The influence of program mediators on eleventh grade outcomes for seventh grade substance users and nonusers

https://doi.org/10.1016/j.drugalcdep.2008.11.010Get rights and content

Abstract

Objectives

In their work examining the effects of the Take Charge of Your Life (TCYL) program, Sloboda and colleagues (This Issue) found that the TCYL program had significant positive effects on baseline marijuana users and significant negative effects on baseline nonusers of cigarettes and alcohol.

Methods

Mediational analyses were used to understand why the program had these differential impacts on baseline users and nonusers.

Results

Path models for binary outcomes revealed significant program impacts on marijuana normative beliefs and refusal skills. The treatment impacts were between 1.5 and 3 times larger for the baseline users than for nonusers. These direct effects of the program on normative beliefs and refusal skills mediated the treatment impact on use for baseline marijuana users. In contrast, the negative treatment effects on alcohol and cigarette use could not be explained by the program's targeted mediators (normative beliefs, refusal skills, consequences, attitudes and intentions). The direct effects of treatment on use for the baseline nonusers of cigarettes and alcohol remain unexplained.

Conclusions

Possible explanations for this pattern and implications for strengthening universal prevention programs that are delivered to both users and nonusers are discussed. The importance of mediational analyses for programs that show negative impacts, as well as for those that show positive impacts is stressed.

Introduction

In their work examining the effects of the Take Charge of Your Life (TCYL) Program, Sloboda et al. (2009) found, significant positive effects of the intervention for seventh grade marijuana users (baseline users) and significant negative effects of the intervention for seventh grade nonusers of cigarettes and alcohol (baseline nonusers). The program was not, however, designed to have direct effects on substance use. As discussed in Stephens et al. (2009), the program was designed to impact intentions to use alcohol, tobacco and marijuana by addressing normative beliefs about the prevalence of substance use among adolescents, perceptions of the harmful consequences of substance use, substance use attitudes, and to provide students with life skills such as decision making, communications and refusal skills. Based on the program model, any effects on substance use thus should be mediated through these factors. In their mediational analyses, conducted to examine the effects of these factors on intentions and eventually use, Stephens et al. (2009) found in general these mediators worked as intended in the control group. However, the relationships among these mediators and outcomes differed by substance use outcome.

Although there is evidence that school-based substance abuse prevention programs are effective at reducing youth substance abuse in general (Botvin et al., 1995, Ellickson et al., 2003, Pentz et al., 1989, Tobler et al., 1999), information as to what is working and for whom it is working provides a more inconsistent perspective. Early studies of these programs were criticized as not reaching diverse populations, particularly those adolescents that were considered to be particularly vulnerable to substance use. Several researchers have addressed these criticisms demonstrating that prevention programs were effective with these subgroups (Flay et al., 1985, Botvin et al., 2001, Ellickson et al., 2003). However, there is also evidence that programs can be detrimental to these groups as well, leading to increased rather than decreased use of a variety of substances (Ary et al., 1990, Donaldson et al., 1995, White and Pitts, 1998, Peterson et al., 2000, Poulin et al., 2001, Werch and Owen, 2002, Skara and Sussman, 2003, Gorman, 2003, Wiehe et al., 2005, Thomas and Perera, 2006).

Several prevention researchers have conducted mediational analyses examining what factors targeted by an intervention were related to substance use outcomes, when the programmatic impacts were positive. However, although inconsistent and negative findings have been reported in general or among subgroups, few researchers have gone the “next steps” to determine what factors in the intervention may directly or indirectly be related to these negative outcomes. When such analyses are utilized, investigations of the effects of the treatment on the program's targeted mediators can reveal where the program is having unintended consequences and can provide the developers with important information for revision. In order to understand the “treatment process,” such mediational analyses are warranted not only when the program is successful, but also particularly when it has no impact or an iatrogenic impact. Thus far, such investigations of mediation, when programs have negative effects, are rare.

