- Email: [email protected]
Preventing substance use among native american youth
Addictive Behaviors, Vol. 25, No. 3, pp. 387–397, 2000 Copyright © 2000 Elsevier Science Ltd. Printed in the USA. All rights reserved 0306-4603/00/$–see front matter
Pergamon
PII S0306-4603(99)00071-4
PREVENTING SUBSTANCE USE AMONG NATIVE AMERICAN YOUTH: THREE-YEAR RESULTS STEVEN P. SCHINKE,* LELA TEPAVAC,† and KRISTIN C. COLE* *Columbia University School of Social Work; and †Delta Consulting Group
Abstract — This study developed and tested skills- and community-based approaches to prevent substance abuse among Native American youth. After completing pretest measurements, 1,396 third- through fifth-grade Native American students from 27 elementary schools in five states were divided randomly by school into two intervention arms and one control arm. Following intervention delivery, youths in all arms completed posttest measurements and three annual follow-up measurements. Youths in schools assigned to the intervention arms learned cognitive and behavioral skills for substance abuse prevention. One intervention arm additionally engaged local community residents in efforts to prevent substance use among Native American youth. Outcome assessment batteries measured youths’ reported use of smoked and smokeless tobacco, alcohol, and marijuana. Over the course of the 3.5-year study, increased rates of tobacco, alcohol, and marijuana use were reported by youths across the three arms of the study. Though cigarette use was unaffected by intervention, follow up rates of smokeless tobacco, alcohol, and marijuana use were lower for youths who received skills intervention than for youths in the control arm. Community intervention components appeared to exert no added beneficial influence on youths’ substance use, beyond the impact of skills intervention components alone. Finally, gender differences were apparent across substances, measurements, and study arms, with girls smoking more cigarettes and boys using more smokeless tobacco, alcohol, and marijuana. © 2000 Elsevier Science Ltd. Key Words. Native American youth, Substance use, Prevention.
Compared with adolescents of other American ethnic-racial groups, Native American youths use tobacco, alcohol, and other drugs earlier, at higher rates, and with more severe health, social, and economic consequences. Cigarette smoking prevalence among Native Americans exceeds rates for every other ethnic-racial group; and heavy use of cigarettes—consumption of at least a pack of cigarettes a day for the last 30 days—is particularly evident among Native Americans (Moncher, Schinke, Holden, & Aragon, 1989; Office of Applied Studies, 1998). Use rates of smokeless tobacco products, including snuff and chewing tobacco, are also higher among Native American children and youth than among their majority culture counterparts (Hall & Dexter, 1988; Schinke, Gilchrist, Schilling, Walker, Locklear, et al., 1986). Smokeless tobacco use is especially pronounced for young Indian women who consume the products at higher levels compared to young white women and to young men from most other ethnicracial groups of men (Schinke, Schilling, Gilchrist, Ashby, & Senechal, 1987). Research reported in this paper was funded by the National Cancer Institute (CA44903), and the National Institute on Drug Abuse (DA03277). Without the collaborative partnership of American Indian tribes, reservations, and communities in North and South Dakota, Idaho, Montana, and Oklahoma, this research would not have been possible. We thank these collaborators not only for allowing access to institutions and youth under their auspices, but more important for dedicating themselves to realizing the vision of a future generation of Native people empowered to release the grip of substance abuse on their lives and communities. We are very grateful to Joseph Trimble, Michael Moncher, and Greg Cornell for their contributions to planning, logistics, and field trial phases of the research. Finally, we recognize the invaluable contributions of Lawrence Scheier who added significantly to the article’s methodological considerations and rigor. Requests for reprints should be sent to Dr. Steven P. Schinke, 622 West 113th Street, New York, NY 10025; E-mail: [email protected] 387
388
S. P. SCHINKE et al.
Indian youth also drink alcohol more than their non-Indian peers (Beauvais & LaBoueff, 1985; Beauvais, Oetting, & Edwards, 1985). Relative to non-Indians, Native youth drink earlier, consume greater amounts of alcohol, and suffer higher levels of drinking-related consequences than non-Indian youth (Beauvais, 1992; Jumper-Thurman, 1995; Oetting & Beauvais, 1989). In 1993, 71% of Indian youth in grades 7 through 12 reported having ever used alcohol, and 55% reported having ever been drunk (Beauvais, 1996). Drug use is another serious problem for Native American young people. Regional and national survey data show that for nearly all drugs, use rates are higher among Native American youth than among non-Indian youth (Okwumaba & Duryea, 1987). Illustrative are findings on marijuana use. For the most recent reporting period, nearly 50% of Native American students in grades 7 through 12 reported having ever used marijuana, compared to 12% of non-Indian youth (Beauvais, 1998). Several explanations have been advanced for problems of substance abuse among Native Americans. These explanations include the spiritual meaning of intoxication, the recreational value of drinking and drug taking, peer pressure toward substance abuse, and cultural conflicts between Native Americans and the larger society (Bagley, Angel, Dilworth-Anderson, Liu, & Schinke, 1995; Jones-Saumty, Hochhaus, Dru, & Zeiner, 1983; Lewis, 1982; Pedigo, 1983; Schinke, Gilchrist, Schilling, Walker, Kirkham, et al., 1986; Walker & Kivlahan, 1984). Accounting for such explanatory factors and addressing influences associated with substance use among all American adolescents—regardless of ethnic-racial background—theory-based and empirically tested prevention programs have attempted to reduce substance abuse among Native American youth (Moncher & Schinke, 1994; Moran & Wolf, 1992; Schinke, 1996; Schinke, Moncher, & Singer, 1994; Schinke, Singer, Cole, & Contento, 1996). The more successful of those substance abuse prevention programs have considered the two cultures in which many Indian youth live: Native American culture and the dominant non-Indian culture (LaFromboise, 1988; Moran, 1992). Prevention efforts aimed to help Indian young people acquire bicultural competence are guided by theory and data suggesting that Native American youth who understand the prerogatives of both cultures may prosper over their counterparts who may identify with only one or the other culture (LaFromboise & Rowe, 1983; Oetting et al., 1983). In the present context, bicultural competence is the ability of Indian youth to adaptively interact and thrive within Indian and non-Indian cultures. Besides addressing bicultural issues, prior programs have helped Indian youth avoid problems with tobacco, alcohol, and drugs by gaining mastery in problem solving, decision making, communication, and media analysis skills. Such life skills, included in most contemporary prevention programs that enjoy empirical support, as salient for Indian youth as for their non-Indian peers (cf. Botvin & Schinke, 1997; Botvin, Schinke, & Orlandi, 1995). Despite the modest success of past programs to prevent substance use and abuse among Native American young persons, the search for long-lasting, effective programs must be redoubled. With each new generation, Indians in this country are inheriting a longer and sadder legacy of problems with tobacco, alcohol, and drug use. These problems are extraordinary, serious, and, in many instances, life-threatening. Though treating substance abuse among Native people is a viable option to vitiate some of these problems, preventing harmful substance use appears more humane, effective, and cost-beneficial. Accordingly, prevention approaches that lend themselves to early intervention with Indian youths in the elementary school years are sorely needed.
Preventing substance use
389
In particularly short supply are substance abuse prevention programs that can produce sustained effects, yet that do not require significant outlays of resources to deliver. Prevention efforts that resonate to young members of specific American Indian groups and that are suitable for a panoply of tribal affiliations and geographic settings must also be researched. Finally, tobacco, alcohol, and drug abuse prevention programs need to explore the additive value of engaging not only youths in responsive, skills-based approaches, but also that draw upon potentially salubrious resources from the larger environmental settings that distinguish Native American communities (cf. Hall et al., 1995). Toward meeting these needs and challenges, research reported in this paper describes a study of skills-based intervention delivered together with and separately from a community mobilization intervention in the service of helping Native American youth avoid problems with substance abuse. The research employed a clinical trial design and collected postintervention follow-up data for 3.5 years after baseline measurement. In so doing, the study sought to determine whether the skills approach with and without the community mobilization approach had a demonstrable long-term impact on smoked and smokeless tobacco consumption, alcohol drinking, and marijuana smoking among a large and representative sample of female and male Native American youth.
