APPENDIX C

A Survey of Previous Research

For some drugs the practice of distinguishing among various using patterns is well established. Social drinking has been distinguished from alcoholism, for example, even if somewhat imprecisely. But for marihuana, psychedelics, and opiates, the importance of making such distinctions has only recently been recognized. A literature search using the computerized MEDLINE File (which catalogues articles from more than 1,000 journals) revealed only one study specifically concerned with occasional illicit drug use between January 1969 and November 1972. Until recently, research studies have tended to reflect in three related ways the reigning cultural belief that all illicit drug use is abusive.

First, many if not most investigators have not differentiated between moderate and excessive use. Margaret Heller, in her 1972 review of thirtyfive studies of opiate use that had been published since 1954, concluded that their "most serious" flaw was that they "lumped together all drug users without considering the extent of their use."

Second, what few attempts were made to define the variety of drug-using styles only resulted in a proliferation of vague and confusing terms. This was most evident in regard to moderate patterns of use-ranging all the way from near-abstinence to near-compulsion-but it was also apparent in descriptions of extreme use (usually heroin addiction), which had been studied more carefully and intensively. The most inclusive terms, like "illicit drug use" and nonmedical drug use," were too generic to have much descriptive value. More specific terms-such as "chipper," "experimenter," or "head"; "occasional," "habitual," "chronic," "heavy,`- "ritual," "recreational," "addicted," "committed," "casual," or "regular user"; "drug-dependent person"; and "abuser"-were employed to indicate the intensity and frequency of use or to describe some characteristic of the user. These terms, however, were applied so casually and inconsistently that, as the National Commission on Marihuana and Drug Abuse reported (1973), "It is often difficult to ascertain who is being described, what kind of behavior is being evaluated ... or what actual or potential risks of such behavior are likely." Isidore Chein et al. (1964) found that even the apparently precise term "heroin addict" was used in so many different ways that it was "meaningless to identify an individual as an addict."' Bruce D. Johnson (1977) appropriately described "addict" as an accordion word which expands or contracts to please the audience being played to" and which was sometimes defined "so broadly as to include anyone who was ever a daily heroin user."

Third, the few investigators who acknowledged the existence of moderate or noncompulsive using styles concentrated on marihuana and neglected the psychedelics and opiates, apparently because of the prevasive popular belief that the drug variable-that is, the chemistry of each drug-was the main determinant of the outcome of use. For example, sociologist Erich Goode (1969) defined the occasional marihuana smoker as "someone comparable to the social drinker, to whom marihuana is a pleasant but largely dispensable and not particularly important aspect of life," and he suspected that such benign using patterns predominated among marihuana users. Further, Herbert Blumer et al. (1967) noted the possibility of continued noncompulsive use of marihuana but believed that heroin could not be used on an occasional basis because it developed greater dependency. This failure to discern and describe moderate patterns of psychedelic or opiate use might be understandable if the ratio of moderate to compulsive users of these two types of drugs had been found to be low relative to the ratio for marihuana use. But satisfactory data were not yet available on those ratios; in fact, surveys of illicit drug use in the general population were not made until relatively recently. According to the National Commission on Marihuana and Drug Abuse (1973), no such studies were undertaken prior to 1965, and even though several were available by the mid-1970s, many of these had major shortcomings. "Most were studies of young populations . . . concerned with marihuana at the expense of both other illicit drugs and alcohol . . . . Simple measures of `ever used' and 'present use' were obtained and finer distinctions were seldom made."

This failure of research studies to differentiate between moderate and excessive drug use, to develop a precise and accurate terminology, and to describe the moderate use of psychedelics and opiates was strikingly evident in 1972, when my DAC study was launched. Because that study and its successor (sponsored by NIDA), and therefore this book, place the main emphasis on the use of opiates, the following review of the research literature published between 1957 and 1981 focuses on studies of opiate use. It shows that although a growing number of investigators are now seriously considering the differences among various patterns of use, there is still a relative lack of information about nonaddictive use. Hundreds of studies deal with some aspect of addiction, but fewer than a dozen give substantial consideration to occasional use. In addition, a large number of the articles describing these studies deal solely with the "captive" or medical population of treatment centers while only a few are concerned with the nonmedical populations of users "at large."

Factors Inhibiting the Study of Occasional Opiate Use

There are at least four reasons why nonaddictive and nontreatment opiate use have not been adequately studied.

The first is related to the persistence of cultural beliefs that heroin is an inherently "bad" drug. Joseph R. Gusfield (1975) has noted that because researchers are not immune to the cultural atmosphere in which they work, the prevailing attitudes toward heroin use may have made it difficult for them to appreciate that much of the conventional wisdom contained in the literature was inaccurate. The tendency to regard any kind of heroin user as a "junkie" made it difficult for investigators to delineate groups that did not fit that cultural stereotype or to recognize how valuable the study of such groups could be.

A second, and probably more important, reason is that noncompulsive and therefore nontreatment users are very difficult to locate and study (Geber 1969; Catton & Shain 1976; Rittenhouse 1977). Opiate use is still so rare that the identification of users of any kind in the general population is a very expensive task. It is generally agreed that fewer than 1% of Americans are nonmedical opiate users: hence 10.000 persons would have to be interviewed to generate a representative sample of only 100 users, and only some of those would use the drug on an occasional basis. It is true, of course, that data on opiate users can be gathered efficiently from treatment, law-enforcement, correctional, and other captive populations. But such sources provide a highly selective sample of users who "have encountered significant personal, medical, social, or legal problems in conjunction with their drug use, and thus often represent the most pathological end of the using spectrum" (Greene, Nightingale & DuPont 1975).

