volvement in illegal drug use changed from being a set of experiences and behaviors concentrated mostly in fairly small, identifiable subgroups on the periphery of mainstream American society to becoming a widespread and more normative set of experiences and behavior. In terms of numerical increases of new users and previously observed and expected numbers, these illegal drug use experiences came to be characterized as an epidemic, and drug use became an important “generation gap” issue, with youths and adults often holding radically different views of the acceptability of drug use—particularly the use of certain drugs such as marijuana and hallucinogens (National Commission on Marijuana and Drug Abuse, 1972).

At the peak of this epidemic of illegal drug use, between 1975 and 1980, fully two thirds of American adolescents had tried an illegal drug by the time they finished high school (Johnston, O'Malley, & Bachman, 2002a; see also Substance Abuse and Mental Health Services Administration (SAMHSA), 2002a). This represented a dramatic increase over the prevalence proportions observed in the mid-1960s, when the epidemic began (Johnston, 1973; National Commission on Marijuana and Drug Abuse, 1972). During this interval, illegal drug use came to be associated with political beliefs—particularly, being against the Vietnam War—and with certain lifestyle orientations, as reflected in the counterculture (Johnston, 1973; Zinburg & Robertson, 1972). At the population level, there were also associations with other forms of rule-breaking behavior, deviance, or delinquency unrelated to using drugs. These associations remain, even though many drug-using youths are otherwise rule abiding and do not show other conduct problems, and some youths with conduct problems do not take illegal drugs (Jessor & Jessor, 1977; Johnston, 1973; Osgood, Johnston, O'Malley, & Bachman, 1988).

If the Vietnam War and other historical events of the 1960s and early 1970s accounted for the dramatic increase in the epidemic of illegal drug use, one might have expected a downward trend with passage of time since these events. That turned out not to be the case. The Vietnam War ended in 1973, but the rise in drug use continued into the late 1970s, albeit with some fading of the symbolic meanings of illegal drug use. During this interval, the illegal drug use epidemic developed its own forward momentum.

As gauged in relation to the number of new adolescent users year by year, it was not until the late 1970s that the epidemic trend turned downward. Even so, the proportion of young people continuing to use drugs did not decline across the board until after 1985. The decline in each year's prevalence proportion persisted into the early 1990s. During this interval of time, norms among young people against the use of many of the illegal drugs strengthened considerably. Our surveys showed an increased appreciation of harms associated with illegal drug use, in particular marijuana, cocaine, crack, and phenlcyclidine (PCP) (Bachman, Johnston, & O'Malley, 1990, 1998; Johnston, 2003).

This change in perceptions about drugs and a concomitant decline in prevalence of illegal drug use during the late 1980s, however, may have helped to sow the seeds of its own reversal and to create a context for emergence of a noteworthy “failure of success.” By this we mean that important new historical events were emerging in the late 1980s and early 1990s to take the center stage of the popular imagination. Headlines and media attention shifted away from domestic matters such as illegal drug use and toward concerns about terrorism abroad, the Iraqi invasion of Kuwait, and the emerging Gulf War. With these declines in coverage of the drug issue by the media, Congressional attention and budgeting for drug prevention programs shrank considerably, and political attention to the issue declined generally. Young people were hearing and reading much less about illegal drug use, and perhaps more importantly, their perceptions reflected less familiarity with hazards that go along with illegal drug use (Johnston, 1991, 2003). In our retrospective analysis of the domains of influence that govern youths' appreciation of the hazards of illegal drug use, we build from a foundation of observations by historian David Musto and others, highlighting (a) media coverage, (b) drug prevention programming, (c) personal experiences, and (d) vicarious experiences (i.e., learning vicariously from peers, parents, or other relatives of the drugs' hazards experienced by others). To the extent that society achieves suc

cess in dampening the prevalence of illegal drug use (by whatever means), we may expect controllable declines in the first two domains: (a) media coverage falls, and (b) support for prevention programming wanes. Part and parcel with declines in the prevalence of illegal drug use are less personal experience with these drugs, fewer chances to try the drugs, and fewer young people experiencing the hurt that often goes along with drug taking. In addition, there is less vicarious experience with the associated hazards as can be gained by personal acquaintance with other young people or adults whose lives have been harmed by illegal drug use. Hence, on the downward side of an epidemic curve of illegal drug use, the same processes that fuel the continuing decline in the behavior of illegal drug use are fueling a decline in adolescents' personal and vicarious knowledge of the hazards of illegal drug use. In this sense, a “success” in the form of declining prevalence of illegal drug use sows the seeds for a “failure” and later rebound—to the extent that the knowledge of drug-associated hazards helps to promote resistance when the young person faces the first or subsequent chance to try an illegal drug.

