(p. 459) Prevention of Substance Use Disorders
(p. 460) Extensive epidemiologic research has identified a set of interrelated problem behaviors, typically originating during childhood and adolescence, that are critically important from a public health standpoint. For youth, central among these risk-related health behaviors are alcohol, tobacco, and other drug use (Substance Abuse and Mental Health Services, 2015). Prevalence rates of alcohol, tobacco, and marijuana use among adolescents remain high. For example, recent prevalence of lifetime alcohol use among 12th-graders was 64%; for cigarette use the lifetime rate was 31%, and for marijuana it was 45% (Johnston, O’Malley, Miech, Bachman, & Schulenberg, 2016a, 2016b). Early initiation and progression in the use of addictive substances is associated with a wide range of externalizing behaviors, including conduct disorder (CD), risky sexual practices, and behaviors that result in unintentional and intentional injuries (Iacono, Malone, & McGue, 2008; Taylor et al., 2010). Thus, legal and moral implications aside, adolescent substance abuse must be regarded as a public health issue. The effective prevention and treatment of adolescent substance abuse, like that for any public health problem, requires a clear understanding of causes and the context in which these causes operate.
The review of the epidemiology and neurobiological underpinnings of drug use in Chapter 17 provides important background for the study of prevention. This research has uncovered important determinants of drug use that can guide interventions to prevent the uptake and progression of this behavior. In this chapter, we review the major theories that have guided the development of interventions to reduce drug use in adolescence and examine the evidence that has accumulated regarding the success of these efforts. We then describe major directions for the future development of effective prevention programs.
Theoretical and Conceptual Models of Prevention and Change
There is no generally agreed-upon theoretical or conceptual model for prevention of substance abuse. What is clear is that the uptake of drugs during adolescence is a product of a wide array of individual, social, and cultural factors that affect adolescent interest in and access to drugs. This is evident in the long-term trends for recent alcohol, cigarette, and marijuana use among adolescents as tracked by the Monitoring the Future study (MTF) and shown below. Since the mid-1970s, recent use of all three drugs has decreased, and quite dramatically for alcohol and cigarettes (Johnston et al., 2016a, 2016b). Although marijuana use has also declined, it is now more prevalent than cigarettes, reflecting a cultural change in the United States of greater acceptance of this drug among adults (see Chapter 17 and Fig. 19.1). Associated with these declines are changes in perceptions of the harmfulness of these drugs, with the correlations between aggregate perceptions of harm and recent use of alcohol (−.83), cigarettes (−.80), and marijuana (−.59) quite high. Over this period, federal education policy has encouraged schools to educate adolescents about drugs, and these programs have proliferated (Kumar, O’Malley, Johnston, & Laetz, 2013; Ringwalt et al., 2011). Media campaigns to discourage the use of tobacco (National Cancer Institute, 2008) and other drugs have also expanded (Ferri, Allara, Gasparrini, & Faggiano, 2013). In addition, the age at which adolescents can purchase alcohol was raised to 21, which researchers believe has reduced the use of and the number of injuries associated with this drug (DeJong & Blanchette, 2014). Although there have been massive efforts over this period to educate adolescents and the public about the harms of alcohol and tobacco, how these perceptions and behavior changed is not well understood.
Risk and Protective Factors: The Complex Causes of Adolescent Substance Abuse
It has been recognized for over 40 years that the risk for becoming a substance abuser is not equally distributed in the population. Originally this observation came from research that followed children into (p. 461) adulthood and used childhood demographic and psychological data to uncover pathways to an adolescent or adult substance abuse disorder. Subsequently, this view was bolstered by epidemiologic surveys in the United States that revealed that only 27% of individuals who have experimentally used drugs six or more times actually progress to become daily drug users, and only about a half of young-adult daily drug users go on to develop a drug abuse or dependence disorder (Robins & Regier 1991). While it is possible that chance plays a role in the acquisition of a substance abuse problem, it is more likely that the complex interplay of risk and protective factors determines who progresses from experimentation to regular use and from regular use to problematic involvement. Furthermore, this interplay of risk and protective factors exists in a maturational context such that at some stages of human development certain biological, psychological, or social factors may be totally benign, while at other stages of development these same factors may confer considerable risk for problematic involvement with drugs of abuse. These risk factors are subject to effects of gender and ethnicity, so risk factors may operate differently in boys and girls, and in different ethnic groups. To further complicate the issue, individual risk and protective factors must be viewed against a backdrop of laws, cultural and social norms, drug availability, economic circumstances, and regional and community factors. For example, a white adolescent male living in the United States who has a variety of individual risk factors for alcohol dependence might develop alcohol problems, but if he were raised in Saudi Arabia (where drinking alcohol is forbidden), it is less likely that he would develop an alcohol problem. However, it is possible that these risk factors might manifest themselves in other forms of problematic behavior (e.g., aggressive behavior). Thus, substance abuse is a multifaceted problem.
Geneticists refer to multidetermined problems like substance abuse as “complex disorders” because a multiplicity of individual biological and behavioral factors interact with environmental factors (e.g., social and societal phenomena) in complicated ways across human development to produce different outcomes. To (p. 462) the best of our knowledge, there is no single cause of adolescent substance abuse, so it is unlikely that there will be a single preventive measure to forestall its development. For this reason, the reader is cautioned to be skeptical of overly simplistic causal explanations for our substance abuse problems and facile and obvious solutions. The likelihood that approaches guided by conventional wisdom will achieve their promised results is diminished by the realities of our current understanding of the complex pathways to a substance use disorder.
Influential Theories of the Development of Adolescent Substance Abuse
Theories develop as an effort to summarize and explain research data generated by observation and experimentation. Theories are used to organize future research studies that ultimately test the validity of the original theory and provide an opportunity for it to evolve and undergo revision. Thus, theories are scientific “works in progress.” Several influential theories have guided our understanding of the origins of adolescent substance abuse and provide a framework for ongoing research in this area. These theories also provide a useful structure to guide approaches to the prevention and treatment of adolescent substance abuse problems. The following are among the most influential of these theories. There are many areas of commonality and overlap, yet each has contributed and advanced our understanding of the origins of substance abuse.
