Epidemiology Of Drug Abuse

Epidemiology Of Drug Abuse

One of the best ways to introduce an article on the epidemiology of drug use and drug dependence is to ask some basic questions that epidemiologic studies can answer but laboratory and clinical studies cannot. Here are some examples:

In the late 1990s in the United States, about how many ages 12 to 17 had used cocaine at least once?

In the late 1990s, within which U.S. population subgroups were active cocaine users most likely to be found?

Within the United States in the early 1990s, among those aged 15 to 24 who had used cocaine, what proportion had become dependent on it?

In the early 1990s, which age group within the U.S. population was most likely to have experimented with cocaine, and which age group was most likely to have developed cocaine dependence?

For a young adult living in the United States, what is the risk of developing the problem of alcohol abuse or dependence between one year and the next?

Is the risk for alcohol dependence greater for some young adults than for others?

Which subgroups of young adults are at especially high risk for alcohol dependence?

Are these same subgroups of young adults at especially high risk of becoming dependent on psychoactive drugs such as marijuana or cocaine?

To answer questions of this type, it is necessary to step outside the laboratory and clinical settings where drug users receive treatment. This step can be taken during the course of epidemiologic surveys that seek information about all aspects of the population's drug experience; the surveys take into account not only the relatively modest numbers of drug users who have received counseling and treatment, but also those who never have received any kind of health care or social services. The answers to these questions, based on epidemiologic surveys conducted in the United States between 1980 and the present, are as follows:

In the late 1990s, among those aged 12 to 17 in the United States, an estimated 496,000 to 682,000 had used cocaine at least once. As a proportion, this amounted to about 2.5 percent of those 12 to 17 in the United States at that time.

Within the United States in the late 1900s, young adult men aged 18 to 29 were more likely to be active cocaine users than any other population subgroup categorized by age and sex. For example, slightly more than 2.5 percent of men 18 to 25 were active cocaine users, as compared with 1.4 percent of men 26 to 34, 1.3 percent of women aged 18 to 25, and 0.9 percent of women aged 26 to 34.

Within the United States in the early 1990s, among those aged 15 to 24 who had used cocaine, an estimated 25 percent had become dependent on it. That is, for every four who had experimented with cocaine, one had become dependent on it.

Within the United States in the early 1990s, people of the 25 to 34-year age group were most likely to have experimented with cocaine; within this age group, about 30 percent of men had tried cocaine at least once, and about 21 percent of women had tried cocaine at least once. Cocaine dependence also was most prevalent in this age group: it affected about 4 percent of all persons aged 25 to 34. Among cocaine users aged 25 to 34, an estimated 16 percent had become dependent on it.

For those 18 to 29 living in the United States, the best available estimate for the risk of developing alcohol abuse or dependence between one year and the next is about 2 to 4 percent.

The risk of succumbing to alcohol abuse or dependence for males aged 18 to 29 is an estimated 6 percent per year, as compared with about 1 percent per year for females aged 18 to 29.

Males between the ages of 18 and 25 are at especially high risk of succumbing to alcohol abuse or dependence.

These same subgroups of young adults are at especially high risk of becoming dependent on psychoactive drugs such as marijuana or cocaine. When all the abuse or dependence syndromes attributable to nonmedical use of these drugs are considered, the estimated risk for males aged 18 to 29 of developing clinically recognizable drug problem is estimated at 4.4 percent per year; for females aged 18 to 20, it is about 1.6 percent.

There is, of course, good reason to wonder whether epidemiologic surveys of drug use and drug dependence have sufficient validity to be trusted. On the one hand, especially among young people, there may be a tendency to exaggerate drug taking, and to falsify survey responses in the direction of more drug taking than has really occurred. On the other hand, some people may be hesitant to disclose their histories of drug taking or drug problems; they might not agree to participate in the survey, or they might falsify their answers in the direction of less drug taking or fewer problems than have actually occurred.

There fortunately is a body of methodologic research that provides some general assurance about the accuracy of estimates in epidemiologic surveys. Accuracy of the survey results seems to be enhanced considerably when special care is taken to guarantee confidentiality of responses, to protect the privacy of the survey respondents, and to develop trust and rapport before asking survey questions about sensitive behavior, alcohol and drug problems, or illegal activities. In particular, except in poorly conducted surveys of very young respondents, there seems to be very little exaggeration of drug involvement, and older adolescents and adults rarely report drug use unless it actually has happened. Moreover, the accuracy of the estimates does not seem to be distorted too much when the surveys concentrate on household residents and do not extend their samples to include homeless or imprisoned segments of the population. Even though homeless people and prisoners often have significant and special needs for alcohol- and other drug-dependent treatment services that society cannot ignore without peril, the number of homeless and incarcerated persons is small relative to the considerably larger number of persons living in households.

