Epidemiology of Drug Use Among Adolescents


This review of the epidemiology of adolescent drug use focuses on typical stages of the drug use career and assesses the diversity of experience that my characterize special subgroups of youth. As shown by earlier longitudinal studies (), drug use during the adolescent years is a dynamic, multistage phenomenon which my best be understood by a conceptualization of progressive stages of involvement. The most recent data on dominant patterns of drug use in the 12 to 21 age group are provided by two series of nationwide surveys: the national survey of high school seniors () and the national survey of household population aged 12 and older (). Both of these surveys are cross-sectional, but because many of the data consist of retrospective drug use histories, nationwide patterns of use in 1979 can be described in terms of sequential stages of drug experience.

As Kandel has noted, the study of stages of adolescent drug use does not imply “that the use of a drug causes the progression to the next level. Nor can we assume that once started adolescents will progress through the entire sequence” (). Rather, during the adolescent years, each stage of the drug use career represents a risk factor with regard to more serious stages of drug involvement. Epidemiologic data can specify the levels of risk associated with progression to each successive stage. Such a conceptualization at once facilitates the identification of various kinds of antecedent factors () and reveals that the progression to each stage of drug use can be predicted with reasonable accuracy by the degree of prior contact and experience with drugs as well as the age at which these earlier drug experiences occurred. The stages-of-drug-use conceptualization is also suited to the study of drug use consequences, for, as Cisin (1979) has noted, the consequences of using a particular drug are most appropriately linked to a specific level of drug experience.

Stages of drug use involve two distinct dimensions:

  • An empirical hierarchy of drugs or drug classes, which in studies of the adolescent population has been examined in terms of the sequence in which the first use of various drugs occurs; and
  • Various degrees of involvement with a single drug or drug class, including drug use entry, the continuation of use, the move to abusive levels of consumption, and discontinuation of use.

These two dimensions of drug use careers – progression through a hierarchy of drugs and involvement with a single substance-are conceptually distinct. Certainly, an individual could experiment with a range of substances without continuing to use any of them; alternatively, a young person could become seriously involved in the use of a single substance without trying any other drugs. However, longitudinal data have indicated that, at least in the adolescent population, the two dimensions are empirically related: Kandel and Faust (1975) found that the heavier the use of a lower-ranked drug at Time 1, the more likely it was that progression to a higher-ranked drug would occur by Time 2, 6 months later. Furthermore, those who continue to use a higher-ranked drug usually also continue to use a lower-ranked drug ().

Stages of drug use thus include progression from lower-ranked to higher-ranked drugs and progressive degrees of involvement with various substances; concurrent use of different drugs often occurs, for the two dimensions typically converge as a young person becomes more (or less) involved in drug use. It should be noted that one aspect of multiple drug use that has not been widely examined is the use of two or more drugs on the same occasion. Data from the study of young men and drugs () indicate that this phenomenon deserves attention, at least insofar as alcohol and marijuana are concerned. Bunce (1977) reports that of the young men who had used alcohol on at least 10 occasions and had also used marijuana on at least 10 occasions, over 80 percent said they had used these drugs in combination. Consequences have typically been studied in terms of a single drug or drug class, but as Bunce has pointed out, different consequences may occur as a result of multiple drug use, particularly simultaneous use of two substances.

In this paper, the most recent findings on dominant patterns of adolescent drug use, which are provided by the two, series of cross-sectional national surveys, will be discussed in terms of the following topic areas: (1) the sequence of first use of various drugs; (2) stages of involvement with marijuana, including the entry into marijuana use, the importance of the opportunity factor, the continuation of marijuana use, and later stages of involvement with this drug; (3) stages of involvement with stronger drugs;2 and (4) the consequences of marijuana use. A final section addresses questions of diversity, that is, the extent to which the experience of various adolescent population groups diverges from the dominant patterns observed in the national studies.