The “Gateway” or Stage Theory
This theory comes from epidemiologic research that has examined the patterning of alcohol and other drug use progression among adolescents. However, this theory has become a battleground for those both for and against the decriminalization of marijuana. The theory is based on the delineation of four stages in the sequence of involvement with drugs. The original findings suggested that surveyed adolescents engage in use of either alcohol or cigarettes (as legal and culturally accepted drugs) and then progress to marijuana, and then on to other illicit drugs, such as heroin and cocaine. The legal drugs are necessary intermediates between nonuse and marijuana. Thus, the use of tobacco, alcohol, and marijuana by adolescents was viewed as a crucial step or “gateway” to the use of other illicit drugs (Kandel, 1975; Yamaguchi & Kandel, 1984).
Why would tobacco and/or alcohol trigger marijuana use? It could be a general liability that increases the risk for any drug, with those most prevalent more likely to be initiated first. A secondary analysis of the 2008 MTF 12th-grade data aimed to check which drug (alcohol, tobacco, or marijuana) was the actual “gateway” drug leading to additional substance use among a nationally representative sample of high school seniors. Results indicated that alcohol represented the “gateway” drug leading to the use of tobacco, marijuana, and other illicit substances (Kirby & Barry, 2012). Or, the route of administration could predispose. According to a Dutch study, early-onset tobacco use does not pose a significantly higher risk of initiating cannabis use than early-onset alcohol use. Thus, the route of administration was not a factor. In this same study, early-onset comorbid use of both tobacco and alcohol was associated with a higher likelihood of initiating cannabis use than in adolescents who had tried either tobacco or alcohol. The authors concluded that the gateway hypothesis was not enough to explain their finding, again favoring the general liability model (van Leeuwen et al., 2011). An interesting study that examined East Asian youth with a genetic variation that yields a deficiency in an enzyme that metabolizes ethanol found that they were just as likely to progress to other drugs as those without the deficiency (Irons, McGue, Iacono, & Oetting, 2007). Thus, being unable to tolerate and use alcohol does not appear to influence the progression to other drugs. Furthermore, surveys of American high school students suggest that by 12th grade, 44.7% of students have tried marijuana, while only 0.8% have tried heroin and 4.0% have tried cocaine (Johnston et al., 2016a, 2016b). The discrepancies in these prevalence rates indicate that although illegal drug users (p. 463) may have started with marijuana, it is clear that marijuana use does not invariably progress to adolescent use of other illegal drugs.
Importance of Age of Initiation
Although the gateway theory has remained controversial, a less controversial aspect of this theory deals with age of initiation of experimentation with drugs of abuse (whether alcohol, tobacco, marijuana, or other illegal drugs) and the timing of stages of regular use and problematic involvement. The literature converges around the observation that the earlier the onset of progressive substance use, the greater the likelihood of problematic involvement later in development (Choi et al., 1997, 2001; Kandel & Logan, 1984; Schuckit & Russell, 1983; Yamaguchi & Kandel, 1984). However, this pattern is also consistent with a general liability theory described below. For example, early appearance of any externalizing behavior increases the risk for later antisocial behavior (McGue & Iacono, 2005). Khurana et al. (2015) also showed that impulsive tendencies reflective of poor executive function predict the progression rather than initiation of drug use in early to middle adolescence. Progression would seem to be the critical factor rather than early initiation, because without progression, there is little chance for addiction or other adverse consequences. Thus, for this reason, substantial effort has been placed on prevention interventions that delay the initiation and progression of substance use during early adolescence.
General Liability for Drug Use and Other Externalizing Behavior
Opponents of the gateway theory suggest that if there were a risk factor that was common to alcohol and other drugs, it could easily account for the relationship between supposed gateway drugs and other drug use. Examples of a theorized “third factor” include the genetic predisposition to drug use, a predisposition toward adolescent risk behavior in general, or shared opportunities to obtain both marijuana and other drugs (Morral, McCaffrey, & Paddock, 2002). The Problem Behavior Theory proposed by Jessor and Jessor (1977) is an early example of a third-factor explanation. They suggested that adolescents differ in a general tendency toward antisocial behavior, perhaps in response to stressors or deviant peers. The theory also posits that substance abuse for some adolescents may be a maladaptive means to cope with the stresses and social pressures that are characteristic of the adolescent stage of development. This theoretical perspective suggests that prevention interventions that offer alternative means of coping and social adaptation might reduce adolescent substance use behavior.
Considerable evidence reviewed by Iocono, Malone, and McGue (2008) suggests that a liability is present for the emergence of a wide range of antisocial behavior in children and adolescents. They characterize this liability as a behavioral disinhibition reflecting lack of control over impulses, which is common in children and adolescents who exhibit such externalizing behavior. Evidence for this pattern is present prior to adolescence, with children showing early signs of disinhibition exhibiting enhanced risk for later drug use (Elkins, McGue, & Iacono, 2007; Moffitt et al., 2011; Wong et al., 2006). More recent research suggests that weakness in executive functions, such as working memory, is an important component of the impulsive tendencies that predict the use of alcohol (Khurana et al., 2013) and other drugs (Khurana et al., 2015).
Consistent with the general liability model of adolescent drug use is evidence for shared genetic risk in twin studies of adolescent drug use (Iocono, Malone, & McGue, 2008). This pattern is further supported by studies showing high heritability of more general externalizing behaviors, including drug use, CD, and other antisocial behavior (Iocono, Malone, & McGue, 2008). Although there are gender differences in the display of these behaviors, the genetic liability appears to be similar for both males and females (Hicks et al., 2007; Kendler, Prescott, Myers, & Neale, 2003). Internalizing tendencies, such as depressed mood and anxiety, are also correlated with externalizing behavior in adolescents and thus are a potential co-occurring liability for drug use (Kreuger & Markon, 2006).