M E T H O D S
The study initially involved 1,396 informed and consenting Native American youths from 10 reservations in North and South Dakota, Idaho, Montana, and Oklahoma. Youths were third-, fourth-, and fifth-grade students in 27 tribal and public schools that served socioeconomically comparable Native American communities. The sample was 49% female and had a mean age of 10.28 years at time of pretest. Most youths (55.6%) lived in two-parent households. At the onset of the research, all youths completed an assessment battery asking them to report their use of cigarettes, smokeless tobacco (including snuff and chewing tobacco), alcohol, and marijuana. Biochemical samples were collected from all youths to increase the accuracy of their self-reported substance use (Harrell, 1985; Murray, O’Connell, Schmid, & Perry, 1987). Prior to the administration of the pretest and each subsequent outcome measurement battery, youths submitted saliva samples which they understood would be assayed to show actual levels of substance use. These samples were gathered through self-administration using prelabeled tubes, capped by subjects and collected by research assistants with completed measurement battery instruments. By school, youths were divided randomly into three arms. Two intervention arms engaged youths in 15 50-minute weekly sessions. Each session involved instruction, modeling, and rehearsal in cognitive-behavioral skills associated with substance abuse prevention (cf. Botvin, Baker, Dusenbury, Tortu, & Botvin, 1990). Intervention occurred during the spring term of the school academic year. Though intervention was derived from a conventional theoretical model of life skills training, it was tailored to the cultural prerogatives and everyday realities of Native American young people in the target western reservation settings. Cultural tailoring of the intervention aimed to provide Indian youth with skills to help them resist pressures toward substance use within Native society and in the larger dominant American society (Schinke et al., 1988). Within the context of contemporary Native American culture, youths learned problem-solving, personal coping, and interpersonal commu-
390
S. P. SCHINKE et al.
nication skills for preventing substance abuse. These skills were first explained by group leaders, and then demonstrated by slightly older peers. Every intervention session incorporated Native American values, legends, and stories. Cultural content not only addressed substance use issues in Indian society, but also addressed positive and holistic concepts of health and health promotion among Native American people. Accounting for wide variations that exist among American Indians of the plains’ reservations, tribes, and groups that were represented in the sample, culturally specific intervention components sought commonalties, yet drew from material particular to individual sites and communities. Cultural content therefore aimed to celebrate traditions that enjoy popularity among many tribal and regional groups and illustrate specific beliefs and ceremonies particular to individual Native American groups and subgroups. In the context of these culturally specific situations, youths acquired new skills by applying them initially to role-play situations, then subsequently to situations volunteered by youths from their daily lives. Skills practice prepared youths for substance use risk situations alone and with peers. As youths received instruction, coaching, and reinforcement from group leaders, they rehearsed strategies for refusing peer influences toward substance use and for initiating positive, healthy alternative activities. Other exercises increased youths’ awareness of Native American cultural traditions that run counter to substance abuse. For example, youths heard and discussed tales and legends of Native people who employed tobacco ceremonially and who lived in harmony with nature without consuming harmful substances. Every session included homework assignments for youths to gather information and testimonies from their surroundings on topics relevant to intervention. For example, youths observed and reported upon such events and activities as the manner in which advertisers associate tobacco and alcohol with enjoyment, youthfulness, and vitality; ways that people avoid tobacco, alcohol, and drugs in their lives; and additional ideas for promoting abstinence from substances. Based on their observations and experiences, youths discussed the functions, perceived benefits, and consequences of substance use, not only in their lives but also in Indian and White societies. Youths in schools assigned randomly to one intervention arm also participated in a community involvement component. Aimed at enhancing the positive outcome effects of skills intervention, this component mobilized Native American constituents in youths’ communities to support substance abuse prevention. Constituents for community involvement were youths’ families, teachers and school guidance counselors, neighborhood residents, law enforcement officials, and commercial establishments frequented by youths. Components of the community intervention included a series of activities to raise awareness of the substance abuse prevention message (cf. Perry, Kelder, Murray, & Klepp, 1992). For example, community intervention involved media releases about the benefits of substance abuse prevention efforts aimed at Native American young people. Also in the community intervention arm, flyers and posters were distributed to businesses, health and social service agencies, schools, and churches. Informational meetings were also held for parents, neighbors, and teachers, informing them about intervention components youths were receiving (Perry et al., 1992). Informational sessions took place at local schools and included poster-making exercises, mural painting, skits, and problem-solving contests. Youths in schools assigned to the control arm did not receive any intervention. Six months after intervention delivery and every 12 months thereafter for 3 years,
Preventing substance use
391
all youths were retested with measures employed at pretest. Semiannually, youths in the two intervention arms received booster sessions indexed to skills intervention alone or to skills intervention plus community components, respectively. Each booster session was delivered in two 50-minute sessions, paralleling procedures used for initial intervention delivery. Drawn from principles employed for initial intervention delivery, booster session content included developmentally appropriate content and tactics, reflecting youths’ increased risks for substance use and abuse. Those risks were due in part to youths’ greater amounts of unsupervised time, increased peer pressure and modeling of smoked and smokeless tobacco use, drinking, and drug use, and accessibility to licit and illicit substances.
R E S U L T S
Over the course of the 3.5-year study, attrition of youths occurred at each measurement following the baseline pretest occasion, but did not differ among arms or sites. Across sites, from baseline to postintervention measurement, a 6-month period, the total sample was reduced to 1,374 youths from the initial number of 1,396, an attrition rate of 1.58%. One year following postintervention measurement, or 18 months after pretest measurement, 1,329 youths remained in the study, a decrease of 3.28% from the 6-month measurement. Thirty months after pretest measurement, 1,268 youths were still available, for an attrition rate of 4.59% from the prior follow-up. Attrition between the 30- and 42-month follow-ups was 5.44%, or 69 youths. From baseline to final follow-up measurement, a span of 3.5 years, 197 youths left the study, for an overall attrition rate of 14.11%. By the method of Hansen and associates, one-way ANOVAs compared retained and nonretained youth on baseline and subsequent prior outcome variable scores (Hansen, Collins, Malotte, Johnson, & Fielding, 1985). These analyses failed to confirm nonrandom attrition patterns between arms or among measurement occasions. A subset of saliva samples provided by each youth at every measurement was analyzed for the presence of cotinine, a nicotine derivative. These samples of saliva cotinine provided correlational evidence of the accuracy of youths’ self-reported tobacco use. Across study arms and measurement occasions, correlations between the presence of cotinine and self-reported smoking or smokeless tobacco averaged 0.53. Biochemical data were principally collected to increase the veracity of youth self-reports. Intentionally, therefore, few saliva samples were analyzed, and the rest discarded. The number of analyzed samples was too small to allow outcome comparisons of cotinine levels by arm or measurement occasion. Self-reported outcome data were examined relative to youths’ cigarette smoking, smokeless tobacco use, alcohol drinking, and marijuana use prior to intervention delivery and during each of the four postintervention periods. Indexed against baseline data collection, these four postintervention periods happened at 6 months, 18 months, 30 months, and 42 months. For data analytic purposes, cigarette smoking was defined as seven or more cigarettes smoked in the week prior to each measurement occasion. Similarly, smokeless tobacco use was defined as seven or more instances of snuff or chewing tobacco use in the week prior to each outcome measurement. Alcohol drinking was defined as four or more alcoholic drinks of any type consumed in the week prior to measurement. To be categorized as alcohol users, youths