The survey of high school seniors is conducted by the Institute for Social Research at the University of Michigan, while the survey of the household population is jointly conducted by the Social Research Group of the George Washington University and Response Analysis Corporation. Both surveys provide data on current patterns of use and past experience for a broad spectrum of substances, including marijuana / hashish, cocaine, hallucinogens such as LSD and PCP, and the nonmedical use of psychotherapeutic drugs available by prescription. The most recent survey in each series was conducted in 1979. The instruments used in the two studies have achieved a high degree of comparability; while one is administered in the schools and the other is administered in the home, the results are strikingly similar (Methods of sampling and response rates are described by Fishburne et al. (1980) and Johnston et al. (1979). Briefly, the national survey of the household population is based on a stratified random sample selected in several stages, including selection of 111 primary geographic areas, 500 smaller areas, specific households, and random selection of respondents from within designated households; the response rate for 12- to 17-year-olds was 86 percent, while the response rate for 18- to 25-year-olds was 84 percent. Overall, 2,165 youths aged 12 to 17 participated in the 1979 study, as did 1,016 18- to 21-year-olds (). About 125 to 130 public and private schools participate in the national survey of high school seniors. Normally, between 66 and 88 percent of the schools originally designated agree to participate; for each refusal, a similar, replacement school is recruited. About 30 percent of all sampled students complete the questionnaire, with absences from class being the chief reason for nonresponse. Altogether, 15,500 high school seniors completed the 1979 questionnaire).

The two sample surveys (both of which are funded by the National Institute on Drug Abuse) represent the vast majority of the total population of adolescents aged 13 to 21; however, certain population groups are, by definition, excluded from each study: The national survey of high school seniors excludes high school dropouts, while the national survey of the household population excludes older adolescents (aged 18 to 21) who have joined the armed forces or who reside in college dorms. (In recent years, between 15 and 20 percent of each age cohort has dropped out of high school before graduation (); and, of those who completed the senior year in 1972, about 80 percent were living in households 2-1/2 years later, i.e., in the fall of 1974, while about 15 percent were residing in college dormitories or fraternity/sorority houses, and about 3 percent resided in military accommodations (). These figures are provided to give the reader a rough idea of the size of the excluded groups. A worldwide survey of alcohol and drug abuse among members of the armed forces has recently been carried out by Burt Associates under the sponsorship of the Department of Defense; results have not yet been made available by DOD. While a number of surveys of college students were conducted circa 1970 (e), these have not been recently replicated). Both surveys also exclude certain fringe groups such as youths who are institutionalized, as well as older adolescents who have transient lifestyles and lack a fixed address.

The most recent findings of the survey of high school seniors are provided in a report of Highlights of the 1979 study (). A canprehensive report of prevalent rates from the 1979 household survey is provided by Fishburne et al. (1980) and in separate volumes of detailed tabulations (); a series of papers based on special analyses of the data from the household population survey is presented in a forthcoming volume (Rittenhouse); in press); and a Highlights report () summarizes and a Highlights report () summarizes form the basis of the discussion of dominant patterns of adolescent drug use, which is organized around the stages-f-drug-use conceptualization suggest& by earlier longitudinal studies.

The Hierarchy of Drug Classes

Recent reports of the order in which young persons recall having first used various drugs confirm earlier results based on scalogram analysis () and those of longitudinal studies (). The use of alcohol and cigarettes typically precedes the first marijuana experience (Rittenhouse, unpublished). About half of those who try marijuana (and/or hashish) in the teen years eventually use stronger drugs such as cocaine and hallucinogens-although this progression is often delayed, and in many cases does not occur until young adulthood. Those who begin marijuana use at earlier ages are more likely than others to progress to the use of these stronger drugs (). When the stronger drug list is extended to include stimulants, sedatives, and tranquilizers, correlational data indicate that the proportion of older adolescent marijuana users who report experience with other illicit drugs increases to a clear majority ().

The first use of marijuana typically occurs prior to the senior year of high school, often between the ages of 14 and 17; by contrast, the first use of stronger drugs such as cocaine and hallucinogens is most likely to take place in late adolescence (). Thus, very different lifetime prevalence levels are observed for the various adolescent groups: less than 10 percent of all 12- to 13-year-olds have ever used marijuana, but the rate increases to about 30 percent for 14- and 15-year-olds, to about 50 percent for 16- and 17-year-olds, then to 60 percent for high school seniors, and to almost 70 percent for the 18 to 21 age group (). Similarly, less than 10 percent of all youths aged 12 to 17 have ever experimented with drugs such as cocaine and LSD, but the rate is close to one-third for 18 to 21-year-olds ().

Stages of On Contact and Involvement with Marijuana

Beginning Marijuana Use

The importance of opportunity as an antecedent to first marijuana use is suggested by a comparison of the opportunity and lifetime prevalence rates for various adolescent age groups shown in table 1. Clearly, during the early adolescent years, lack of opportunity represents an effective barrier to beginning drug use; the large majority of 12-and 13-year-olds have never had the chance to try marijuana, and, indeed, less than 10 percent of this age group has ever used this drug. But by the later adolescent years, opportunities to try marijuana are widespread and the majority have used it. Thus, the opportunity factor explains much – but not all – of the difference in the lifetime prevalence rates observed for various adolescent age groups. As may be deduced from table Marijuana Opportunity and Lifetime Prevalence Rates, by Adolescent Age Croups, the older adolescent age groups are more likely to have taken advantage of the chance to try marijuana. However, the earlier in life the first marijuana opportunity occurs, the more likely it is that the individual will eventually try the drug (Somerville and Miller, forthcoming); when this result is considered with the previously noted finding of early marijuana use as predictive of eventual progression to stronger drugs, the importance of early contact with drugs for subsequent stages of involvement seems clear.