(p. 464) While research clearly reveals that genes are an important determinant of liability for substance abuse, it does not tell us which genes. For other complex traits like high blood pressure, diabetes, or high cholesterol, it is clear that there are multiple genes involved and multiple genetic and biological pathways are involved in producing disease. It is unlikely that there is a single gene for alcohol, cocaine, or nicotine dependence. There may be hundreds or thousands of genes in a given pattern producing risk, and that risk may only be present in a given environmental context. The nature of the genetic risk may be a common factor for abuse across a wide variety of drugs, or a genetic risk for conduct difficulties or problem behaviors, or a set of genes that delay the maturation of the brain so one is less able to control the habituating effects of drugs. The effects of genes may be protective rather than associated with risk, and what we think of as genetic effects producing substance abuse may actually be the absence of protective genes. There is good evidence that specific genetic mutations protect against the development of alcohol use disorder in certain ethnic groups, and some evidence that there is a mutation that protects against smoking. There is some evidence that substance use heritability changes across development. It seems that genetics plays a smaller role for substance initiation, but might gain more importance for the progression of drug use (Meyers & Dick, 2010).
In sum, the finding that deficits in inhibitory control and related liabilities are a primary source of risk for drug abuse and other externalizing behavior has been a focus of several interventions that attempt to enhance self-regulatory skills in children and adolescents. A prominent example of this approach is embodied in theories regarding the role of the family in enhancing such skills in at-risk children.
Patterson’s Developmental Theory
Patterson’s theory was originally proposed to explain the development of juvenile delinquency; consistent with the observation that problem behaviors frequently co-occur in adolescents, it has also been used to understand and address problematic involvement with alcohol and other drugs of abuse. Patterson and colleagues (Dishion et al., 1991; Patterson, DeBaryshe, & Ramsey, 1989) proposed a developmental theory of CD and related externalizing behavior that posits that adolescent problem behavior is a consequence of poor parental family management practices interacting with the child’s own aggressive and oppositional temperament. Here, temperament refers to the early and genetically determined behavioral characteristics that over time lead to externalizing behavior. Deficits in parenting skill, such as harsh and inconsistent punishment, increased parent–child conflict, low parental involvement, and poor parental monitoring result in poor behavior and performance in school. The poorly performing and behaving child may be socially rejected by many peers, but he or she may form friendships with other problematic children. This process of forming close peer relationships is augmented by the negative interactions with caregivers in the home. The impact of CD on further drug addiction outcomes is well documented, with CD conferring additional risk for illicit drug use (Hopfer et al., 2013; Sung, Erkanli, & Costello, 2012).
As the child affiliates with more deviant children, he or she adopts deviant behavior as a norm. Other deviant children become powerful social role models, from whom the child learns further deviant and socially unacceptable behavior, including experimentation and progression in the use of drugs of abuse. These children may therefore be viewed as being on a developmental trajectory of deviancy and substance abuse that begins early in development and is compounded by unskilled parenting and the formation of social relationships with other problem children (Vuchinich, Bank, & Patterson, 1992).
The developmental cascade model shown in Figure 19.2 illustrates the influences featured in Patterson’s theory. In this model, children who show early signs of externalizing behavior move on to experience school failure and attraction to other antisocial children that is exacerbated by poor family management practices. As the child moves into adolescence, the problems (p. 465) escalate into various forms of unhealthy behavior such as drug use. Prevention interventions that are based on this theoretical approach offer parenting skill training to teach parents more effective ways to discipline and monitor their children, and to reduce the negative environment of the home. Tutoring and other forms of education support may be provided to reduce academic failure. Social skills training may also be offered the child in order to reduce peer rejection and provide a mechanism to gracefully resist peer pressure to use alcohol and illicit drugs. These interventions are described in greater detail below.
Summary of Factors Influencing Adolescent Drug Use
While adolescents in the United States are widely exposed to a spectrum of drugs of abuse, research suggests that adolescent substance abuse problems are due to multiple factors. Most theories suggest that genetic, psychological, familial, and nonfamilial environmental factors interact in a complex way to determine an adverse or protective outcome. Thus, genes, temperament, attitudes and beliefs, family environment, peer affiliation, and social norms all influence the relationship between the individual and a substance use disorder outcome. The developmental timing of these factors adds an additional level of complexity. The question of “nature or nurture” has been rendered moot, primarily by research conducted over the last 20 years: it is clear that both “nature” and “nurture” are involved, set against the backdrop of child development (Kendler, Jaffee, & Romer, 2011). Thus, there is no single cause of adolescent substance abuse, and any single prevention approach is unlikely to have broad universal success.
The general liability approach does help us to identify high-risk children for prevention interventions. Clearly, offspring of parents with substance abuse problems are themselves at significant risk for becoming substance abusers. Interventions that improve parenting practices may be important, not only in instilling appropriate disciplinary practices in the parents of high-risk children, but also by enhancing parental involvement and monitoring. Social skills training may keep high-risk children from being rejected by less deviant peers and thereby avoid being attracted to more deviant peer groups. A wide range of personality characteristics that have been linked to drug use (p. 466) progression may also provide guidance for more targeted prevention (Audrain-McGovern & Tercyak, 2011). For example, an intervention designed to reduce drug use was found to reduce the risk that a gene conferred to substance use (Brody, Yu, & Beach, 2015).
Other important liabilities include two facets of impulsivity, acting without thinking and delay discounting (Audrain-McGovern et al., 2009; Khurana et al., 2013; Winstanley et al., 2010). Acting without thinking is a phenotypic expression of impulsivity linked to early involvement in a wide range of externalizing behavior, including drug use, gambling, and fighting (Romer et al., 2009). Delay discounting is a related but separate expression of impulsivity characterized by preference for small immediate rewards over larger but delayed rewards (Shamosh et al., 2008). It has been linked to progression in use of various drugs in adolescents, including alcohol (Khurana et al., 2013) and cigarettes (Audrain-McGovern et al., 2009). Prevention programs that target these specific liabilities have yet to be developed, although universal programs described below have components that can help to constrain these liabilities by enhancing decision making and self-regulation skills.
Types of Prevention Interventions and Model Programs
Prevention programs are often categorized according to the following definitions based on the audience they are designed to reach (Haggerty & Mrazek, 1994).