392
S. P. SCHINKE et al.
must have reported consumption of four or more alcoholic drinks at one sitting, or consumption of one drink at each of four sittings, or any combination of consumption that totaled four drinks in the week before the index outcome measurement. Marijuana use was defined as four or more instances of marijuana or hashish use during the week prior to measurement. Such use could have included multiple instances of marijuana consumption in one day (e.g., several marijuana cigarettes smoked in one sitting), single marijuana uses on multiple days, or any combination thereof to result in youths having reported marijuana use a total of four times in the week prior to each measurement. Outcome analyses on self reported substance use information began with one-way ANOVA tests among the arms, with individual youths as the analytic unit. Significant omnibus F-ratios from the ANOVAs called for Scheffé post-hoc multiple comparison tests to find differences by study arm mean scores. Seen in Table 1, pretest scores did not differ among the three arms on subjects’ use of cigarettes, smokeless tobacco, alcohol, or marijuana. The presence of nonsignificant baseline rates did not implicate the use of covariates in subsequent analyses. Further, at measurements 6 and 18 months after pretest data collection, and following initial and booster session intervention for youths in the intervention arms, all study youths reported similar postintervention weekly rates of substance use. Though rates of weekly cigarette smoking steadily rose over time to over treble their baseline levels at the 3.5 year follow-up measurement occasion, no differences were evident among the three arms at any period. Smokeless tobacco use was lower for subjects in schools assigned to the skills arm than for subjects in schools assigned to the skills plus community arm or to the control arm, as reported at 30- and 42-month follow-up measurements, respectively, F(2, 1261) ⫽ 7.55, p ⬍ .0001; F(2, 1193) ⫽ 5.82, p ⬍ .001. Alcohol use was lower, according to youth reports, in the skills arm relative to the control arm at 30- and 42-month measurement occasions respectively, F(2, 1182) ⫽ 4.75, p ⬍ .01; F(2, 1171) ⫽ 6.03, p ⬍ .001. At the final follow-up, marijuana use rates were lower for Native American
Table 1. Percentages of Native American youths in the study’s three arms reporting weekly use of tobacco, alcohol, or marijuana at pretest, prior to intervention, and at 6-, 18-, 30-, and 42-month follow-up measurements Pretest
6 Months
18 Months
30 Months
Cigarette smoking (seven or more cigarettes in past week) Skills 10.82 14.23 14.31 22.51 Skills ⫹ community 12.05 18.54 19.09 25.95 Control 12.14 16.87 18.76 22.86 Smokeless tobacco use (seven or more snuff or chewing tobacco uses in past week) Skills 4.32 2.21 7.88 6.80a Skills ⫹ community 5.16 2.48 9.32 11.71b Control 7.04 3.12 8.41 10.48b Alcohol consumption (four or more drinks in past week) Skills 9.13 7.65 12.57 15.89a Skills ⫹ community 8.94 7.32 14.43 17.18a,b Control 8.72 8.31 15.55 19.06b Marijuana use (four or more marijuana uses in past week) Skills 7.11 9.17 5.39 4.93 Skills ⫹ community 5.68 8.34 5.44 6.12 Control 5.54 7.73 7.27 5.85
42 Months 35.17 38.44 40.75 10.23a 16.56b 17.83b 22.87a 25.44a,b 30.17b 7.03a 10.15a,b 14.84b
Note. Column means with dissimilar subscripts differ at p ⬍ .01 according to Scheffé post hoc comparisons.
Preventing substance use
393
youths in schools that received skills intervention than in schools that received no intervention, F(2, 1186) ⫽ 7.63, p ⬍ .0001. To discern gender differences in rates of substance use and in intervention effects, 42-month data were graphically compared for women and for men across the three study arms. Figure 1 reveals higher rates of cigarette use for females than for males. Still, rates of smokeless tobacco use, alcohol consumption, and marijuana use were lower for young women than for young men across arms at the final follow-up measurement occasion.
D I S C U S S I O N
These results shed light on substance use patterns and intervention outcomes among a diverse sample of Native American youth. Over the course of the 3.5-year research study, increased rates of tobacco, alcohol, and marijuana use were reported by youths across all three arms. Although rates of cigarette use were unaffected by youths’ receipt of intervention, follow-up rates of smokeless tobacco, alcohol, and marijuana use were lower for youths who received skills intervention than for youths in the control arm. Youths who received the skills plus community intervention had nonsignificantly lower use rates of all measured substances than control arm youths, and nonsignificantly higher use rates of all measured substances than skills intervention alone arm youths. Gender differences were somewhat surprising. Whereas young men reported higher
Fig. 1. Gender differences in weekly percentage use rates of cigarettes, smokeless tobacco, alcohol, and marijuana at the 42-month measurement occasion by study arm.
394
S. P. SCHINKE et al.
rates than young women of smokeless tobacco use, alcohol use, and marijuana use across arms and measurements, cigarette smoking rates for young women exceeded those for young men. Perhaps following trends in the majority culture and in other minority cultures, cigarette use among Native American youth is becoming a pattern more equally shared across genders (Beauvais & Segal, 1992; Schinke et al., 1992; Segal, 1990). As for other substance use, levels for Indian girls are below those for Indian boys most dramatically for snuff and chewing tobacco, next most for alcohol use, and least for marijuana. In fact, young women in the skills plus community arm, as well as those in the control arm, reported final follow-up rates of marijuana use equal to or higher than young men in the skills arm. The absence of strong intervention effects for Native American female and male youths in the skills plus community involvement intervention arm of the study warrants speculation. Possibly, inclusion of community involvement components diluted the combined intervention and vitiated its effectiveness with youths. Contrariwise, skills-only intervention may have provided youths with a concentrated approach to effect a positive impact. Whatever the underlying mechanism, the skills intervention resulted in final follow-up rates of smokeless tobacco, alcohol, and marijuana use that were respectively 43%, 24%, and 53% lower for youths in that arm of study compared with youths in the control arm. Those differences imply that the skills intervention exerted a demonstrable influence over Native American youths’ substance use patterns for up to 3.5 years after initial intervention delivery. Confidence in study findings increases because of methodological controls. With schools divided randomly into the three arms, the likelihood of contamination between and among intervention and control arms is small. No differences in substance use rates were evident at the study’s onset, indicating comparability among the three arms. An assessment of intervention fidelity, field observations showed a close correspondence between implementation of skills and community interventions and written protocols for each intervention arm. Biochemical data collection at each measurement occasion increased the probability of youths’ accurately reporting their substance use. Methodological strengths aside, the research is not flawless. Principal among its potential weaknesses is the study’s use of traditional analytic methods to discern differences among its three arms. Though ANOVA followed by multiple range tests of differences in group means—when significant omnibus effects appear—is a conventional way to compare behavioral outcomes in prevention trials, questions remain about underlying differences that may not emerge through such tests. The present study, for example, randomized schools to arms. But youths served as the units of analysis. This multilevel analytic procedure facilitated the differentiation of substance use outcomes by increasing the available degrees of freedom. Yet, the analysis may not have accounted for school and other nested group effects. Increasingly, prevention researchers recognize that social climate factors and other social groupings are stronger within schools than across schools. Unmindful of that relative strength, investigators can mistakenly attribute effects arising from social climate influences to preventive intervention effects. When randomization is at the school level, intact groups of youths can bring social factors (deviant and otherwise) into their respective study arms. Failing to control for the transfer of such social factors risks the introduction of Type I errors into data analyses, proceeded by attendant erroneous conclusions about outcome effects. For these reasons, prevention investigators attempt to control for the variance transferred from school and other social climate factors by estimating the intraclass
Preventing substance use
395
(e.g., intraschool, intragroup, or within other social groupings) correlation coefficient (ICC) on nested data. The estimated ICC can then be used to analyze study outcomes through random effects modeling or random coefficient modeling. Mixed-model analytic designs can thereby acknowledge the nested nature of data and determine intervention outcome effects with appropriate statistical controls for the ICC (cf. Murray et al., 1994; Siddiqui, Hedeker, Flay, & Hu, 1996). In its conventional design and through its possibly less sensitive treatment of outcome data, the present study did not pursue these newer methodological strategies. The strategies, nevertheless, warrant serious consideration in future prevention research. Notwithstanding unanswered questions about patterns of substance use within schools and within such groupings as age cohorts and genders, data yielded by this research have implications for the development of effective substance abuse prevention approaches aimed at Native American young people. Given the absence of proven interventions for Native American youths at risk for tobacco, alcohol, and other drug use, culturally sensitive interventions are timely. The present research suggests that such interventions can exert a positive effect on substance use. But, the study does not provide empirical evidence that the addition of community involvement components as used in the prevention program justified their expense and logistics. Indeed, the lack of outcome differences for youths in schools in the skills plus community involvement arm implies that that intervention was less effective than skills intervention alone (Schinke, Cole, & Singer, 1999). More research is needed on the prevention of substance abuse among Native American youth. Future research might profitably confirm or refute the results of such culturally tailored skills-based interventions as used in this study. Conceptual work is necessary to develop and test new prevention approaches and to refine existing interventions for Native American youth. Besides intervention studies, epidemiological and correlational research needs to explore the nature of substance use by Native American youth and the progression of substance use as youths mature. Perhaps data reported here will encourage other investigators to advance the emerging base of substance abuse prevention research among Native American youth. R E F E R E N C E S Bagley, S. P., Angel, R., Dilworth-Anderson, P., Liu, W., & Schinke, S. (1995). Adaptive health behaviors among ethnic minorities. Health Psychology, 14, 632–640. Beauvais, F. (1992). Indian adolescent drug and alcohol use: Recent patterns and consequences. American Indian and Alaska Native Mental Health Research, 5, 62–78. Beauvais, F. (1996). Trends in drug use among American Indian students, 1975–1994 with an adjustment for dropouts. American Journal of Public Health, 86, 1594–1598. Beauvais, F. (1998). American Indians and alcohol. Alcohol Health and Research World, 22, 253–259. Beauvais, F., & LaBoueff, S. (1985). Drug and alcohol abuse intervention in American Indian communities. International Journal of the Addictions, 20, 139–171. Beauvais, F., Oetting, E. R., & Edwards, R. W. (1985). Trends in drug use of Indian adolescents living on reservations: 1975–1983. American Journal of Drug and Alcohol Abuse, 11, 209–229. Beauvis, F., & Segal, B. (1992). Drug use patterns among American Indian and Alaskan Native youth: Special rural populations. Drugs and Society, 7, 77–94. Botvin, G. J., Baker, E., Dusenbury, L., Tortu, S., & Botvin, E. (1990). Preventing adolescent drug abuse through a multimodal cognitive-behavioral approach: Results of a three year study. Journal of Consulting and Clinical Psychology, 58, 437–446. Botvin, G. J., & Schinke, S. P. (1997). The etiology and prevention of drug abuse among minority youth. Binghamton, NY: The Haworth Press. Botvin, G. J., Schinke, S. P., & Orlandi, M. A. (Eds.). (1995). Drug abuse prevention with multiethnic youth. Thousand Oaks, CA: Sage Publications. Hall, R. L., & Dexter, D. (1988). Smokeless tobacco use and attitudes toward smokeless tobacco among