It also is important to note the relatively large size of the survey estimates obtained in these epidemiologic surveys. For example, in 1998, as part of the HIGH SCHOOL SENIOR SURVEY (Monitoring the Future), almost 16,000 high school seniors were asked to fill out confidential questionnaires about their use of such drugs as marijuana and cocaine; more than 38 percent reported having taken these drugs illegally, 80 percent reported consuming alcoholic beverages, and more than 60 percent reported having consumed alcohol to the point of getting drunk. In 1998, more than 25,500 American household residents aged 12 years and older participated in a U.S. government-sponsored NATIONAL HOUSEHOLD SURVEY ON DRUG ABUSE and were asked to answer an interviewer's questions about the use of these drugs; illegal drug taking was reported by an estimated 21 percent of those 12 to 17 years, 48 percent of those 18 to 25, 51 percent of those 26 to 34, and 32 percent of older adults. Furthermore, between 1990 and 1992, almost 9,000 Americans aged 15 to 54 completed confidential interviews as part of a U.S. government-sponsored National Comorbidity Survey. According to this survey, one in three tobacco smokers had tobacco problems, signs, and symptoms consistent with their having become dependent on tobacco and one in seven drinkers had alcohol problems, signs, and symptoms consistent with their having developed the clinical syndrome of alcohol dependence. Among those who reported use of marijuana, heroin, or other controlled substances, one in seven reported drug problems, signs, and symptoms consistent with their having become dependent on these drugs. These survey-based estimates are already high enough to provoke social concern. They would be even higher if corrections were to be made to account for respondents who were hesitant to report either their consumption of these drugs or the problems associated with drug use that they had.

DRUG-SPECIFIC ESTIMATES FOR THE U.S. POPULATION

It may be useful if, bearing in mind these potential limitations in the survey methods, one considers each broad drug class one by one, in order to convey the relative frequency of use of tobacco, alcohol, and other drugs in the United States, and to identify population subgroups within which drug use or drug dependence is most common. (From this point on, estimates based on the 1999 survey of high school seniors are labeled MF estimates; those from the 1998 National Household Survey on Drug Abuse are labeled NHSDA estimates; and those from the 1990-1992 National Comorbidity Survey are labeled NCS estimates.) In view of recent attention to the CAFFEINE-dependence syndrome and other health hazards of drinking COFFEE or TEA or consuming other caffeinated products, estimates concerning the use of caffeine and caffeine dependence might seem warranted. There is not yet a stable base of epidemiologic data on caffeine use and caffeine dependence, however; these remain topics that ought to be examined in future epidemiologic studies.

Tobacco Smoking in the Late 1990s.

Monitoring the Future (MF) estimates show that about 65 percent of high school seniors have smoked TOBACCO cigarettes at least once. An estimated 35 percent of high school seniors smoked tobacco cigarettes at least once during the month prior to the survey, and 23 percent had become daily tobacco smokers.

According to the National Household Survey on Drug Abuse (NHSDA), which included household residents age 12 years and older, an estimated 68 to 71 percent smoked tobacco cigarettes at least once, for a total of about 149,021,000 to 155,515,000 smokers. An estimated 29 to 32 percent had smoked in the year prior to the survey, for a total of 64,012,000 to 69,522,000 recently active smokers; most of these had smoked in the month prior to the survey (57,811,000-63,072,000).

There was an important age and sex-related variation in these estimates. For example, among adults past age 34, males were more likely than females to have been recent tobacco smokers (26.9% versus 23.4%). Among those 18 to 25, within the limits of survey error, there essentially were no differences between the sexes in prevalence of smoking, and both estimates were in a range from 37.5 to 45.3 percent. Among those 12 to 17, there also were no statistically reliable differences between the sexes, and the estimated proportions were between 17 and 21 percent; although estimates from earlier years show the proportion of girls smoking in this age group to be numerically greater than that for boys of the same age (NHSDA estimates).