Universal intervention programs are designed to reach the general population, such as all students in a given school or school district, through school education or media campaigns, for example. Broadly speaking, universal interventions represent the most widely used approach to drug abuse prevention. A national survey of school administrators revealed that over 90% of middle and high schools report delivering some type of universal program to students (Kumar et al., 2011). From a universal intervention perspective, public health problems and their solutions are inextricably a part of the community social system; solutions are essentially universal, with some types of universal interventions facilitating access to higher-risk groups within the community that may warrant more intensive intervention. Implementation of these types of interventions is typically supported by local community partnerships or coalitions.
Selective and indicated interventions specifically target persons identified as at risk of drug abuse, such as those who might already have used a drug and therefore are at higher risk of progression than those not having tried one. Examples of such programs, described below, include programs that screen for youth who have used alcohol and then deliver brief therapeutic interventions to prevent progression in use. Indicated interventions focus more clearly on youth who show signs of serious mental health conditions, such as drug abuse, delinquency, or depression, but have not been diagnosed with a substance use disorder. These programs are more intense than either the universal or selective interventions.
In our review of drug prevention programs, we highlight those classes of interventions that have received the most extensive evaluations, as evidenced by Cochrane or other meta-analyses. Most of these evaluations have appeared since the first edition of this volume. Not surprisingly, the majority of interventions have been universal in scope. In some approaches, universal interventions are used to identify higher-risk youth who are then provided with opportunities to receive more intensive programming.
School-Based Universal Interventions
The most widely adopted universal program in schools is the well-known Drug Abuse Resistance Education (DARE), a school-based primary drug prevention curriculum designed for introduction during the last year of elementary education (Kumar et al., 2013). Despite its popularity, early evaluations of DARE failed to demonstrate its effectiveness (Clayton, Cattarello, & Johnstone, 1996; Lynam et al., (p. 467) 1999). In response, an enhanced version of DARE was developed and tested, DARE Plus. Additional components added to the original DARE curriculum include a peer-led parental involvement classroom program called “On the VERGE,” youth-led extracurricular activities, community adult action teams, and postcard mailings to parents. Evidence suggests that DARE Plus program delivered over two years in seventh and eighth grade produced significant reductions in alcohol, tobacco, and polydrug use among boys but had no effect on girls (Perry et al., 2003). Other versions of the DARE program have evolved, with the adoption of the “Keepin’ it REAL” intervention. This program uses interactive exercises that develop skills to Refuse, Explain, Avoid, and Leave situations that encourage drug use. Evaluations of this universal program indicate that it is particularly effective in reducing the uptake of alcohol use (Hecht et al., 2003). It is also effective in reducing the progression of alcohol use among those who have already tried the drug (Kulis, Nieri, Yabiku, Stromwall, & Marsiglia, 2007).
A Cochrane Review of all universal school-based drug prevention programs indicated that programs with an emphasis on decision making, drug resistance, and self-management skills, such as featured in Botvin’s life skills training program (Botvin et al., 1995), have shown efficacy, especially in regard to alcohol use (Foxcroft & Tsertsvadze, 2012). A life skills training program implemented in Germany in fifth grade, with boosters in sixth and seventh grades, found reductions in use of alcohol, cigarettes, and illicit drugs over a 2-year follow-up, with effect sizes ranging from 0.34 to 0.44 (Weichold & Blumenthal, 2016). A program delivered in the early elementary school years, the “Good Behavior Game,” was also found to be a surprisingly effective program for reducing externalizing behavior in high school, including drug use (Kellam, Reid, & Balster, 2008). Although delivered as a universal program in high-risk schools, this program seems to have its effects by reducing impulsive behavior in children most at risk of later externalizing problems.
A review of school-based prevention programs for illicit drugs conducted by Faggiano et al. (2008) concluded that skill-based programs such the life skills training program were most successful in reducing the use of these drugs, with reductions on the order of 20% in drug use progression. A review of school-based drug education programs focusing on marijuana use found an overall effect size of 0.58 in reduced use, which amounted to a success rate of 28% compared to controls (Porath-Waller, Beasley, & Beirness, 2010). This review also suggested that programs delivered to youth older than 13 were more effective than those for younger youth. A study of potential drug prevention mediators in over 7,000 U.S. students assessed from ninth to 11th grade indicated that apart from attitudes and beliefs about the harms of drugs, youth with better problem-solving skills were more likely to report reduced drug use of alcohol, cigarettes, and marijuana in 11th grade (Stephens et al., 2009). This study supports the strategies employed in the more successful universal school-based drug prevention programs that emphasize skills as well as beliefs and attitudes as targets of intervention.
A more recently developed approach that can potentially reach large audiences without using limited school resources is drug prevention programming delivered over the Internet. A review of these programs found some encouraging evidence of effects on reduction in marijuana use (Tait, Spijkerman, & Riper, 2013). These programs have been tested with adolescents and found to have an effect size of 0.17; however, the follow-up periods have been short. Although the effects are smaller than more intense school-based programs, these interventions are still in the early stages of development and may be a useful approach to consider for further refinement.
Mandatory Random Drug Testing in Schools
Another school-based approach to drug use prevention that has gained popularity in recent years is the use of mandatory random drug (p. 468) testing among students involved in sports and other extracurricular activities. These programs require students who wish to play sports or participate in clubs and organizations to agree to be tested at random with biological assays during the school year. It is estimated that approximately 13% of schools have adopted some version of this policy (CDC, 2015). Although this policy has been challenged in the courts, the U.S. Supreme Court has ruled on two occasions that schools are not barred from implementing such policies. Students who screen positively are referred to treatment and are often excluded from participating in school activities for the school year. Only two trials have been conducted to test this strategy (Goldberg et al., 2007; James-Burdumy, Goesling, Deke, & Einspruch, 2012), and the results have been largely negative. In the one trial that found reductions in reported drug use in the past 30 days (James-Burdumy et al., 2012), there was no evidence that the intervention reduced attitudes or intentions toward drug use. A large analysis of MTF data over a period of 14 years (Terry-McElrath, O’Malley, & Johnston, 2013) found that students in schools that employed drug testing reported less use of marijuana but increased use of other drugs that were not subject to testing. Other surveys have found that schools with better social climates had less drug use but that, apart from this school characteristic, testing was not associated with reduced use of tobacco, alcohol, or marijuana (Sznitman, Dunlop, Nakkur, & Romer, 2012; Sznitman & Romer, 2014).