Table Marijuana Opportunity and Lifetime Prevalence Rates, by Adolescent Age Croups

12-13 14-15 16-17 18-21
(671) (721) (773) (1,016)
Percent who have the hance to try marijuana 20% 51% 69% 83%
Percent who have ever used marijuana 8% 32% 51% 69%

Interestingly, few young persons in any age group take advantage of their first chance to try marijuana – even when fairly long-term acquaintance with a user precedes the first opportunity experience (Somerville and Miller, forthcoming). The usual time lapses or time lags across first acquaintance with a marijuana user, first opportunity and first use, suggest that few young persons seek out the chance to try marijuana, that many are at first hesitant to begin use, and that repeated opportunities account for the fact that the substantial majority of those youths who are exposed to marijuana opportunity eventually try the drug.

Continuing Marijuana Use

In earlier years, lack of availability my have constituted a barrier to continuing marijuana use, but this is not the case today – at least not in the older adolescent age groups. The survey of high school seniors indicates that marijuana is almost universally available: 90 percent of the respondents to this survey said that marijuana would be fairly easy or very easy for them to obtain (). Few young persons can be characterized as one- or two-time experimental marijuana users. While 60 percent of high school seniors report experience with marijuana, less than 10 percent say they have used it on only one or two occasions; indeed, almost one-half of the high school seniors who ever tried marijuana say they have used it 40 or more times (). Similarly, in the 18 to 21 age group, use on 100 or more occasions is three times as likely as one- or two-time use. Even among youths aged 12 to 17, only one-fourth of the ever-users say that their experience has been limited to one- or two-time use; a similar number have used marijuana on 100 or more occasions. More than one-half of all adolescents who ever tried this drug are current users, i.e., reported use during the month prior to interview (). Thus, it seems clear that once in young person tries marijuana, he or she is likely to repeat the behavior and to continue use during the adolescent years.

Tabulations showing current use rates among 18- to 25-year-olds by their recalled age at first use () indicate that those who began use earlier in life are more likely to be current users; thus, the earlier the age at the first use of marijuana, the more likely the young person is to continue use through the young adult years. Again, the importance of the onset of the drug use career is emphasized.

Finally, it should be noted that almost all current marijuana users are also current alcohol users; indeed, among 18- to 25-year-olds who are current users of both drugs, the number of days on which alcohol was used during the month prior to interview correlates with the number of days on which marijuana was used (). Thus, continuing marijuana users are also continuing alcohol users, and the simultaneous or combined use of these substances clearly deserves investigation.

Later Stages of Involvement with Marijuana

Turning to heavier patterns of marijuana consumption, how likely is it that a young person will use this drug on a daily or near-daily basis? While no data are available on the number of youths who ever used marijuana on a daily or near-daily basis, both major surveys indicate that among current marijuana users in the adolescent age range, about one-fourth or more (depending on age) used this drug on at least 20 days or occasions during the month prior to interview: 17 percent of all youths aged 12 to 17 are past-month users and 4 percent of this entire age group reports use on at least 20 days during the month prior to interview (). Johnston reports that 36.5 percent of all high school seniors are current users, and that 10.3 percent of this high school class used marijuana on 20 or more occasions during the preceding 30 days. The corresponding figures for 18- to 21-year-olds, reported by Response Analysis (1980), are 40 percent (current use) and 12 percent (daily or near-daily use).

How often do youths discontinue the use of this drug? Quitting, or discontinuation of marijuana use, appears to be an unlikely outcome – at least until the adolescent years are past. Focusing on 12- to 17-year-olds who had first used marijuana more than 6 months prior to interview and who had used the drug on more than 10 occasions, Parry and Cisin (forthcoming) found that only about 10 percent of this user group reported no use during the past 6 months. Discontinuation of use is somewhat more likely in the years that follow adolescence if the young person marries and forms a family of his/her own.