Using data from the National Comorbidity Survey of Americans aged 15 to 54, it has been possible to estimate the proportion of tobacco smokers and other drug users who have developed drug-dependence syndromes, as defined in relation to a set of diagnostic criteria for drug dependence that were developed by the American Psychiatric Association in 1987. Before the diagnoses of drug dependence are made, the survey must produce evidence that drug users experienced signs or symptoms of dependence such as going through withdrawal or taking drugs to avoid withdrawal symptoms. Applied to the tobacco smokers identified in the Naitonal Comorbidity Survey, these diagnostic methods indicated that almost one-third of tobacco smokers in the survey population had developed tobacco dependence. That is, for every three tobacco smokers, one had developed tobacco dependence and was found to have met the American Psychiatric Association's diagnostic criteria for dependence on this drug. Of the more than 70 percent of respondents who had smoked tobacco at least once, a truly remarkable proportion of about 24 percent was found to have a history of currently active or former tobacco dependence (NCS estimates).

Smokeless Tobacco Use in the Late 1990s.

An estimated 23 percent of high school seniors had tried smokeless tobacco at least once, and about 8.4 percent had used it during the month prior to the survey (MF estimates). Household survey estimates indicate somewhat lower values, except among males aged 18 to 25. For example, among 12- to 17-year-olds, an estimated 8.9 percent had tried smokeless tobacco, and just over 1 percent had used it in the month prior to the survey. By comparison, slightly more than 24 percent of 18- to 25-year-olds had tried smokeless tobacco; corresponding estimates for 26- to 34-year-olds and those over age 34 were 23.4 and 15.6 percent, respectively. Males aged 18 to 25 were also more likely to be recent smokeless tobacco users; more than 10 percent had used it during the month prior to the survey, while an additional 16 percent had used it at some time before the past month (NHSDA estimates).

Alcohol Use in the Late 1990s.

An estimated 80 percent of high school seniors have consumed ALCOHOL at least once. About 74 percent had consumed alcoholic beverages in the year prior to the survey, and 51 percent had done so during the month prior to the survey. About 3.4 percent had become daily drinkers (MF estimates).

An estimated 62.3 percent of high school seniors had been drunk at least once—almost 53 percent during the year prior to the survey and almost 33 percent during the month prior to the survey. About 3.4 percent reported having become daily drinkers (MF estimates).

Among household residents aged 12 and older, an estimated 80 to 82 percent have consumed alcoholic beverages; this represents from 174,928,000 to 179,975,000 individuals. During the month prior to the survey, an estimated 51 percent had consumed alcohol. As might be expected, the prevalence values for 18- to 25-year-olds were somewhat higher than they were for the high school seniors, especially in relation to recent drinking: Almost 60 percent of the 18- to 25-year-olds had consumed alcoholic beverages during the month prior to the survey. The values for 12- to 17-year-olds were lower: About 37 percent in this age group had tried alcoholic beverages at least once, and about 19 percent had consumed alcohol during the month prior to the survey (NHSDA estimates).

An estimated 22.4 percent of respondents of all age groups from 12 years upward reported drinking at least once per week or more during the year prior to the survey. Corresponding estimates for respondents aged 12 to 17, 18 to 25, 26 to 34, and 35 + were 4.6, 24.5, 23.8, and 24.6 percent, respectively (NHSDA estimates).

Alcohol dependence was found to have affected 15 percent of those who had consumed alcoholic beverages: Out of every six or seven persons who had tried alcohol, about one had become dependent on alcohol. In relation to the total survey population that included drinkers as well as abstainers, an estimated 14 percent were found to qualify for the diagnosis of drug dependence, according to the American Psychiatric Association's criteria (NCS estimates).

Other Illicit Drug Use in the Late 1990s.

When controlled substances such as MARIJUANA, cocaine, and heroin, as well as INHALANT drugs, were considered, it was found that an estimated 55 percent of respondents had used these drugs on at least once occasion, 42 percent during the year prior to the survey. About 26 percent had taken one or more of these drugs during the month prior to the survey (MF estimates).

The National Household Survey on Drug Abuse reported that an estimated 34 to 37 percent of the population aged 12 and older had engaged in illicit drug use at lease once: this amounts to about 75 to 81 million drug takers. The number of recently active drug takers was lower; they represented 6 to 7 percent of the population (NHSDA estimates).