Environmental Universal Interventions
Another class of universal interventions implements policies that attempt to influence the social and legal environment of entire communities, including schools, parents, drug dispensaries (e.g., bars, shops, and other retail outlets), and police regarding legal purchase age and other restrictions on drug use, such as driving under the influence of alcohol. We first review some environmental strategies that have been implemented through statewide policy and then examine efforts that include greater enforcement of restrictions on drug access and use in local communities.
Mass Media Campaigns
Mass media campaigns represent an environmental approach sometimes combined with school-based programs. The most frequent uses of mass media involve campaigns to reduce the uptake of cigarettes. These programs typically highlight the harms of drug use and the benefits of cessation for those who have already initiated. A Cochrane Review of seven programs that met the best standards of evaluation found evidence to support their efficacy (Brinn, Carson, Esterman, Chang, & Smith, 2010). However, such programs are difficult to evaluate and the evidence was regarded as weak in relation to programs delivered in typical randomized clinical trials. A longstanding campaign implemented in California appears to have had success (Pierce, White, & Gilpin, 2005), and indeed California has among the lowest rates of adolescent smoking in the United States (California Department of Public Health, 2015). Campaigns that raised awareness about the ways that the tobacco industry has misled young people about the hazards of smoking have had success (The Truth campaign). This message, which has been tested both in national samples by the American Legacy Foundation (Thrasher et al., 2004) and in selected states (e.g., Sly, Heald, & Ray, 2001), appears capable of reducing the uptake of smoking in adolescents. Nevertheless, these evaluations relied on quasi-experimental designs that do not meet the highest standards of evidence. With the passage of the Family Smoking Prevention and Tobacco Control Act of 2009, the U.S. Food and Drug Administration, in collaboration with the U.S. Centers for Disease Control and Prevention, has developed smoking cessation media campaigns directed toward both adults and adolescents. Initial results of these efforts in regard to adult outcomes suggest that the campaigns encourage quitting (McAfee et al., 2013). However, effects on youth initiation have not as yet been evaluated.
Other programs have been delivered to reduce the uptake of illegal drugs, such as (p. 469) marijuana. A Cochrane Review of these programs also found some evidence of efficacy; however, the evidence was not consistent (Ferri et al., 2013). The National Youth Anti-Drug Media Campaign, supported by the U.S. Office of National Drug Control Policy (ONDCP), was intensely evaluated. From September 1999 to June 2004, three nationally representative cohorts of U.S. youths ages 9 to 18 years were surveyed at home four times. Main outcomes were self-reported lifetime, past-year, and past-30-day marijuana use and related cognitions. Most analyses showed no effects of the campaign (Hornik, Jacobsohn, Orwin, Piesse, & Kalton, 2008). It was concluded that through June 2004, the campaign was unlikely to have had favorable effects and may have actually had delayed unfavorable effects. In particular, exposure at round three predicted marijuana initiation at round four, potentially because the campaign increased perceptions that peers were using marijuana and that its use was appealing despite the risks.
The failure of the ONDCP media campaign led to the development of a new program that is currently sponsored by the nonprofit Partnership for Drug-Free Kids. This program features a message that encourages youth to live “Above the Influence” of illicit drug use. The campaign does not highlight peer use of drugs but rather features the importance of being in control of one’s life. An evaluation of the program using a randomized community design indicated reductions in use of marijuana over a 2-year follow-up period (Slater, Kelly, Lawrence, Stanley, & Comello, 2011). The study was also able to evaluate the concurrent national campaign using the same message strategy. Exposure to both programs was found to produce less use of marijuana. In addition, the message that drug use interferes with personal life aspirations was found to mediate campaign effects.
Raising Drug Prices
Making drugs more costly is another universal approach that can affect adolescent drug use. This approach is most amenable to intervention for legal drugs, such as tobacco and alcohol, for which taxes can be applied that increase the cost to the consumer. Price interventions for cigarettes vary considerably across states, and thus evaluations can be conducted to assess their success. These studies tend to find that price does affect adolescent uptake of cigarettes (Huang & Chaloupka, 2012), although there is some dispute about the breadth of the effect (Fletcher, Deb, & Sindelar, 2009). Price effects on alcohol are similar, but efforts to raise taxes on this product have been much less successful (National Research Council, 2004). Indeed, taxes on alcohol have remained relatively low in comparison to inflation in the United States.
Age of Purchase
Legal age of purchase of alcohol has been aggressively pursued as official policy since 1984, when the U.S. Congress voted to withhold federal funds for highway construction in states that did not adopt 21 as the legal age for purchase of alcohol. This policy has been credited with reducing motor vehicle crashes attributable to alcohol in adolescents (DeJong & Blanchette, 2014). However, it is noteworthy that the policy was adopted following an aggressive media campaign by Mothers Against Drunk Driving, which highlighted the dangers of alcohol use, especially in regard to driving. This campaign was also active in Canada, which also experienced a parallel reduction in adolescent motor vehicle crashes despite having legal purchase ages of 18 and 19, depending on the province (Hedlund, Ulmer, & Preusser, 2001). Thus, at least some of the effects of this policy are likely to be attributed to changes in societal views about the dangers of using alcohol, as reflected in the steady decline in alcohol use among adolescents illustrated in Figure 19.1. At the same time, rates of heavy drinking among college students have been lower in Canada than in the United States (Kuo et al., 2002), suggesting that the lower legal drinking age in Canada has not encouraged excessive drinking in this young transitioning adult group. Adolescent driving habits have also changed, with less reporting of driving while under the influence of (p. 470) alcohol (Terry-McElrath, O’Malley, & Johnston, 2014) and greater use of seatbelts (Carpenter & Stehr, 2008). These changes are also likely to be the result of societal influences apart from the age 21 law (Hedlund et al., 2001).
