthood than do females (3). The lack of a gender difference in early use is perhaps less surprising in this particular population, however, since adult rates of substance use disorder are much more comparable across gender than they are in both the United States more generally and in other American Indian populations (3). The roots of early substance use are likely to be somewhat different among girls and boys, with risks related to early puberty (29) and childhood sexual trauma (30) perhaps more important in risk for girls and early conduct problems more common precursors for boys.

In this report, we have focused on patterns of substance use initiation only; it will be important to follow this effort with an examination of what happens after that first smoke or drink, examining the developmental course of continuation, escalation, or cessation of use. Another shortcoming is that this study included only one reservation community. Although it cannot represent the diversity of American Indian youth experience across other communities and tribes, these rich data facilitate a template for investigating American Indian youth substance use more broadly.

Another limitation is that findings presented here are merely descriptive; they provide a picture of emergent substance use patterns in this American Indian population, but do not speak to etiology. A confluence of factors in this community are likely at the root of early use, including high rates of poverty, early and chronic exposure to traumatic and stressful life events (including neglect), and limited access to both educational opportunities and health care. Next steps in our data analyses will be to examine factors that put these young adolescents at particularly high risk for early substance use.

Taken together, these findings have important implications for the prevention efforts in this high-risk population. First, intervention needs to begin early. Intervention efforts in high school will be secondary prevention for many of these youth, focusing on reducing use among the majority who have already initiated. Primary prevention must take place earlier. Second, prevention programs developed for other populations and imported into this community will need to be adapted to the culture and distinct patterns of initiation within this context. An emphasis on alcohol prevention, for example, may miss the mark among these youth who come to alcohol somewhat later in their substance use trajectories. Finally, prevention programs for these youth will need to address the unique risk factors within this context, although further work is needed to understand what might be at the root of those risks. In particular, future work should include examinations of how both cultural norms related to substance use and the environmental availability of different substances impact initiation patterns within American Indian reservation contexts. In addition, our understanding of the roots of these substance use patterns could be richly informed by ethnographic studies, exploring more fully the interpretations offered by the EAST Council and RRB. Such studies could marry indigenous knowledge with scientific inquiry to deepen our understanding of the etiology of early substance use in this population.