According to the National Comorbidity Survey estimates, out of every seven persons who had tried marijuana, cocaine, or other controlled substances and inhalant drugs, one had developed drug dependence (14.7%). In light of the fact that about 51 percent of this survey population of 15- to 54-year-olds reported a history of illicit drug use, the resulting estimate for the prevalence of dependence on controlled substances was 7.5 percent. That is, in the total population of individuals (including both drug users and never users), about one in fourteen had fulfilled the criteria for drug dependence (NCS estimates).

Cannabis Use in the Late 1990s.

An estimated 50 percent of high school seniors had tried marijuana or HASHISH (Cannabis) on at least one occasion, and about 38 percent had smoked cannabis during the year prior to the survey. An estimated 23 percent had smoked cannabis during the month prior to the survey, and an estimated 6 percent reported daily cannabis use (MF estimates).

Within the age ranges of 12 to 17 and among persons aged 35 and older, there are many individuals who have not yet started to use illicit drugs such as cannabis, as well as many others who never will start to use these drugs. As a result, one might expect lower prevalence values in these age groups as compared to the values for other age ranges. In fact, this is precisely what the national survey estimates indicate. Overall, an estimated 32 to 34 percent of respondents reported having tried cannabis, but among 12- to 17-year-olds the estimate was only 18.9 percent, and among those aged 35 years and older it was 29.4 percent. Prevalence of cannabis use was most common among 26- to 34-year-olds (47.9%) and among 18- to 25-year-olds (44.6%). This also was true for recent cannabis use during the month prior to the survey: There was a prevalence of 5.0 percent for the population overall, 8.3 percent for 12- to 17-year-olds, 13.8 percent for 18- to 25-year-olds, 5.5 percent for 26- to 34-year-olds, and 2.5 percent for older adults (NHSDA estimates).

Among cannabis users, about 9 percent were found to have developed cannabis dependence. Among all 15- to 54-year-olds (including both users and never users), 4.2 percent had become dependent on cannabis (NCS estimates).

Inhalant Use in the Late 1990s.

INHALANTS had been used by an estimated 15 percent of high school seniors—about 6 percent within the year prior to the survey and about 2 percent during the month prior to the survey. Very few respondents (well under 1 percent) reported daily inhalant use (MF estimates).

The National Household Survey on Drug Abuse indicated that about 5.8 percent of its survey population had tried inhalants at least once; about 1 percent had done so during the year prior to the survey, and from 0.3 to 0.4 percent had used these drugs during the month prior to the survey. It was found, when considering age and sex, that the subgroup most likely to have used inhalant drugs during the month prior to the survey was that of males aged 18 to 25; in this group, 1.9 percent reported recently active inhalant use (NHSDA estimates).

An estimated 2.3 to 5.1 percent of the inhalant users have been found to qualify for the diagnosis of dependence on inhalant drugs. Translated into an overall prevalence estimate for both users and nonusers, this amounts to about 0.3 percent prevalence of inhalant dependence in the total survey population (NCS estimates).

Use of Psychedelic Drugs in the Late 1990s.

PSYCHEDELIC drugs (primarily LYSERGIC ACID DIETHYMIDE, or LSD) had been used by an estimated 14 percent of high school seniors. Almost two-thirds of these users (9.4%) had used them in the year prior to the survey, and about one-quarter (3.5%) had used them during the month prior to the survey. PHENCYCLIDINE (PCP) users were in the minority within this group of drug users; only 3.4 percent of the high school seniors had ever tried PCP (MF estimates).

Among persons aged 12 years and older, from 9.1 to 10.7 percent of individuals had tried psychedelic drugs such as LSD, but for the most part these drug experiences were not recent: Only 0.5 to 0.9 percent reported taking psychedelic drugs during the month prior to the survey. Peak prevalence values for recent use of the psychedelic drugs were observed in the years of adolescence and early adulthood; only for 12- to 17-year-olds and 18- to 25-year-olds did these values exceed a threshold of 1 percent (1.8 and 2.7%, respectively); otherwise, they were at the 0.4 percent level or lower (NHSDA estimates).

About 5 percent of the users of psychedelic drugs were found to qualify for the diagnosis of a dependence syndrome, defined in relation to the American Psychiatric Association criteria. Thus, about 0.5 percent of the survey population of 15- to 54-year-olds had become dependent on psychedelic drugs.

Cocaine Use in the Late 1990s.