396
S. P. SCHINKE et al.
Native Americans and other adolescents in the Northwest. American Journal of Public Health, 78, 1586–1588. Hall, R., Lichtenstein, E., Burhansstipanov, L., Davis, S. M., Hodge, F., Schinke, S., Singer, B., Fredericks, L., & Glasgow, R. E. (1995). Tobacco use policies and practices in diverse Indian settings. American Indian Culture and Research Journal, 19, 165–180. Hansen, W. B., Collins, L. M., Malotte, E. K., Johnson, C. A., & Fielding, J. E. (1985). Attrition in prevention research. Journal of Behavioral Medicine, 8, 261–275. Harrell, A. (1985). Validation of self-report. In B. A. Rouse, N. J. Kozel, & L. G. Richards (Eds.), Selfreport methods as estimating drug use (pp. 12–21). Rockville, MD: National Institute on Drug Abuse. Jones-Saumty, D., Hochhaus, L., Dru, R., & Zeiner, A. (1983) Psychological factors of familial alcoholism in American Indians and Caucasians. Journal of Clinical Psychology, 39, 783–790. Jumper-Thurman, P. (1995). Native American community alcohol prevention research. In P. A. Langton (Ed.), The challenge of participatory research: Preventing alcohol-related problems in ethnic communities (pp. 245–258). Rockville, MD: Center for Substance Abuse Prevention. LaFromboise, T. D. (1988). American Indian mental health policy. American Psychologist, 43, 388–397. LaFromboise, T. D., & Rowe, W. (1983). Skills training for bicultural competence: Rationale and application. Journal of Counseling Psychology, 30, 589–595. Lewis, R. G. (1982). Alcoholism and the Native American: A review of the literature. In National Institute on Alcohol Abuse and Alcoholism: Alcohol and Health Monograph 4: Special Population Issues (pp. 315–328). Washington, DC: U.S. Government Printing Office. Moncher, M. S., & Schinke, S. P. (1994). Group intervention to prevent tobacco use among Native American youth. Research on Social Work Practice, 4, 160–171. Moncher, M. S., Schinke, S. P., Holden, G. W., & Aragon, S. (1989). Tobacco use by American Indian youth. Journal of the American Medical Association, 262, 1469–1470. Moran, J. (1992). The relevance of cultural sensitivity for alcohol prevention research. Paper presented at the working group meeting, Alcohol Prevention Research in Minority Communities, sponsored by the National Institute on Alcohol Abuse and Alcoholism, Washington, DC. Moran, J., & Wolf, W. (1992). Walking in balance: Alcohol intervention from an American Indian perspective. Paper presented at the meeting, Emerging Perspectives: Drug and Alcohol Interventions with People of Color, Denver, CO. Murray, D. M., O’Connell, C. M., Schmid, L. A., & Perry, C. L. (1987). The validity of smoking self-reports by adolescents: A reexamination of the bogus pipeline procedure. Addictive Behaviors, 12, 7–15. Murray, D. M., Rooney, B. L., Hannan, P. J., Peterson, A. V., Ary, D. V., Biglan, A., Botvin, G. J., Evans, R. I., Flay, B. R., Futterman, R., Getz, J. G., Marek, P. M., Orlandi, M., Pentz, M. A., Perry, C. L., & Schinke, S. P. (1994). Intraclass correlation among common measures of adolescent smoking: Estimates, correlates, and applications in smoking prevention studies. American Journal of Epidemiology, 140, 1038–1050. Oetting, E. R., & Beauvais, F. (1989). Epidemiology and correlates of alcohol use among Indian adolescents living on reservations. In D. L. Spiegler, D. A. Tate, S. S. Aitken, & C. M. Christian (Eds.), Alcohol use among U.S. ethnic minorities (pp. 239–267) (NIAAA Research Monograph No. 18, DHHS Pub. No. ADM 89-1435). Washington, DC: U.S. Government Printing Office. Oetting, E. R., Beauvais, F., Edwards, R., Waters, M. R., Velarde, J., & Goldstein, G. S. (1983). Drug use among Native American youth. Fort Collins, CO: Colorado State University. Office of Applied Studies. (1998). Prevalence of substance use among racial and ethnic subgroups in the United States 1991–1993. Rockville, MD: Substance Abuse and Mental Health Services Administration. Okwumaba, J. O., & Duryea, E. J. (1987). Age of onset, periods of risk, and patterns of progression in drug use among American Indian high school students. International Journal of the Addictions, 22, 1269– 1276. Pedigo, J. (1983). Finding the “meaning” of Native American substance use: Implications for community prevention. Personnel and Guidance Journal, 61, 273–277. Perry, C. L., Kelder, S. H., Murray, D. M., & Klepp, K. I. (1992). Community-wide smoking prevention: Long-term outcomes of the Minnesota Heart Health Program and the Class of 1989 study. American Journal of Public Health, 82, 1210–1217. Schinke, S. P. (1996). Behavioral approaches to illness prevention for Native Americans. In P. M. Kato & T. Mann (Eds.), Handbook of diversity issues in health psychology (pp. 367–387). New York: Plenum Press. Schinke, S. P., Cole, K. C., & Singer, B. (1999). Tobacco use prevention and dietary modification among American Indian youth in the Northeast. In C. S. Glover & F. S. Hodge (Eds.), Native outreach: A report to American Indian, Alaska Native, and Native Hawaiian communities (pp. 93–101). Rockville, MD: National Cancer Institute. Schinke, S. P., Gilchrist, L. D., Schilling, R. F., Walker, R., Kirkham, M., Bobo, J., Trimble, J., Cvetkovich, G., & Richardson, S. (1986). Prevention of drug and alcohol abuse in American Indian youths. Social Work Research and Abstracts, 22, 18–19. Schinke, S. P., Gilchrist, L. D., Schilling, R. F., Walker, R. D., Locklear, V. S., & Kitajima, E. (1986). Smokeless tobacco use among Native American adolescents. New England Journal of Medicine, 314, 1051–1052.
Preventing substance use
397
Schinke, S. P., Moncher, M. S., & Singer, B. R. (1994). Native American youths and cancer risk reduction: Effects of software intervention. Journal of Adolescent Health, 15, 105–110. Schinke, S. P., Orlandi, M. A., Botvin, G. J., Gilchrist, L. D., Trimble, J. E., & Locklear, V. S. (1988). Preventing substance abuse among American Indian adolescents: A bicultural competence skills approach. Journal of Counseling Psychology, 35, 87–90. Schinke, S. P., Orlandi, M. A., Vaccaro, D., Espinoza, R., McAlister, A., & Botvin, G. J. (1992). Substance use among Hispanic and non-Hispanic adolescents. Addictive Behaviors, 17, 117–124. Schinke, S. P., Schilling, R. F., Gilchrist, L. D., Ashby, M. R., & Senechal, V. A. (1987). Health effects of smokeless tobacco. Journal of the American Medical Association, 257, 781. Schinke, S. P., Singer, B., Cole, K., & Contento, I. R. (1996). Reducing cancer risk among Native American adolescents. Preventive Medicine, 25, 146–155. Segal, B. (1990). Drug-taking behavior among school-aged youth: The Alaska experience and comparisons with lower-48 states. New York: The Haworth Press, Inc. Siddiqui, O., Hedeker, D., Flay, B. R., & Hu, F. B. (1996). Intraclass correlation estimates in a school-based smoking prevention study: Outcome and mediating variables, by sex and ethnicity. American Journal of Epidemiology, 144, 425–433. Walker, R. D., & Kivlahan, D. R. (1984). Definitions, models, and methods in research on sociocultural factors in American Indian alcohol use. Substance and Alcohol Actions/Misuse, 5, 9–19.