The legal age for purchase of cigarettes has been 18 in most states. However, efforts are under way at the time of this writing to raise the age to 21, just as for alcohol. Support for this policy also derives from research suggesting that it would deter the progression of cigarette use in adolescents, when most initiate the use of this drug.
Marijuana is now a legally purchased product in several states with the age of purchase set at 21.
Community Environmental Interventions
Community environmental interventions use various strategies to enforce restrictions on the sale of drugs to youth, especially alcohol and cigarettes, or to encourage responsible alcoholic beverage service as a way to reduce the adverse effects of alcohol use in bars and other retail outlets (Saltz, Grube, & Treno, 2015). Based on the evidence that statewide policies regarding purchase and access to alcohol and tobacco can reduce its uptake in young people, the various strategies employed in community environmental interventions attempt to increase the effects of these policies through greater enforcement and awareness of restrictions. In their review of these interventions, Saltz et al. (2015) note that the evidence that has accumulated over time finds that these programs can produce modest reductions in a variety of outcomes, including reductions in alcohol-related injuries and use of alcohol by young people. Here again, however, the studies that have evaluated these interventions tend to require quasi-experimental designs that do not meet the highest standards of evidence. In addition, the interventions tend to combine a variety of strategies, so it is not clear which ones worked or the theoretical processes that mediated the effects (Saltz et al., 2015). Nevertheless, the finding that such programs can reduce the adverse effects of alcohol use and tobacco uptake suggests that communities that wish to address youth access and use of drugs can do so with a variety of strategies.
Another important category of universal interventions are family-focused ones. Such programs are designed to strengthen bonds to family, school, and community and facilitate participant development of family-connected skills consistent with etiological research. Educating on the dangers of substance abuse is not the main focus, since, by this model, substance abuse is understood as having important family components, at several levels: parental monitoring, effective parent–child communication and disciplining, and improved conflict resolution. They have been adapted for various age groups and can be used as selective and indicated programs as well. An early version of this approach called the Supporting Families Program is illustrated in the causal model in Figure 19.3. As is evident, the family is only one component of influences on drug use in adolescents. In this social ecology model, family bonding is central to establishing effective supervision of the adolescent, which then influences family and peer norms for drug use. Family bonding also encourages greater attachment to school, which enhances norms against drug use. All of these influences are embedded within the larger community, which also affects peer and family norms. Kumpfer, Alvarado, and Whiteside (2003) argued, based on a literature review at the time, that family-focused interventions are up to nine times more effective than school-based programs in reducing alcohol use. A Cochrane Review found that a majority of family-based programs reduced alcohol use in adolescents (Foxcroft & Tsertsvadze, 2012). A review of the effects of various universal as well as selective family-based interventions in regard to marijuana and other illicit drugs found that these programs have some success in reducing the initiation of marijuana use (Vermeulen-Smit, Verdurmen, & Engels, 2015). However, evidence of success in reducing the uptake of other illicit drugs was less robust.
(p. 471) Spoth, Trudeau, Guyll, and Redmond (2009) have produced a shorter version of this intervention that has been rigorously evaluated and found to be effective, the Strengthening Families Program for Parents and Youth Ages 10–14 (SFP 10-14). Implementation of the SFP entails seven weekly sessions. The SFP has separate sessions for parents and children that run concurrently for 1 hour and focus on skills building. During the second hour parents and children participate together in a joint hour-long family session, during which they practice the skills learned in their separate sessions. The family session affords the opportunity for higher-risk families and those with special needs to identify available services. In addition, the family session includes activities designed to encourage family cohesiveness and positive involvement of the child in family activities. For the parental sessions, the essential content and key concepts of the program are also presented on videotape. Further detail regarding the SFP is provided in a recent review (Spoth, Redmond, Mason, Schainker, & Borduin, 2015). A 10-year follow-up of one test of the intervention found reduced levels of illicit drug use at age 21 mediated by reduced progression of drug use produced by the intervention (Spoth, Trudeau, Guyll, & Shin, 2012).
Because family-based interventions have positive effects on family functioning, they can also reduce other outcomes that are associated with drug use, such as internalizing symptoms of depression and anxiety. In a trial that employed SFP approaches, Trudeau et al. (2007) found that the intervention reduced internalizing symptoms, especially in girls, as well as polydrug use over a period of several years. Other crossover effects include reductions in risky sexual behavior and enhanced school performance (reviewed in Spoth et al., 2015). The growing evidence in support of SFP also includes cost-effectiveness analyses showing clear benefits for each dollar spent on the program (Spoth et al., 2015).
Some interventions merge family-focused approaches with other community programs, such as school-based interventions and media programs. An example of this type of expanded program is Project STAR (Pentz et al., 1989). (p. 472) This intervention model attempts to involve the entire community with a comprehensive school program, a mass media campaign, a parent program, a community organizing component, and a health policy change component. Project STAR has been shown to be effective in terms of reductions in drug use behavior in high school for those youth who began the program in middle school. Another noteworthy example is the program of university–community partnerships developed by Spoth et al. (2011). In this program, communities collaborate with university researchers to implement a menu of effective drug prevention interventions. A demonstration of this approach in 28 school districts in rural Iowa and Pennsylvania found sizeable reductions in drug use compared to controls over a 7.5-year period (Spoth et al., 2011, 2013).
Indicated Intervention Programs
One of the limitations of universal drug prevention programs for schools is their need for multiple sessions over periods of 1 to 2 years (Gottfredson & Wilson, 2003). The review by Porath et al. (2010) found that programs with fewer than 15 sessions were less successful in reducing use of marijuana. Another approach is to implement more concentrated and brief interventions (e.g., 5 hours) that target youth already using drugs. A review of these programs that focus on prevention of alcohol-use progression found that such programs, when delivered individually, can have a positive effect (Hennessy & Tanner-Smith, 2015). Ideally, such programs would be more feasible if they could be delivered in group settings; however, group interventions also can introduce iatrogenic effects resulting from peer modeling among high-risk youth (Dishion & Dodge, 2005). An approach that screens for and then treats higher-risk youth in schools is discussed in Chapter 31.