These seeds for a resurgence of illegal drug use among American adolescents had been sown in the late 1980s and early 1990s, and for most illegal drugs, the epidemic curve turned upward in the early 1990s, with a generally persistent trend of increasing proportions of new users and continuing users into the late 1990s. Marijuana use exhibited the sharpest rise during this “relapse phase” in the epidemic, but use of most of the drugs in the ever-growing list of alternatives increased during this period as well (Johnston, O'Malley, & Bachman, 2003b; SAMHSA, 2002a,b).

By the late 1990s there were once again signs of improvement, with prevalence of inhalant use beginning to decline in 1996, and prevalence of marijuana use peaking or stabilizing by 1997. Other drugs began to decline at various points, including LSD (lysergic acid diethylamide), cocaine, and finally heroin. But one drug was growing sharply in popularity in the late 1990s— namely, ecstasy, or MDMA (3,4-methylenedioxy-methamphetamine). With respect to prevalence of recent use, this newcomer to the list of popular drugs reached peak values in the 2000–2002 interval, and there now may be a persisting downward trend in prevalence of MDMA use, concomitant with increases in the proportion of young people perceiving adverse effects and hazards of MDMA use, as happened with a number of other drugs in prior years (Johnston et al., 2003b).

The ecstasy epidemic among youth and young adults illustrates one very important feature of the larger epidemic of illegal drug use over the past three decades: that there has been a continuing march of new drugs onto the scene, each presenting youths with new alternatives. While American youth may have learned about the dangers of most of the existing alternatives on the menu, we may expect continuing innovations and “designer drugs,” perhaps with claims that the new drug compounds have no adverse consequences. As an elaboration of the “failure-of-success” concept, we now appreciate that when one birth cohort of adolescents comes to appreciate the dangers of a drug, by learning about its consequences through the media, prevention programming, and personal and vicarious experience, and enters the lower-risk developmental period of adulthood, the more recently born cohorts of children enter adolescence without the history of accumulated knowledge and belief about the hazards connected with the drug. These adolescents are thus prone to reexperience and relearn these dangers on their own.

The ecstasy/MDMA example illustrates another important point of particular relevance to prevention. A careful examination of the trends over time of the various classes of illegal drugs will show that to a considerable degree, they each have unique cross-time profiles (Johnston et al., 2003a; SAMHSA, 2002a). The use of one drug may be rising at the same time that the use of another is falling and perhaps a third is holding steady. This means that the different drugs are responding to influences that are specific to them—very likely, factors such as the perceived benefits of using that drug as well as the perceived dangers of doing so. Although there may be some larger social forces, such as the Vietnam War, that change the overall proportions of youth willing to engage in illegal drug use gen

erally, there are also many important drug-specific influences that must be taken into account. This means that drug education, communication, and persuasion efforts may be most valuable when they address each class of drug separately. The attitudes and beliefs related to drugs are so varied that in addition to pursuing goals common to all drugs, such as strengthening resistance to peer influence in favor of any illegal drug, there should be educational efforts specific to individual drugs (Johnston et al., 2003a; SAMHSA, 2002a).

There are several ongoing survey series available for assessing the size and nature of the American adolescent drug experience, based on scientifically selected national samples. Two are based on in-school surveys using self-completed paper-and-pencil questionnaires administered to students in group settings: the Center for Disease Control and Prevention's (CDC's) Youth Risk Behavior Study (YRBS), and the University of Michigan's Monitoring the Future (MTF) study, sponsored by the National Institute on Drug Abuse. The MTF was launched in 1975 and the YRBS in 1991.

In the YRBS, a nationally representative sample of some 13,000 to 14,000 students in grades 9 through 12, enrolled in about 150 public and private high schools, are surveyed by means of self-administered, optically scanned questionnaires (Grunbaum et al., 2002; Kann, 2002). Data are gathered biennially. Measurement is spread across a range of risk behaviors for adolescents, so little information is gathered on attitudes, beliefs, or social surroundings specifically related to drug use. This is a repeated cross-section design.