Pergamon
PII S0306-4603(99)00071-4
PREVENTING SUBSTANCE USE AMONG NATIVE AMERICAN YOUTH: THREE-YEAR RESULTS STEVEN P. SCHINKE,* LELA TEPAVAC,† and KRISTIN C. COLE* *Columbia University School of Social Work; and †Delta Consulting Group
Abstract — This study developed and tested skills- and community-based approaches to prevent substance abuse among Native American youth. After completing pretest measurements, 1,396 third- through fifth-grade Native American students from 27 elementary schools in five states were divided randomly by school into two intervention arms and one control arm. Following intervention delivery, youths in all arms completed posttest measurements and three annual follow-up measurements. Youths in schools assigned to the intervention arms learned cognitive and behavioral skills for substance abuse prevention. One intervention arm additionally engaged local community residents in efforts to prevent substance use among Native American youth. Outcome assessment batteries measured youths’ reported use of smoked and smokeless tobacco, alcohol, and marijuana. Over the course of the 3.5-year study, increased rates of tobacco, alcohol, and marijuana use were reported by youths across the three arms of the study. Though cigarette use was unaffected by intervention, follow up rates of smokeless tobacco, alcohol, and marijuana use were lower for youths who received skills intervention than for youths in the control arm. Community intervention components appeared to exert no added beneficial influence on youths’ substance use, beyond the impact of skills intervention components alone. Finally, gender differences were apparent across substances, measurements, and study arms, with girls smoking more cigarettes and boys using more smokeless tobacco, alcohol, and marijuana. © 2000 Elsevier Science Ltd. Key Words. Native American youth, Substance use, Prevention.
Compared with adolescents of other American ethnic-racial groups, Native American youths use tobacco, alcohol, and other drugs earlier, at higher rates, and with more severe health, social, and economic consequences. Cigarette smoking prevalence among Native Americans exceeds rates for every other ethnic-racial group; and heavy use of cigarettes—consumption of at least a pack of cigarettes a day for the last 30 days—is particularly evident among Native Americans (Moncher, Schinke, Holden, & Aragon, 1989; Office of Applied Studies, 1998). Use rates of smokeless tobacco products, including snuff and chewing tobacco, are also higher among Native American children and youth than among their majority culture counterparts (Hall & Dexter, 1988; Schinke, Gilchrist, Schilling, Walker, Locklear, et al., 1986). Smokeless tobacco use is especially pronounced for young Indian women who consume the products at higher levels compared to young white women and to young men from most other ethnicracial groups of men (Schinke, Schilling, Gilchrist, Ashby, & Senechal, 1987). Research reported in this paper was funded by the National Cancer Institute (CA44903), and the National Institute on Drug Abuse (DA03277). Without the collaborative partnership of American Indian tribes, reservations, and communities in North and South Dakota, Idaho, Montana, and Oklahoma, this research would not have been possible. We thank these collaborators not only for allowing access to institutions and youth under their auspices, but more important for dedicating themselves to realizing the vision of a future generation of Native people empowered to release the grip of substance abuse on their lives and communities. We are very grateful to Joseph Trimble, Michael Moncher, and Greg Cornell for their contributions to planning, logistics, and field trial phases of the research. Finally, we recognize the invaluable contributions of Lawrence Scheier who added significantly to the article’s methodological considerations and rigor. Requests for reprints should be sent to Dr. Steven P. Schinke, 622 West 113th Street, New York, NY 10025; E-mail: [email protected] 387
388
S. P. SCHINKE et al.
Indian youth also drink alcohol more than their non-Indian peers (Beauvais & LaBoueff, 1985; Beauvais, Oetting, & Edwards, 1985). Relative to non-Indians, Native youth drink earlier, consume greater amounts of alcohol, and suffer higher levels of drinking-related consequences than non-Indian youth (Beauvais, 1992; Jumper-Thurman, 1995; Oetting & Beauvais, 1989). In 1993, 71% of Indian youth in grades 7 through 12 reported having ever used alcohol, and 55% reported having ever been drunk (Beauvais, 1996). Drug use is another serious problem for Native American young people. Regional and national survey data show that for nearly all drugs, use rates are higher among Native American youth than among non-Indian youth (Okwumaba & Duryea, 1987). Illustrative are findings on marijuana use. For the most recent reporting period, nearly 50% of Native American students in grades 7 through 12 reported having ever used marijuana, compared to 12% of non-Indian youth (Beauvais, 1998). Several explanations have been advanced for problems of substance abuse among Native Americans. These explanations include the spiritual meaning of intoxication, the recreational value of drinking and drug taking, peer pressure toward substance abuse, and cultural conflicts between Native Americans and the larger society (Bagley, Angel, Dilworth-Anderson, Liu, & Schinke, 1995; Jones-Saumty, Hochhaus, Dru, & Zeiner, 1983; Lewis, 1982; Pedigo, 1983; Schinke, Gilchrist, Schilling, Walker, Kirkham, et al., 1986; Walker & Kivlahan, 1984). Accounting for such explanatory factors and addressing influences associated with substance use among all American adolescents—regardless of ethnic-racial background—theory-based and empirically tested prevention programs have attempted to reduce substance abuse among Native American youth (Moncher & Schinke, 1994; Moran & Wolf, 1992; Schinke, 1996; Schinke, Moncher, & Singer, 1994; Schinke, Singer, Cole, & Contento, 1996). The more successful of those substance abuse prevention programs have considered the two cultures in which many Indian youth live: Native American culture and the dominant non-Indian culture (LaFromboise, 1988; Moran, 1992). Prevention efforts aimed to help Indian young people acquire bicultural competence are guided by theory and data suggesting that Native American youth who understand the prerogatives of both cultures may prosper over their counterparts who may identify with only one or the other culture (LaFromboise & Rowe, 1983; Oetting et al., 1983). In the present context, bicultural competence is the ability of Indian youth to adaptively interact and thrive within Indian and non-Indian cultures. Besides addressing bicultural issues, prior programs have helped Indian youth avoid problems with tobacco, alcohol, and drugs by gaining mastery in problem solving, decision making, communication, and media analysis skills. Such life skills, included in most contemporary prevention programs that enjoy empirical support, as salient for Indian youth as for their non-Indian peers (cf. Botvin & Schinke, 1997; Botvin, Schinke, & Orlandi, 1995). Despite the modest success of past programs to prevent substance use and abuse among Native American young persons, the search for long-lasting, effective programs must be redoubled. With each new generation, Indians in this country are inheriting a longer and sadder legacy of problems with tobacco, alcohol, and drug use. These problems are extraordinary, serious, and, in many instances, life-threatening. Though treating substance abuse among Native people is a viable option to vitiate some of these problems, preventing harmful substance use appears more humane, effective, and cost-beneficial. Accordingly, prevention approaches that lend themselves to early intervention with Indian youths in the elementary school years are sorely needed.
Preventing substance use
389
In particularly short supply are substance abuse prevention programs that can produce sustained effects, yet that do not require significant outlays of resources to deliver. Prevention efforts that resonate to young members of specific American Indian groups and that are suitable for a panoply of tribal affiliations and geographic settings must also be researched. Finally, tobacco, alcohol, and drug abuse prevention programs need to explore the additive value of engaging not only youths in responsive, skills-based approaches, but also that draw upon potentially salubrious resources from the larger environmental settings that distinguish Native American communities (cf. Hall et al., 1995). Toward meeting these needs and challenges, research reported in this paper describes a study of skills-based intervention delivered together with and separately from a community mobilization intervention in the service of helping Native American youth avoid problems with substance abuse. The research employed a clinical trial design and collected postintervention follow-up data for 3.5 years after baseline measurement. In so doing, the study sought to determine whether the skills approach with and without the community mobilization approach had a demonstrable long-term impact on smoked and smokeless tobacco consumption, alcohol drinking, and marijuana smoking among a large and representative sample of female and male Native American youth.