The Reconnecting Youth Program (Eggert et al., 1994, 1995) is another example of a school-based indicated intervention. This program is for adolescents in grades nine to 12 who show signs of poor school achievement and the potential to drop out. The program teaches skills to build resiliency toward risk factors and to moderate early signs of drug abuse. It consists of several components, such as a Personal Growth Class designed to enhance self-esteem, decision making, personal control, and interpersonal communications; a Social Activities and School Bonding Program to establish drug-free peer relationships; and a School System Crisis Response Plan to address suicide prevention. Evaluations of this intervention have documented only short-term benefits, with long-term studies yet to be done.
Indicated programs for older adolescents and young adults have been developed primarily to address heavy use of alcohol in college students. These programs employ Motivational Interviewing, which is typically delivered individually by a skilled interventionist (e.g., McCambridge & Strang, 2004). Toumbourou et al. (2007) identified these as effective interventions for this age group.
Applying Knowledge of Genetics to Reduce Drug Use
The enormous toll that drug dependence takes may also lead prevention experts to consider ways that genetic risk information might be used to identify high-risk subgroups of youth who might benefit from more intensive or tailored prevention approaches. As reviewed in greater detail elsewhere in regard to the prevention of adolescent smoking (Audrain-McGovern & Tercyak, 2011; Lerman, Patterson, & Shields, 2003; Wilfond et al., 2002), there are many ethical challenges and considerations regarding this strategy. From a scientific perspective, research on genetics and tobacco use is still in its infancy. There is no single “tobacco use gene,” as for any other drug, and as such, risk estimates will need to take into account multiple interactions among genetic, social, and psychological factors. Even considering genetic variants with widely validated effects on smoking behavior, these effects are likely to be small, and risk estimates will be highly probabilistic. Additional risks of genetic testing of adolescents include stigmatization, discrimination, and potential adverse psychological effects (p. 473) (Audrain-McGovern & Tercyak, 2011; Lerman, Patterson, & Shields, 2003).
Multilevel Intervention Programs
Multilevel intervention programs are typically combinations of the above intervention models. They include universal, selected, and indicated strategies gauged to the needs of the adolescent.
The Family Check-Up (Dishion & Kavanagh, 2000) is an example of a multilevel intervention designed to address the needs of families of young adolescents who present with a range of problem behaviors and diverse developmental histories, as illustrated in Figure 19.2. This ambitious program incorporates universal, selective, and indicated prevention components. The universal intervention is offered to all students in middle schools in the form of a school-based family management program. This intervention is a shortened version of Botvin’s life skills training program, with only six sessions. Families whose children need further attention are offered the selective intervention, which offers a more intense three-session assessment of family interaction patterns, with suggestions for ways to improve those interactions to support the at-risk youth’s healthier development. Motivational Interviewing techniques are used to encourage more adaptive family conflict resolution and parent engagement strategies to prevent the escalation of unhealthy behavior. The program provides a menu of services that includes a brief family intervention, school monitoring system, parent groups, behavioral family therapy, and case management services. In one study, the selective Family Check-Up intervention condition was related to increased levels of students’ self-regulation skills from sixth to seventh grades, which in turn reduced the risk for growth in antisocial behavior, involvement with deviant peers, and alcohol, tobacco, and marijuana use through the eighth grade (Fosco, Frank, Stormshak, & Dishion, 2013). A 10-year follow-up study showed that the selective intervention reduced the progression in use of alcohol, tobacco, and marijuana through age 21 (Veronneau, Dishion, & Connell, 2014).
In a review of this program of research, Dishion et al. (2015) emphasized the value of their multilevel approach to drug use prevention for its effectiveness, efficiency, and cost. Families with the lowest-risk children receive only the universal intervention, which saves on resources and time without compromising the benefits of this program. Greater resources are devoted primarily to families with greater needs, and these families show the strongest preventive effects. Thus, the program is able to allocate attention to the children who would otherwise not receive adequate preventive services in a universal-only program. Finally, the selective and indicated interventions are only provided to families who seek those services. However, experience in the program suggests that families recognize the need for those services and willingly accept them. Finally, the program can be tailored to the ethnic background of the families, thus increasing their effectiveness and acceptability.
Another example of a multilevel intervention with a family focus is the Triple-P (Positive Parenting Program) (Sanders, 2012; Sanders et al., 2014). This program employs a mass media intervention to reach and engage parents across the risk spectrum and to identify those at greater need for intervention. Up to five levels of selective and indicated interventions are delivered in various venues to parents (e.g., through face-to-face meetings with a practitioner, phone meetings, self-directed materials, or in small groups). An extensive meta-analysis involving over 16,000 children across the United States, Europe, and Australia found that each level of the program produced improvements in parenting behavior and child outcomes related to the emergence of externalizing behavior in adolescence (Sanders et al., 2014). The program has been replicated in poor urban neighborhoods as well as in larger jurisdictions, thus providing a model of prevention that can be easily brought to scale.
Future Challenges for Preventive Interventions
Since the first edition of this book, numerous efforts have been undertaken to identify and disseminate descriptive information about model preventive interventions. Reviews of (p. 474) these interventions typically include descriptions of selection criteria or rules of evidence applied and summaries of the intervention review process. Many also delineate salient characteristics of the types of programs that have proven to be effective. A major issue for the field is the variability in the rules of evidence and intervention selection criteria, with the level of scientific rigor applied varying considerably. A commonly expressed concern in the Cochrane Reviews is the absence of rigor in the evaluation of drug prevention interventions (Foxcroft & Tesertsvadze, 2012). There is concern regarding the presence of bias in many evaluations, with program designers being involved in the tests of their interventions. One review of early school-based drug prevention programs that were evaluated by researchers unconnected with the design of the programs found no evidence of efficacy (Flynn, Falco, & Hocini, 2015). While this is unlikely to explain all of the favorable evidence we have reviewed, it is a source of concern.