Stages of Involvenent with Stronger Drugs

Patterns of beginning use of stronger drugs are similar to those discussed for the entry to marijuana use: Acquaintance with a cocaine or hallucinogen user precedes the young person’s first opportunity to use one of these drugs. Few take advantage of their first chance to use either of these drugs, but many eventually try one or both – although the tendency eventually to try these substances is not as strong as the tendency to use marijuana (Somerville and Miller, in press). Stronger drug opportunity typically occurs subsequent to marijuana opportunity but may precede marijuana use; nonetheless, the decision to use a stronger drug is typically delayed until after marijuana has been tried.

Once a stronger drug has been used, continued use is less likely than was the case for marijuana. While the figures vary depending on the specific drug and the specific age group in question, it is clear both that experimental one- or two-time use is a more likely outcome for stronger drugs than for marijuana, and that current use, considered as a proportion of ever use, is lower for stronger drugs than for marijuana. Yet, one-third or more of the older adolescents who have ever tried cocaine report current use; e.g., 15 percent of high school seniors have tried cocaine and 6 percent of this high school class report current use (). Almost all current users of stronger drugs also report current use of both marijuana () and alcohol (Response Analysis 1980).

Among current users of stronger drugs, daily or near-daily use is rare (). As Johnston notes:

Less than 1% of the respondents report daily use of any of the illicit drugs other than marijuana. Still 0.6% report unsupervised daily use of amphetamines, and the comparable figure for cocaine and hallucinogens …now stands at 0.2%. While very low, these figures are not inconsequential considering that 1% of each high school class represents over 30,000 individuals. ()

Discontinuing use of stronger drugs has not yet been examined, in part because of the relatively small numbers of adolescent respondents who have been classifiable as continuing stronger drug users.

Consequences of Marijuana Use

Given that so many young persons not only try marijuana, but also continue the use of this drug – in some cases on a daily or near-daily basis – the consequences of marijuana use are of increasing concern. Some current survey data are available on the perceived consequences of use among 18- to 25-year-olds, i.e., older adolescents and young adults. These data are limited to two selected consequence areas: impaired driving ability and the loss of motivation. Reports of experience with these adverse effects of marijuana are most likely among young adults who have used the drug on 100 or more occasions: 50 percent of these most experienced users say they performed less well when driving “shortly after getting really high,” and 30 percent reported that at one time they “stopped caring and didn’t try as hard” because of steady or daily use ().

While these data do indicate that large numbers of young persons may be at risk for negative consequences of drug use, they are limited in a number of ways. This is particularly evident for the data on the amotivational syndrome, since the defining phrase, “stopped caring and didn’t try as hard,” was not further specified as to duration or severity of motivation loss, nor as to the perceived seriousness or ramifications of the experience. Nonetheless, it seems clear that respondents are willing to admit that they have experienced negative outcomes of marijuana use, and further investigations along these lines would seem justified.

The dominant patterns of drug use among adolescents today may be summarized as follows: alcohol use precedes marijuana use, which, in turn, is a prerequisite for stronger drug use. As children progress through the teen years, they are increasingly likely to experience opportunities to use marijuana. Though teenagers are at first hesitant to use the drug, the majority of those who have the chance eventually do try it. Once marijuana has been tried, there is a strong tendency to continue the use of this drug (as well as alcohol) throughout the adolescent years. By the senior year of high school 60 percent have tried marijuana and 37 percent report current use; more than one-fourth of these current users (10 percent of the entire senior class) say they consume the drug on a daily or near-daily basis. By the late teens and early twenties, about half of those who began with marijuana have gone on to use stronger drugs such as cocaine and hallucinogens. The continuing user of marijuana and the stronger drug experimenter or occasional user is likely to be a young person whose first contact with marijuana occurred at an early age. Many of the older adolescents who have had extensive experience with marijuana report adverse effects of use, but an assessment of the seriousness of these consequences must await more careful definition of item in future studies. These represent the dominant patterns for the adolescent population as a whole. But important questions remain regarding the extent of diversity that may characterize the drug use experience of adolescents in various cultural groups and in different geographic locations.

Diversity of Experience

The broad view of variation in drug use experience that the national studies provide indicates an increasing homogeneity across major demographic groups in the adolescent population. Young white and nonwhite respondents now report similar levels of marijuana experience () and the earlier study of young men and drugs also showed similar levels of drug experience for the younger cohorts of black and white males (). Trend analyses indicate that the predictive power of several other demographic variables has also diminished in the past decade; for example, rural drug use rates are now catching up with national prevalence levels (see special report on rural drug use). And while prevalence levels are lower in some regions than in others, the differences are not particularly striking ().