M E T H O D S
The study initially involved 1,396 informed and consenting Native American youths from 10 reservations in North and South Dakota, Idaho, Montana, and Oklahoma. Youths were third-, fourth-, and fifth-grade students in 27 tribal and public schools that served socioeconomically comparable Native American communities. The sample was 49% female and had a mean age of 10.28 years at time of pretest. Most youths (55.6%) lived in two-parent households. At the onset of the research, all youths completed an assessment battery asking them to report their use of cigarettes, smokeless tobacco (including snuff and chewing tobacco), alcohol, and marijuana. Biochemical samples were collected from all youths to increase the accuracy of their self-reported substance use (Harrell, 1985; Murray, O’Connell, Schmid, & Perry, 1987). Prior to the administration of the pretest and each subsequent outcome measurement battery, youths submitted saliva samples which they understood would be assayed to show actual levels of substance use. These samples were gathered through self-administration using prelabeled tubes, capped by subjects and collected by research assistants with completed measurement battery instruments. By school, youths were divided randomly into three arms. Two intervention arms engaged youths in 15 50-minute weekly sessions. Each session involved instruction, modeling, and rehearsal in cognitive-behavioral skills associated with substance abuse prevention (cf. Botvin, Baker, Dusenbury, Tortu, & Botvin, 1990). Intervention occurred during the spring term of the school academic year. Though intervention was derived from a conventional theoretical model of life skills training, it was tailored to the cultural prerogatives and everyday realities of Native American young people in the target western reservation settings. Cultural tailoring of the intervention aimed to provide Indian youth with skills to help them resist pressures toward substance use within Native society and in the larger dominant American society (Schinke et al., 1988). Within the context of contemporary Native American culture, youths learned problem-solving, personal coping, and interpersonal commu-
390
S. P. SCHINKE et al.
nication skills for preventing substance abuse. These skills were first explained by group leaders, and then demonstrated by slightly older peers. Every intervention session incorporated Native American values, legends, and stories. Cultural content not only addressed substance use issues in Indian society, but also addressed positive and holistic concepts of health and health promotion among Native American people. Accounting for wide variations that exist among American Indians of the plains’ reservations, tribes, and groups that were represented in the sample, culturally specific intervention components sought commonalties, yet drew from material particular to individual sites and communities. Cultural content therefore aimed to celebrate traditions that enjoy popularity among many tribal and regional groups and illustrate specific beliefs and ceremonies particular to individual Native American groups and subgroups. In the context of these culturally specific situations, youths acquired new skills by applying them initially to role-play situations, then subsequently to situations volunteered by youths from their daily lives. Skills practice prepared youths for substance use risk situations alone and with peers. As youths received instruction, coaching, and reinforcement from group leaders, they rehearsed strategies for refusing peer influences toward substance use and for initiating positive, healthy alternative activities. Other exercises increased youths’ awareness of Native American cultural traditions that run counter to substance abuse. For example, youths heard and discussed tales and legends of Native people who employed tobacco ceremonially and who lived in harmony with nature without consuming harmful substances. Every session included homework assignments for youths to gather information and testimonies from their surroundings on topics relevant to intervention. For example, youths observed and reported upon such events and activities as the manner in which advertisers associate tobacco and alcohol with enjoyment, youthfulness, and vitality; ways that people avoid tobacco, alcohol, and drugs in their lives; and additional ideas for promoting abstinence from substances. Based on their observations and experiences, youths discussed the functions, perceived benefits, and consequences of substance use, not only in their lives but also in Indian and White societies. Youths in schools assigned randomly to one intervention arm also participated in a community involvement component. Aimed at enhancing the positive outcome effects of skills intervention, this component mobilized Native American constituents in youths’ communities to support substance abuse prevention. Constituents for community involvement were youths’ families, teachers and school guidance counselors, neighborhood residents, law enforcement officials, and commercial establishments frequented by youths. Components of the community intervention included a series of activities to raise awareness of the substance abuse prevention message (cf. Perry, Kelder, Murray, & Klepp, 1992). For example, community intervention involved media releases about the benefits of substance abuse prevention efforts aimed at Native American young people. Also in the community intervention arm, flyers and posters were distributed to businesses, health and social service agencies, schools, and churches. Informational meetings were also held for parents, neighbors, and teachers, informing them about intervention components youths were receiving (Perry et al., 1992). Informational sessions took place at local schools and included poster-making exercises, mural painting, skits, and problem-solving contests. Youths in schools assigned to the control arm did not receive any intervention. Six months after intervention delivery and every 12 months thereafter for 3 years,
Preventing substance use
391
all youths were retested with measures employed at pretest. Semiannually, youths in the two intervention arms received booster sessions indexed to skills intervention alone or to skills intervention plus community components, respectively. Each booster session was delivered in two 50-minute sessions, paralleling procedures used for initial intervention delivery. Drawn from principles employed for initial intervention delivery, booster session content included developmentally appropriate content and tactics, reflecting youths’ increased risks for substance use and abuse. Those risks were due in part to youths’ greater amounts of unsupervised time, increased peer pressure and modeling of smoked and smokeless tobacco use, drinking, and drug use, and accessibility to licit and illicit substances.
R E S U L T S
Over the course of the 3.5-year study, attrition of youths occurred at each measurement following the baseline pretest occasion, but did not differ among arms or sites. Across sites, from baseline to postintervention measurement, a 6-month period, the total sample was reduced to 1,374 youths from the initial number of 1,396, an attrition rate of 1.58%. One year following postintervention measurement, or 18 months after pretest measurement, 1,329 youths remained in the study, a decrease of 3.28% from the 6-month measurement. Thirty months after pretest measurement, 1,268 youths were still available, for an attrition rate of 4.59% from the prior follow-up. Attrition between the 30- and 42-month follow-ups was 5.44%, or 69 youths. From baseline to final follow-up measurement, a span of 3.5 years, 197 youths left the study, for an overall attrition rate of 14.11%. By the method of Hansen and associates, one-way ANOVAs compared retained and nonretained youth on baseline and subsequent prior outcome variable scores (Hansen, Collins, Malotte, Johnson, & Fielding, 1985). These analyses failed to confirm nonrandom attrition patterns between arms or among measurement occasions. A subset of saliva samples provided by each youth at every measurement was analyzed for the presence of cotinine, a nicotine derivative. These samples of saliva cotinine provided correlational evidence of the accuracy of youths’ self-reported tobacco use. Across study arms and measurement occasions, correlations between the presence of cotinine and self-reported smoking or smokeless tobacco averaged 0.53. Biochemical data were principally collected to increase the veracity of youth self-reports. Intentionally, therefore, few saliva samples were analyzed, and the rest discarded. The number of analyzed samples was too small to allow outcome comparisons of cotinine levels by arm or measurement occasion. Self-reported outcome data were examined relative to youths’ cigarette smoking, smokeless tobacco use, alcohol drinking, and marijuana use prior to intervention delivery and during each of the four postintervention periods. Indexed against baseline data collection, these four postintervention periods happened at 6 months, 18 months, 30 months, and 42 months. For data analytic purposes, cigarette smoking was defined as seven or more cigarettes smoked in the week prior to each measurement occasion. Similarly, smokeless tobacco use was defined as seven or more instances of snuff or chewing tobacco use in the week prior to each outcome measurement. Alcohol drinking was defined as four or more alcoholic drinks of any type consumed in the week prior to measurement. To be categorized as alcohol users, youths
392
S. P. SCHINKE et al.
must have reported consumption of four or more alcoholic drinks at one sitting, or consumption of one drink at each of four sittings, or any combination of consumption that totaled four drinks in the week before the index outcome measurement. Marijuana use was defined as four or more instances of marijuana or hashish use during the week prior to measurement. Such use could have included multiple instances of marijuana consumption in one day (e.g., several marijuana cigarettes smoked in one sitting), single marijuana uses on multiple days, or any combination thereof to result in youths having reported marijuana use a total of four times in the week prior to each measurement. Outcome analyses on self reported substance use information began with one-way ANOVA tests among the arms, with individual youths as the analytic unit. Significant omnibus F-ratios from the ANOVAs called for Scheffé post-hoc multiple comparison tests to find differences by study arm mean scores. Seen in Table 1, pretest scores did not differ among the three arms on subjects’ use of cigarettes, smokeless tobacco, alcohol, or marijuana. The presence of nonsignificant baseline rates did not implicate the use of covariates in subsequent analyses. Further, at measurements 6 and 18 months after pretest data collection, and following initial and booster session intervention for youths in the intervention arms, all study youths reported similar postintervention weekly rates of substance use. Though rates of weekly cigarette smoking steadily rose over time to over treble their baseline levels at the 3.5 year follow-up measurement occasion, no differences were evident among the three arms at any period. Smokeless tobacco use was lower for subjects in schools assigned to the skills arm than for subjects in schools assigned to the skills plus community arm or to the control arm, as reported at 30- and 42-month follow-up measurements, respectively, F(2, 1261) ⫽ 7.55, p ⬍ .0001; F(2, 1193) ⫽ 5.82, p ⬍ .001. Alcohol use was lower, according to youth reports, in the skills arm relative to the control arm at 30- and 42-month measurement occasions respectively, F(2, 1182) ⫽ 4.75, p ⬍ .01; F(2, 1171) ⫽ 6.03, p ⬍ .001. At the final follow-up, marijuana use rates were lower for Native American
Table 1. Percentages of Native American youths in the study’s three arms reporting weekly use of tobacco, alcohol, or marijuana at pretest, prior to intervention, and at 6-, 18-, 30-, and 42-month follow-up measurements Pretest
6 Months
18 Months
30 Months
Cigarette smoking (seven or more cigarettes in past week) Skills 10.82 14.23 14.31 22.51 Skills ⫹ community 12.05 18.54 19.09 25.95 Control 12.14 16.87 18.76 22.86 Smokeless tobacco use (seven or more snuff or chewing tobacco uses in past week) Skills 4.32 2.21 7.88 6.80a Skills ⫹ community 5.16 2.48 9.32 11.71b Control 7.04 3.12 8.41 10.48b Alcohol consumption (four or more drinks in past week) Skills 9.13 7.65 12.57 15.89a Skills ⫹ community 8.94 7.32 14.43 17.18a,b Control 8.72 8.31 15.55 19.06b Marijuana use (four or more marijuana uses in past week) Skills 7.11 9.17 5.39 4.93 Skills ⫹ community 5.68 8.34 5.44 6.12 Control 5.54 7.73 7.27 5.85
42 Months 35.17 38.44 40.75 10.23a 16.56b 17.83b 22.87a 25.44a,b 30.17b 7.03a 10.15a,b 14.84b
Note. Column means with dissimilar subscripts differ at p ⬍ .01 according to Scheffé post hoc comparisons.