The analyses presented in this paper follow the main effects analyses of the Take Charge of Your Life (TCYL) evaluation discussed in detail in a companion article in this journal issue (Sloboda et al., 2009) and draw upon the mediational analysis of the conceptual foundation for TCYL that was conducted by Stephens et al. (2009) also appearing in this journal issue. Utilizing multilevel logistic regression models, Sloboda et al. found that TCYL had significant positive effects on past 30 day use of marijuana in the eleventh grade, for adolescents who used marijuana in the seventh grade (baseline users). In addition, Sloboda et al. found significant negative effects of treatment on past 30 day use of cigarettes and alcohol in the eleventh grade, for adolescents who did not use cigarettes and alcohol in the seventh grade (baseline nonusers). Two important and thus far unanswered questions are why the program had these differential impacts on baseline users and nonusers and to what extent these outcomes are related to exposure to TCYL. Information regarding the intervention and its delivery is provided in our companion piece (Sloboda et al., 2009).

Stephens et al. detail and test the conceptual framework that underlies the Take Charge of Your Life (TCYL) program. Specifically, Stephens et al. report the TCYL program was intended to effect social skills (refusal skills, communication skills, and decision-making skills), and social influences (normative beliefs, perceptions of consequences, and attitudes toward substance use). These social skills and social influences are measured following the ninth grade intervention (during the ninth grade post-test). Stephens et al. argue that these social influences and social skills should, over time, reinforce intentions not to use substances (measured in the tenth grade follow-up). These intentions not to use substances should then predict refraining from use in the eleventh grade (Stephens et al., 2009). Their analyses, tested in the control group, suggest that overall, the conceptual model appears to be supported by data, although the fit of the model varies by outcome of interest (alcohol versus cigarettes and marijuana).

Section snippets

Description of study population

The data for these analyses come from the Adolescent Substance Abuse Prevention Study (ASAPS) a national randomized controlled trial of the TCYL substance abuse prevention curriculum delivered when students were in the 7th and 9th grades. Eighty-three school clusters consisting of a high school and all of its feeder middle schools were randomly assigned either to receive the TCYL program (n = 41) or to a control condition (n = 42). These school clusters were selected from six metropolitan areas

Results

As proposed by the program model tested by Stephens et al., the treatment impact should be primarily on the program's targeted mediators (normative beliefs, consequences, and refusal, decision-making and communication skills). These analyses partially replicated Sloboda et al. findings of significant effects of the treatment variable on cigarette use (ES = .183; p = .021) and alcohol use (ES = .213; p = .016) for the baseline nonuser groups (see Table 1); these effects are not in the expected

Discussion

In this paper we attempted to uncover the mechanisms by which the TCYL program had negative impacts on baseline nonusers of tobacco and alcohol and positive impacts on baseline marijuana users. Our results suggest that the treatment had an impact on marijuana normative beliefs and refusal skills and these in turn impacted behavioral intentions not to use marijuana, which ultimately impacted marijuana use in the eleventh grade. In contrast, the TCYL program maintained direct effects on alcohol

Conflict of interest

All authors declare that they have no conflicts of interest.

Role of the funding source

Funding for ASAPS was provided by the Robert Wood Johnson Foundation under Grant Nos. 039223 and 040371. The Robert Wood Johnson Foundation had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.

Contributors

Each author contributed in many ways to this article.

Brent Teasdale had the lead as primary author of the paper and conducted the analysis.

Zili Sloboda, Richard C. Stephens, and Peggy Stephens shared responsibility for the design and conduct of the study; the collection, management, analysis, and interpretation of the data; and preparation, review, and approval of the manuscript.

Scott Grey was consulted on the analyses and interpretation of the data; and preparation, review, and approval of the

References (26)

  • P.L. Ellickson et al.

    New inroads in preventing adolescent drug use: results from a large-scale trial of project ALERT in middle schools

    Am. J. Public Health

    (2003)
  • B.R. Flay et al.

    Are social-psychological smoking prevention programs effective? The Waterloo Study

    J. Behav. Med.

    (1985)
  • A. Goldstein et al.

    Skillstreaming the Adolescent: New Strategies and Perspectives for Teaching Prosocial Skills

    (1997)
  • Cited by (15)

    • Evaluation of substance abuse prevention and treatment programs

      2017, Research Methods in the Study of Substance Abuse
    View all citing articles on Scopus
    View full text