Despite the dominant picture of relative homogeneity that the nationwide surveys suggest, there is evidence of considerable diversity, including the potential for numerous pockets of drug abuse. The national surveys indicate that boys are still more likely than girls to be more heavily involved in substance use (), and that older adolescents (and young adults) who have left the parental home to live with friends or roommates are characterized by extremely high prevalence rates. For example, 90 percent of this group have tried marijuana and 65 percent are current users (). Furthermore, as noted in the most recent Marijuana and Health report:

Nationwide statistics may obscure considerable local variation. For example, in Maryland and Maine, where drug surveys were conducted in 1978, higher levels of daily or near-daily use of marijuana were found than among high school seniors nationwide (10.7 percent of seniors nationally). In Maryland, use “daily or several times a week” was reported by a quarter (25.3 percent) of the twelfth graders….In Maine nearly one in six high school students reported daily marijuana use…. (National Institute on Drug Abuse 1980; see also State of Maine 1979; Maryland Department of Health and Mental Hygiene 1979).

Considerable diversity may also characterize various cultural and minority groups; for instance, Kandel’s study of New York high school students indicated high drug use rates for American Indian youths and low rates for Asian-American children (). But, while a good deal of information appears to exist on drug use by members of minority groups (Austin et al. 1978), current knowledge falls short of adequate epidemiologic description. Neither nationwide surveys of general population groups nor small studies conducted at the neighborhood or community level can adequately describe the experience of minority groups that, although relatively small in comparison to the total population, may be marked by internal cultural diversity.

A major research effort has recently been directed toward one such group. In 1977, 3,000 Native American adolescents from several tribes located in different regions of the country were surveyed by Oetting and Goldstein (). This large-scale study revealed considerable diversity across the various tribes and also showed that, as a whole, Native American children are characterized by ‘prevalence rates that are substantially higher than those reported in national studies, for alcohol, marijuana, various stronger drugs, and particularly inhalants. In addition to presenting lifetime and current use prevalence data, the investigators report the prevalence of various types of drug use patterns. These data suggest that among Native American youths, stages of experimentation and involvement are similar to the dominant patterns observed at the national level; the primary difference seems to be that at each stage Indian children are at higher risk of progressing to further drug involvement.

In sum, an increasing homogeneity characterizes the experience of major adolescent population groups; dominant patterns described in detail by national surveys indicate the importance of the earliest drug experiences and document high levels of risk for the typical young person today. But even higher risks obtain for certain identifiable population groups. Some pockets of drug abuse can be identified in nationwide research efforts; for instance, the exceedingly high prevalence levels observed for older adolescents who have substituted peer-group living for the daily influence of family life. However, the full extent of diversity cannot be portrayed by studies with a national focus. Statewide surveys and the study of Native American youths indicate that fairly large-scale studies directed at smaller populations are necessary to uncover concentrations of drug use in certain adolescent populations.

Finally, the adolescent groups that are by definition excluded from the national survey frames deserve special concern, since there is reason to believe that high school drop-outs, young members of the armed forces, and residents of college dormitories may be characterized by higher levels of involvement with drugs than their same-age peers who are routinely surveyed.

Future Research Directions

The surveillance function of epidemiologic research continues to be well served by national studies that describe risk factors for the majority of the adolescent population. Given the wealth of data that has now been amassed by the two series of nationwide studies, important questions concern future research directions. Certainly, large populations that are ordinarily excluded from national survey frames deserve separate study. Two other research directions should also be considered: First, data that already exist could be further mined to increase our knowledge of drug use in certain population groups. For instance, each year the survey of the household population includes a relatively small number of high school drop-outs; by pooling data across recent survey years, it would be possible to obtain reasonably reliable information on this under-studied population group. Second, in planning future national studies, it might be advisable each year to select a special alternative population group for oversampling sufficient to allow separate analysis and to identify pockets of drug abuse that deserve further investigation in specially designed research efforts.

Thus, the mass of data that has been gathered in successive waves by the two national surveys should be subjected to further analysis, and future waves of such studies can be made more sensitive to detecting pockets of drug abuse in special adolescent population groups. In considering separate studies of special groups, priority should be given to those groups of young persons who are by definition excluded from the national surveys – and for whom we have almost no data at the present time.

Judith Droitcour Miller, M.A.


Selections from the book: “Drug Abuse And The American Adolescent”. Dan J. Lettieri, Ph.D., and Jacqueline P. Ludford, M.S., eds. Review Report, emphasizing use of marijuana: epidemiology, socio-demographic and personality factors, family and peer influence, delinquency, and biomedical consequences. National Institute on Drug Abuse Research Monograph 38, 1981.