Among high school seniors, an estimated 9.8 percent had tried cocaine; within this group of COCAINE users, roughly one-half had tried CRACK-cocaine. About 6 percent of high school seniors had used cocaine (including crack) during the year prior to the survey, and just over 2.6 percent had used it in the month prior to the survey. In the MF sample of about 16,000 high school seniors, daily cocaine smoking was too rare to estimate precisely (MF estimates).

An estimated 10 to 11 percent of the National Household Survey's population reported having tried cocaine or crack smoking (or both) at least once. The corresponding value for 12- to 17-year-olds was only 2.2 percent, and there was age-related variation: 10.0 percent of the 18- to 25-year-olds had taken cocaine (including crack); 17.1 percent of the 26- to 34-year-olds had done so, and the prevalence estimate for older adults was 10.4 percent. Translated into absolute numbers, an estimated 21 to 25 million Americans aged 12 and older had tried cocaine or crack smoking. Recent use was substantially less common: Only 0.7 to 1.0 percent of the survey population reported having used these drugs during the month prior to the survey; this represented about 1.4 to 2.1 million recently active cocaine users in the survey population.

By the early 1990s, the second American epidemic of cocaine use had peaked and waned. Crack smoking had sustained the epidemic for a time, but in the early 1990s it became clear that crack smoking had not diffused broadly through the U.S. population. The relatively low prevalence values for crack smoking among high school seniors was reflected in the National Household Survey on Drug Abuse, which found that only 1.8 to 2.3 percent of its survey population had tried crack smoking; this amounted to 3.9 to 5.1 million individuals. The age groups with most crack-smoking experience were the 18- to 25-year-olds, with a prevalence value of 2.7 percent, and the 26- to 34-year-olds, with a prevalence value of 3.9 percent. Prevalence of crack smoking during the month prior to the 1998 survey was uniformly under 1 percent for all age and sex groups under study (NHSDA estimates).

For every six individuals who had tried cocaine at least once, one had developed cocaine dependence. That is, among these cocaine users, an estimated 15.2 to 18.2 percent had become sufficiently dependent upon cocaine to qualify for the American Psychiatric Association diagnosis. In relation to all persons in the survey population, whether they had tried cocaine or not, an estimated 2.7 percent qualified for the diagnosis of cocaine dependence (NCS estimates).

Use of Non-Cocaine Stimulants in the Late 1990s.

The nonmedical use of stimulants other than cocaine (such as AMPHETAMINES) was actually more prevalent than cocaine use among high school seniors. An estimated 16.3 percent of high school seniors had taken these stimulant drugs without any doctor's orders; 10 percent had done so in the year prior to the survey, and 4.5 percent had done so during the month prior to the survey. Methamphetamine or "ice" smoking reemerged among youth in the 1990s. Among high school seniors, 4.8 percent had ever tried "ice," 1.9 percent had done so in the year prior to the survey, and 0.8 percent had used during the prior month (MF estimates).

For reasons not well understood, the Monitoring the Future sample of high school seniors yields prevalence estimates for non-cocaine stimulant usage that are considerably larger than corresponding estimates from the national household survey. Overall, the household survey population estimate for nonmedical use of these stimulant drugs was 4.4 percent, and the age group with the highest prevalence value was that made up of 26- to 34-year-olds, at 5.1 percent. Nonetheless, within the survey population, recent use of the stimulant drugs was found to be 3.2 to 4.9 percent for the 18- to 25-year-olds, the age group whose level of use most resembled that of the high school seniors (NHSDA estimates).

Slightly more than 11 percent of the persons who had used these stimulant drugs were found to have become dependent on them. This number of stimulant-dependence cases represents about 1.7 percent of all persons in the survey population aged 15 to 54 (NCS estimates).

Use of Anxiolytic, Sedative, and Hypnotic Drugs in the Late 1990s.

About 9 percent of high school seniors had used tranquilizers (anxiolytic) or SEDATIVE-HYPNOTIC (e.g., BARBITURATE) drugs without a doctor's orders. About 5.8 percent had done so during the year prior to the survey, and 2.5 percent had done so during the month prior to the survey (MF estimates).

About 3 to 4 percent of the national household survey population reported nonmedical use of tranquilizers or anxiolytic drugs, while 2 to 3 percent reported nonmedical use of sedative-hypnotic drugs without a doctor's orders. For tranquilizers, this amounted to 6.8 to 8.8 millions of nonmedical users. For sedative-hypnotics, the total was 4.0 to 5.4 millions of nonmedical users. The estimated number of recently active users was less substantial; they represented less than 0.5 percent of the survey population for tranquilizers (under 1 million nonmedical users) and for the sedative-hypnotics (under 500,000 nonmedical users).