Another issue for future development of drug abuse prevention programs is determining the best ages to deliver programs in schools and for families (see Spoth, Greenberg, & Turisi, 2009, for an overview of interventions that target different age groups). Family interventions tend to focus on childhood and early adolescence. School-based programs also tend to be delivered in middle school, with follow-ups that only extend into the middle of high school. These programs were designed in part to prevent the early initiation of drug use. However, given that progression continues well into high school, it is important to deliver programs to youth at these ages as well. One review found that such programs tended to be more effective in reducing the use of marijuana, which is likely to emerge later in high school, than programs delivered in middle school. A family-based program using SFP principles delivered to older adolescents in the rural South was noted to be one of the few designed for older adolescents (Brody et al., 2012).
An issue often noted in the field of prevention research is the absence of theoretical models that distill the essential mechanisms underlying intervention efficacy. Interventions are designed to include multiple components that target important sources of efficacy. However, program effects are often evaluated with a focus on drug use endpoints without assessing the mediating mechanisms. This problem has been noted for the evaluation of tobacco cessation programs by Baker et al. (2011), with a call for greater attention to short-term mediating mechanisms that can help to determine the efficacy of intervention components as they proceed. For example, it is not clear across the various prevention approaches reviewed here what the mediating mechanisms are that underlie program efficacy and the time courses of their influence. Unless we understand those mechanisms, it will be difficult to determine the correct mix and strength of program components. Attempts to isolate such components, such as the Multiphase Optimization Strategy (MOST) of Collins et al. (2011), are important steps that could advance the evaluation of prevention programs. However, to achieve the full benefits of MOST, greater attention to effects on mediating mechanisms will be required.
Prevention professionals have broadened the target of their interventions, going beyond substance abuse to address the global quality of youth development. Advocates of positive youth development approaches emphasize that efforts to address public health concerns by preventing youth problem behaviors must be pursued in concert with youth-related health promotion goals. The need to integrate prevention and youth-related health promotion—or positive youth development—has emerged as a consequence of the observation that problem-free youth are not necessarily fully prepared youth (Pittman, 2000). A review of these programs is presented in Chapter 26.
A final challenge regarding effective interventions is the failure of schools and communities to employ evidence-based programs. An excellent source of evidence-based programs is available in the Blueprints registry (http://www.blueprintsprograms.com/). In addition, both the National Institute on Drug Abuse (NIDA) and the Substance Abuse and Mental Health Services Administration (SAMHSA) (p. 475) maintain lists of effective drug abuse prevention programs. However, schools adopt a wide range of programs that vary in effectiveness (Kumar et al., 2011). They may also resort to strategies that sound effective but have little evidence to support them (e.g., drug testing). The emerging field of implementation science is an attempt to identify the best ways to translate efficacious programs into acceptable and enduring practices in communities (Neta et al., 2015; Pas & Bradshaw, 2015; Spoth et al., 2013). A future challenge in the field of prevention science is to identify methods that encourage adoption of effective programs that can be implemented with fidelity in community settings.
Chapter 30 outlines recent advances in the implementation of evidence-based interventions (EBIs) for child and adolescent healthcare. It is striking how these EBIs have been aided by the need to enhance the cost-effectiveness of medical care across all provider systems. Although significant challenges remain, the successes achieved using Internet education to hasten the dissemination of EBIs to providers suggests that similar strategies will be helpful to prevention specialists. For example, although registries such as Blueprints identify effective programs, they do little to help users select the right program for their specific needs, and they do not always link to resources that can be used to train interventionists to implement the program faithfully. Based on our review of EBIs, it is clear that schools are a major venue for delivery of prevention programs and that health providers in those settings should be a target for the dissemination of EBIs. In addition, community-based prevention activities could be administered by county health officials who are responsible for promoting the health of their localities. A major challenge for the successful dissemination and adoption of EBIs will be the demonstration of return on investment, especially in regard to whatever funds are currently devoted to activities that better prevention practices would curtail (e.g., drug enforcement, school failure, and youth injury). There is some evidence that prevention and health promotion programs are cost-effective (see also Chapter 26). Greater efforts to make these programs attractive to local stakeholders will be a critical element in future efforts to increase adoption of effective prevention programs.
It is fitting to close this chapter on prevention by highlighting the need for a national youth development strategy. Planning for a comprehensive strategy to foster positive youth development and to prevent youth substance-related problems necessitates a sustained, well-organized effort, with inputs from a range of community interventionists, scientists, and policymakers at the state and federal levels. One potential contribution to this larger planning effort is a design for universities and communities to partner together to foster a higher prevalence of capable and problem-free youth (Spoth et al., 2011). However, the wide range of tasks for those involved in community–university partnerships, the many barriers to task accomplishment, and the limited resources available highlight the challenges to the design of developmentally appropriate preventive interventions. There are well-known predictors of risk for drug abuse and reliable screening instruments, and some early risk factors can be easily recognized based on clinical presentation, such as fighting or school absence. There are also effective multilevel programs, such as Triple-P, using a public health approach that can serve as a model for the effective delivery of preventive interventions. However, an integrated program of prevention activities, as a national policy, is still lacking.
Concerns about inadequate prevention programming are all the more urgent given trends in drug use that affect adolescents, a prime example being various forms of legalization of marijuana. Although this drug may be less harmful than alcohol, its legalization will create new challenges to prevention programs, which will have to deal with a newly legal but addictive drug (Caulkins, Hawken, Kilmer, & Kleiman, 2012). There are also challenges to (p. 476) the increased use of prescription drugs, with a lifetime prevalence of 18.3%, including narcotics that are often prescribed for the reduction of pain (8.4%) (Johnston et al., 2016a, 2016b). Adolescents are increasingly trying these drugs, sometimes with unfortunate consequences. Finally, there is the new method of delivering nicotine, the electronic cigarette, that is also being tried by adolescents (16.2% past 30-day prevalence) (Johnston et al., 2016a, 2016b). Although use of cigarettes has declined in recent years, overall experimentation with tobacco products still remains high, at around 26% (Arrazola et al., 2015). All of these emerging ways of delivering addictive products will pose new challenges to the field of drug abuse prevention. It is hoped that the potential benefits of efficient and effective intervention strategies that employ the full range of universal, selective, and indicated interventions will become a common feature of an overall positive youth development program delivered through existing community, state, and national structures.