Preventing substance use
393
youths in schools that received skills intervention than in schools that received no intervention, F(2, 1186) ⫽ 7.63, p ⬍ .0001. To discern gender differences in rates of substance use and in intervention effects, 42-month data were graphically compared for women and for men across the three study arms. Figure 1 reveals higher rates of cigarette use for females than for males. Still, rates of smokeless tobacco use, alcohol consumption, and marijuana use were lower for young women than for young men across arms at the final follow-up measurement occasion.
D I S C U S S I O N
These results shed light on substance use patterns and intervention outcomes among a diverse sample of Native American youth. Over the course of the 3.5-year research study, increased rates of tobacco, alcohol, and marijuana use were reported by youths across all three arms. Although rates of cigarette use were unaffected by youths’ receipt of intervention, follow-up rates of smokeless tobacco, alcohol, and marijuana use were lower for youths who received skills intervention than for youths in the control arm. Youths who received the skills plus community intervention had nonsignificantly lower use rates of all measured substances than control arm youths, and nonsignificantly higher use rates of all measured substances than skills intervention alone arm youths. Gender differences were somewhat surprising. Whereas young men reported higher
Fig. 1. Gender differences in weekly percentage use rates of cigarettes, smokeless tobacco, alcohol, and marijuana at the 42-month measurement occasion by study arm.
394
S. P. SCHINKE et al.
rates than young women of smokeless tobacco use, alcohol use, and marijuana use across arms and measurements, cigarette smoking rates for young women exceeded those for young men. Perhaps following trends in the majority culture and in other minority cultures, cigarette use among Native American youth is becoming a pattern more equally shared across genders (Beauvais & Segal, 1992; Schinke et al., 1992; Segal, 1990). As for other substance use, levels for Indian girls are below those for Indian boys most dramatically for snuff and chewing tobacco, next most for alcohol use, and least for marijuana. In fact, young women in the skills plus community arm, as well as those in the control arm, reported final follow-up rates of marijuana use equal to or higher than young men in the skills arm. The absence of strong intervention effects for Native American female and male youths in the skills plus community involvement intervention arm of the study warrants speculation. Possibly, inclusion of community involvement components diluted the combined intervention and vitiated its effectiveness with youths. Contrariwise, skills-only intervention may have provided youths with a concentrated approach to effect a positive impact. Whatever the underlying mechanism, the skills intervention resulted in final follow-up rates of smokeless tobacco, alcohol, and marijuana use that were respectively 43%, 24%, and 53% lower for youths in that arm of study compared with youths in the control arm. Those differences imply that the skills intervention exerted a demonstrable influence over Native American youths’ substance use patterns for up to 3.5 years after initial intervention delivery. Confidence in study findings increases because of methodological controls. With schools divided randomly into the three arms, the likelihood of contamination between and among intervention and control arms is small. No differences in substance use rates were evident at the study’s onset, indicating comparability among the three arms. An assessment of intervention fidelity, field observations showed a close correspondence between implementation of skills and community interventions and written protocols for each intervention arm. Biochemical data collection at each measurement occasion increased the probability of youths’ accurately reporting their substance use. Methodological strengths aside, the research is not flawless. Principal among its potential weaknesses is the study’s use of traditional analytic methods to discern differences among its three arms. Though ANOVA followed by multiple range tests of differences in group means—when significant omnibus effects appear—is a conventional way to compare behavioral outcomes in prevention trials, questions remain about underlying differences that may not emerge through such tests. The present study, for example, randomized schools to arms. But youths served as the units of analysis. This multilevel analytic procedure facilitated the differentiation of substance use outcomes by increasing the available degrees of freedom. Yet, the analysis may not have accounted for school and other nested group effects. Increasingly, prevention researchers recognize that social climate factors and other social groupings are stronger within schools than across schools. Unmindful of that relative strength, investigators can mistakenly attribute effects arising from social climate influences to preventive intervention effects. When randomization is at the school level, intact groups of youths can bring social factors (deviant and otherwise) into their respective study arms. Failing to control for the transfer of such social factors risks the introduction of Type I errors into data analyses, proceeded by attendant erroneous conclusions about outcome effects. For these reasons, prevention investigators attempt to control for the variance transferred from school and other social climate factors by estimating the intraclass
Preventing substance use
395
(e.g., intraschool, intragroup, or within other social groupings) correlation coefficient (ICC) on nested data. The estimated ICC can then be used to analyze study outcomes through random effects modeling or random coefficient modeling. Mixed-model analytic designs can thereby acknowledge the nested nature of data and determine intervention outcome effects with appropriate statistical controls for the ICC (cf. Murray et al., 1994; Siddiqui, Hedeker, Flay, & Hu, 1996). In its conventional design and through its possibly less sensitive treatment of outcome data, the present study did not pursue these newer methodological strategies. The strategies, nevertheless, warrant serious consideration in future prevention research. Notwithstanding unanswered questions about patterns of substance use within schools and within such groupings as age cohorts and genders, data yielded by this research have implications for the development of effective substance abuse prevention approaches aimed at Native American young people. Given the absence of proven interventions for Native American youths at risk for tobacco, alcohol, and other drug use, culturally sensitive interventions are timely. The present research suggests that such interventions can exert a positive effect on substance use. But, the study does not provide empirical evidence that the addition of community involvement components as used in the prevention program justified their expense and logistics. Indeed, the lack of outcome differences for youths in schools in the skills plus community involvement arm implies that that intervention was less effective than skills intervention alone (Schinke, Cole, & Singer, 1999). More research is needed on the prevention of substance abuse among Native American youth. Future research might profitably confirm or refute the results of such culturally tailored skills-based interventions as used in this study. Conceptual work is necessary to develop and test new prevention approaches and to refine existing interventions for Native American youth. Besides intervention studies, epidemiological and correlational research needs to explore the nature of substance use by Native American youth and the progression of substance use as youths mature. Perhaps data reported here will encourage other investigators to advance the emerging base of substance abuse prevention research among Native American youth. R E F E R E N C E S Bagley, S. P., Angel, R., Dilworth-Anderson, P., Liu, W., & Schinke, S. (1995). Adaptive health behaviors among ethnic minorities. Health Psychology, 14, 632–640. Beauvais, F. (1992). Indian adolescent drug and alcohol use: Recent patterns and consequences. American Indian and Alaska Native Mental Health Research, 5, 62–78. Beauvais, F. (1996). Trends in drug use among American Indian students, 1975–1994 with an adjustment for dropouts. American Journal of Public Health, 86, 1594–1598. Beauvais, F. (1998). American Indians and alcohol. Alcohol Health and Research World, 22, 253–259. Beauvais, F., & LaBoueff, S. (1985). Drug and alcohol abuse intervention in American Indian communities. International Journal of the Addictions, 20, 139–171. Beauvais, F., Oetting, E. R., & Edwards, R. W. (1985). Trends in drug use of Indian adolescents living on reservations: 1975–1983. American Journal of Drug and Alcohol Abuse, 11, 209–229. Beauvis, F., & Segal, B. (1992). Drug use patterns among American Indian and Alaskan Native youth: Special rural populations. Drugs and Society, 7, 77–94. Botvin, G. J., Baker, E., Dusenbury, L., Tortu, S., & Botvin, E. (1990). Preventing adolescent drug abuse through a multimodal cognitive-behavioral approach: Results of a three year study. Journal of Consulting and Clinical Psychology, 58, 437–446. Botvin, G. J., & Schinke, S. P. (1997). The etiology and prevention of drug abuse among minority youth. Binghamton, NY: The Haworth Press. Botvin, G. J., Schinke, S. P., & Orlandi, M. A. (Eds.). (1995). Drug abuse prevention with multiethnic youth. Thousand Oaks, CA: Sage Publications. Hall, R. L., & Dexter, D. (1988). Smokeless tobacco use and attitudes toward smokeless tobacco among