Grouping the users of the tranquilizer or anxiolytic drugs together with the users of the sedative and hypnotic drugs, the National Comorbidity Survey team found that about 9 percent of these drug users had become dependent on them. In considering this prevalence value, it is important to note that in this survey nonmedical drug use was defined to include not only use of the drug to get high, but also taking more of the drug than was prescribed or in ways not consistent with accepted medical practice. Overall, the prevalence of dependence on these drugs was at a level of 1.2 percent in the survey population (NCS estimates).

EPIDEMIOLOGY OF DRUG USE AND DRUG DEPENDENCE OUTSIDE THE UNITED STATES

Each year, the United States allocates more resources to epidemiologic surveys of drug use than does any other country in the world. For this reason, it has been possible to assemble a wealth of epidemiologic survey data on the prevalence of drug use and drug dependence within the United States. Other countries also have conducted surveys of this type and have produced valuable evidence about their experience with tobacco, alcohol, and other drugs. (See the bibliography for some references that can be consulted to gain more information about the results of these surveys.)

OTHER ASPECTS OF EPIDEMIOLOGY AS APPLIED TO DRUG USE AND DRUG DEPENDENCE

A broad range of research questions must be answered in order to gain a complete understanding of the epidemiology of drug use and drug dependence. The focus in this article has been on quantity: How many people in the population (or what proportion) have been affected by drug use and by drug dependence? Although many epidemiologists now devote their research careers to surveys that are needed to answer this kind of basic question, more stress ought to be placed on the other central questions for epidemiology, especially when the answers to these questions can guide society toward effective strategies for prevention of drug use and drug dependence. These questions are:

Where in the population are the affected cases located (in which subgroups, in which places, during which seasons, years, or epochs)? This is a question of location.

What accounts for some people becoming affected, whereas others do not become affected? This is a question about CAUSES.

By what processes or sequence of conditions do people become dependent on drugs? This is a question about mechanisms and linked sequences of causal conditions.

What can we do to prevent and reduce the suffering? This is a question about prevention and amelioration.

At its best, epidemiology provides critically important answers to each of these questions, and it works to ensure that new findings are translated rapidly into effective strategies for prevention. This is the future agenda for epidemiologic research on drug use and drug dependence.

BIBLIOGRAPHY

ANTHONY, J. C. (1995). International databases on drug dependence. In L. Eisenberg & R. DesJarlais (Eds.), International behavioral and mental health: A sourcebook. Cambridge: Oxford University Press.

ANTHONY, J. C., & HELZER, J. E. (1995). Epidemiology of drug dependency. In M. Tsuang, M. Tohen, & G. Zahner (Eds.). Textbook of psychiatric epidemiology. New York: John Wiley and Sons.

ANTHONY, J. C., WARNER, L. A., & KESSLER, R. C. (1994). Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalant drugs: Basic findings from the National Comorbidity Survey. Experimental and Clinical Psychopharmacology, 2, 1-24.

HELZER, J. E., ET AL. (1990). Alcoholism—North America and Asia. Archives of General Psychiatry, 47, 313-319.

JOHNSTON, L. D., O' MALLEY, P. M., & BACHMAN, J. G. (2000). National survey results on drug use from the Monitoring the Future Study, Overview of Key Findings, 1999. NIH Publication no. 00-4690. Rockville, MD: National Institute on Drug Abuse.

MURRELLE, L. (1990). Epidemiologic report on the use and abuse of psychoactive substances in 16 countries of Latin America and the Caribbean. In Drug abuse, World Health Organization, Pan American Health Organization Scientific Publication No. 522. Washington, DC: Pan American Health Organization.

PETERSEN, R. C. (ED.). (1978). The international challenge of drug abuse. National Institute on Drug Abuse Research Monograph. DHEW Publication No. ADM-78-654. Washington, DC: U.S. Government Printing Office.

U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. (1999). National Household Survey on Drug Abuse: Population estimates 1998. DHHS Publication No. (SMA) 99-3327. Rockville, MD: Substance Abuse and Mental Health Services Administration.

JAMES C. ANTHONY

MARSHA F. ROSENBERG