In the MTF, some 45,000 students in grades 8, 10, and 12, enrolled in public and private schools in the coterminous United States, are surveyed annually. Self-administered, optically scanned questionnaires are used in this study as well. Extensive information is gathered on attitudes, beliefs, and various social influences from the family, school, work, and mass-media environments. In addition, representative panels of high school seniors are selected for follow-up each year and are then surveyed by mail for some years after high school graduation in this cohort-sequential study design.

The third source of population survey data derives from national household surveys that include and report separately on youth 12 to 17 years old—the National Household Survey on Drug Use and Health (NHSDUH), known until very recently as the National Household Survey on Drug Abuse (NHSDA). Data were gathered from adolescents for many years by personal interview in combination with private answer sheets on drug use, but very recently the methodology has shifted to computer-assisted interviewing. This series began in 1971 with a survey for the National Commission on Marijuana and Drug Abuse (1972). Over the years there have been a number of changes in measurement content, measurement methods, and sample sizes, all of which have made accurate trend estimation more of a challenge. In general, however, trends have been reasonably parallel to those generated by MTF and YRBS. The NHSDUH uses a repeated cross-section design, with surveys conducted at various intervals in the past, but on an annual basis in recent years.

The National Comorbidity Survey (NCS; Kessler, 1994) represents a somewhat different approach to generating prevalence data on drug use, in that it measures the prevalence and correlates of DSM III-R disorders, as well as the connection of drug dependence and related disorders to the various other psychiatric disorders (Anthony, Warner, & Kessler, 1994). One key finding is that there is a significant level of such comorbidity among individuals with a drug de

pendence syndrome (see, for example, Kessler et al., 1996, 2001). Unfortunately, the relatively small samples of adolescents in the first iteration in the NCS put limitations on the precision of estimates specific to adolescents. The NCS household sample included around 8,000 people from ages 15 to 54 in the noninstitutionalized population of the United States, and it was fielded between 1990 and 1992. Sequel surveys are now being carried out that will, among other things, provide trend estimates on many of the conditions measured in the first wave. In addition, and of particular relevance here, a supplementary sample of 10,000 adolescents is included in the current work to specifically examine the prevalence and correlates of drug dependence and other disorders among adolescents.

Still another, completely different, type of information is gathered nationwide from hospital emergency rooms and coroners' offices as part of the Drug Abuse Warning Network (DAWN; e.g., SAMHSA, 2002b), in which case counts are made of people treated for medical emergencies involving any of a range of drugs and of people who die with identifiable evidence of drug use present. Unlike the population surveys, which attempt to estimate prevalence and trends in drug use in major segments of the national pop ulation, the DAWN system is intended to generate data on case counts of drug-related “casualties.”

Prevalence rates of adolescent use of specific substances (heroin, cocaine, alcohol, and tobacco especially) are given in the discussion of those substances later in this chapter. Here we focus on the prevalence rates and trends of use across various substances.

The 2002 MTF survey shows that a quarter (25%) of today's young people have tried some illegal drug before finishing eighth grade—that is, by ages 13 or 14—and more than half (53%) have done so by the end of high school (Johnston et al., 2003b). If inhalants are included in the definition of illegal drugs, the numbers are even higher (32% and 55%, respectively). Prevalence rates for any illegal drug, marijuana, cigarettes, and binge drinking of alcohol observed among 8th-, 10th-, and 12th-grade students in these nationally representative surveys are given in Figures 17.1–17.4. (Grade 8 students are 13 or 14 years old for the most part, and grade 12 students are mostly 17 or 18 years old.) Trend data for the three grades illustrate a number of the points made above, including the dynamic na

ture of this class of problem behaviors among youth, and that different drugs tend to vary independently of the others.

By comparison, Tables 17.1 and 17.2 provide trends in the proportion of 8th-, 10th-, and 12th-grade students in MTF who reported receiving treatment for their use of alcohol and/or illegal drugs in recent years. As would be expected, the values rise with age (and, therefore, with prevalence), but not as much as one might expect. Among 12th graders (the only ones asked about the distinction) the number receiving outpatient treatment or counseling exceeds the number receiving inpatient treatment by a ratio of 3:1 to 5:1 in recent years, but by larger ratios in earlier years. The lifetime prevalence of treatment also rose toward the end of the 1990s, no doubt reflecting the relapse in the epidemic of adolescent drug use from the early to mid-1990s.

Other important survey series that attempt to measure clinically defined cases of drug depen