396
S. P. SCHINKE et al.
Native Americans and other adolescents in the Northwest. American Journal of Public Health, 78, 1586–1588. Hall, R., Lichtenstein, E., Burhansstipanov, L., Davis, S. M., Hodge, F., Schinke, S., Singer, B., Fredericks, L., & Glasgow, R. E. (1995). Tobacco use policies and practices in diverse Indian settings. American Indian Culture and Research Journal, 19, 165–180. Hansen, W. B., Collins, L. M., Malotte, E. K., Johnson, C. A., & Fielding, J. E. (1985). Attrition in prevention research. Journal of Behavioral Medicine, 8, 261–275. Harrell, A. (1985). Validation of self-report. In B. A. Rouse, N. J. Kozel, & L. G. Richards (Eds.), Selfreport methods as estimating drug use (pp. 12–21). Rockville, MD: National Institute on Drug Abuse. Jones-Saumty, D., Hochhaus, L., Dru, R., & Zeiner, A. (1983) Psychological factors of familial alcoholism in American Indians and Caucasians. Journal of Clinical Psychology, 39, 783–790. Jumper-Thurman, P. (1995). Native American community alcohol prevention research. In P. A. Langton (Ed.), The challenge of participatory research: Preventing alcohol-related problems in ethnic communities (pp. 245–258). Rockville, MD: Center for Substance Abuse Prevention. LaFromboise, T. D. (1988). American Indian mental health policy. American Psychologist, 43, 388–397. LaFromboise, T. D., & Rowe, W. (1983). Skills training for bicultural competence: Rationale and application. Journal of Counseling Psychology, 30, 589–595. Lewis, R. G. (1982). Alcoholism and the Native American: A review of the literature. In National Institute on Alcohol Abuse and Alcoholism: Alcohol and Health Monograph 4: Special Population Issues (pp. 315–328). Washington, DC: U.S. Government Printing Office. Moncher, M. S., & Schinke, S. P. (1994). Group intervention to prevent tobacco use among Native American youth. Research on Social Work Practice, 4, 160–171. Moncher, M. S., Schinke, S. P., Holden, G. W., & Aragon, S. (1989). Tobacco use by American Indian youth. Journal of the American Medical Association, 262, 1469–1470. Moran, J. (1992). The relevance of cultural sensitivity for alcohol prevention research. Paper presented at the working group meeting, Alcohol Prevention Research in Minority Communities, sponsored by the National Institute on Alcohol Abuse and Alcoholism, Washington, DC. Moran, J., & Wolf, W. (1992). Walking in balance: Alcohol intervention from an American Indian perspective. Paper presented at the meeting, Emerging Perspectives: Drug and Alcohol Interventions with People of Color, Denver, CO. Murray, D. M., O’Connell, C. M., Schmid, L. A., & Perry, C. L. (1987). The validity of smoking self-reports by adolescents: A reexamination of the bogus pipeline procedure. Addictive Behaviors, 12, 7–15. Murray, D. M., Rooney, B. L., Hannan, P. J., Peterson, A. V., Ary, D. V., Biglan, A., Botvin, G. J., Evans, R. I., Flay, B. R., Futterman, R., Getz, J. G., Marek, P. M., Orlandi, M., Pentz, M. A., Perry, C. L., & Schinke, S. P. (1994). Intraclass correlation among common measures of adolescent smoking: Estimates, correlates, and applications in smoking prevention studies. American Journal of Epidemiology, 140, 1038–1050. Oetting, E. R., & Beauvais, F. (1989). Epidemiology and correlates of alcohol use among Indian adolescents living on reservations. In D. L. Spiegler, D. A. Tate, S. S. Aitken, & C. M. Christian (Eds.), Alcohol use among U.S. ethnic minorities (pp. 239–267) (NIAAA Research Monograph No. 18, DHHS Pub. No. ADM 89-1435). Washington, DC: U.S. Government Printing Office. Oetting, E. R., Beauvais, F., Edwards, R., Waters, M. R., Velarde, J., & Goldstein, G. S. (1983). Drug use among Native American youth. Fort Collins, CO: Colorado State University. Office of Applied Studies. (1998). Prevalence of substance use among racial and ethnic subgroups in the United States 1991–1993. Rockville, MD: Substance Abuse and Mental Health Services Administration. Okwumaba, J. O., & Duryea, E. J. (1987). Age of onset, periods of risk, and patterns of progression in drug use among American Indian high school students. International Journal of the Addictions, 22, 1269– 1276. Pedigo, J. (1983). Finding the “meaning” of Native American substance use: Implications for community prevention. Personnel and Guidance Journal, 61, 273–277. Perry, C. L., Kelder, S. H., Murray, D. M., & Klepp, K. I. (1992). Community-wide smoking prevention: Long-term outcomes of the Minnesota Heart Health Program and the Class of 1989 study. American Journal of Public Health, 82, 1210–1217. Schinke, S. P. (1996). Behavioral approaches to illness prevention for Native Americans. In P. M. Kato & T. Mann (Eds.), Handbook of diversity issues in health psychology (pp. 367–387). New York: Plenum Press. Schinke, S. P., Cole, K. C., & Singer, B. (1999). Tobacco use prevention and dietary modification among American Indian youth in the Northeast. In C. S. Glover & F. S. Hodge (Eds.), Native outreach: A report to American Indian, Alaska Native, and Native Hawaiian communities (pp. 93–101). Rockville, MD: National Cancer Institute. Schinke, S. P., Gilchrist, L. D., Schilling, R. F., Walker, R., Kirkham, M., Bobo, J., Trimble, J., Cvetkovich, G., & Richardson, S. (1986). Prevention of drug and alcohol abuse in American Indian youths. Social Work Research and Abstracts, 22, 18–19. Schinke, S. P., Gilchrist, L. D., Schilling, R. F., Walker, R. D., Locklear, V. S., & Kitajima, E. (1986). Smokeless tobacco use among Native American adolescents. New England Journal of Medicine, 314, 1051–1052.
Preventing substance use
397
Schinke, S. P., Moncher, M. S., & Singer, B. R. (1994). Native American youths and cancer risk reduction: Effects of software intervention. Journal of Adolescent Health, 15, 105–110. Schinke, S. P., Orlandi, M. A., Botvin, G. J., Gilchrist, L. D., Trimble, J. E., & Locklear, V. S. (1988). Preventing substance abuse among American Indian adolescents: A bicultural competence skills approach. Journal of Counseling Psychology, 35, 87–90. Schinke, S. P., Orlandi, M. A., Vaccaro, D., Espinoza, R., McAlister, A., & Botvin, G. J. (1992). Substance use among Hispanic and non-Hispanic adolescents. Addictive Behaviors, 17, 117–124. Schinke, S. P., Schilling, R. F., Gilchrist, L. D., Ashby, M. R., & Senechal, V. A. (1987). Health effects of smokeless tobacco. Journal of the American Medical Association, 257, 781. Schinke, S. P., Singer, B., Cole, K., & Contento, I. R. (1996). Reducing cancer risk among Native American adolescents. Preventive Medicine, 25, 146–155. Segal, B. (1990). Drug-taking behavior among school-aged youth: The Alaska experience and comparisons with lower-48 states. New York: The Haworth Press, Inc. Siddiqui, O., Hedeker, D., Flay, B. R., & Hu, F. B. (1996). Intraclass correlation estimates in a school-based smoking prevention study: Outcome and mediating variables, by sex and ethnicity. American Journal of Epidemiology, 144, 425–433. Walker, R. D., & Kivlahan, D. R. (1984). Definitions, models, and methods in research on sociocultural factors in American Indian alcohol use. Substance and Alcohol Actions/Misuse, 5, 9–19.