Third, questions of ethical responsibility have probably impeded the study of nonaddictive use. Some researchers, aware of the popular belief that any use of heroin results in depravity and enslavement and that this belief undoubtedly serves to discourage heroin use, may have been reluctant to undertake the study of occasional use lest they undermine that attitude. Furthermore, some investigators may have feared that if they reported the existence of occasional users they might be accused of endorsing experimentation with opiates (Zinberg 1881). These. ethical concerns have, of course, troubled me as well.

Fourth, the widespread confusion of terms relating to the level and quality of opiate use has been a substantial obstacle to research (Lasagna 1965; Smart 1974; Robins 1977; Apsler 1978). This confusion has not only complicated the comparison and interpretation of findings from different studies but has prompted at least one investigator to ask: "Is the occasional user of 'dangerous' or narcotic drugs an addict? Is the habitual user of heroin which is so 'cut' as to be incapable of creating serious tolerance or physical dependence addicted'? Is a person who has been off drugs for years and then takes one shot of heroin `in relapse'?" (Lasagna 1965).

In the literature the terms "occasional use" and "chipping" are most often used to indicate a nonaddictive pattern of opiate use, but they do not necessarily connote moderate use. Rather, they tend to indicate less than daily use and the absence of physical dependence. Further, they do not necessarily imply anything about the consistency of use over time or about the frequency or quality of the concurrent use of other drugs.

The Emergence of Research Interest in Occasional Opiate Use

Alfred Lindesmith (1947), a pioneer in the study of heroin, was one of the first twentieth-century researchers to refer to nonaddictive use. He defined the "pleasure user" or "joy popper" as follows: "A `joy popper' is simply an individual who uses the drug intermittently and who has never been `hooked"' (original emphasis). Having only a few examples of joy poppers and no long term data about them, Lindesmith speculated that most such users eventually became addicted. He did, however, emphasize the uniqueness of this group: "As long as they indulge in sufficient moderation to avoid severe withdrawal symptoms, they are sharply distinguished from those who have been `hooked' and they do not regard themselves as addicts."

Jordan M. Scher (1961, 1966), through his work at the Cook County (Illinois) Narcotics Court and jail, came in contact with "addicts" who had confined their use to weekends or parties. Although he felt that this pattern often progressed to more frequent use, he observed several cases "in which relatively gainful and steady employment has been maintained for two or three years while the user was on what might be called a regulated or controlled habit" (original emphasis).

Isidore Chein et al. (1964) stated that they had been able to establish the existence of "long continued nonaddictive users" of heroin, but not in sufficient numbers to permit a study of the type they wished to undertake.

When Harold Alksne, L. Lieberman and L. Brill discussed a model of the life cycle of addiction in 1967, they indicated that some drug-takers seemed able to "control their use" and to continue in occasional or limited use for "an indefinite period of time." The authors noted that "no research reports are available for this kind of use," and they felt it likely that these "experimental" users would either become more involved with opiates or stop using altogether. At the same time, however, they expressed the need for future research to document their theory.

Unlike many of the earlier investigators of occasional use, Erich Goode (1972) felt that "the occasional (weekend) heroin user is probably a good deal more common than most of us realize." He did not differ from most other investigators, however, in his view that the majority of occasional heroin users eventually become addicted. He based this opinion on the lack of cultural norms (which I call sanctions and rituals) to limit heroin use, the rapidity with which a user could become physiologically addicted, and the tendency of new users to discount the risk of becoming addicted. In elaborating the last point he referred to Alan G. Sutter (1969), who had argued that the "magical belief' that one could "chippy around" and still control heroin use was a self-deception that contributed to the development of addiction. "If a person knows from experience that he has always been able to control his drug use and still 'take care of business,' he will be convinced of his strong 'will power' and will believe that only 'weak-minded people get hooked.' This belief lays the groundwork for addiction" (original emphasis).

In view of these repeated declarations that in a majority of cases occasional use quickly leads to addiction, it is not surprising that the first study to focus specifically on occasional opiate use did not appear till 1973, when Douglas H. Powell published data on twelve occasional users who had responded to his newspaper advertisements. He concluded that these subjects had never become addicted even though they had all been using heroin for at least three consecutive years. Some of them, however, appear to have followed patterns of opiate use that, according to the standards of my project, could hardly be defined as "well controlled": one, for example, had used heroin for "six weeks straight" within the previous six months. Nor did Powell clearly describe the use of other intoxicants by the group. Nevertheless, his work established the existence of patterns of opiate use that differed significantly from the classic model of intensive and destructive heroin consumption. His investigation, which legitimized further studies (including my own) that were aimed at describing these alternative patterns, touched off increased research interest in occasional opiate use.

Occasional Opiate Users in Treatment Centers

The fact that occasional users turn up and are identified in treatment populations has raised concern about the propriety of giving treatment to such users. Guidelines suggested by the American Medical Association (1972), which state that the "mere" use of an opiate, "even if periodic or intermittent ... cannot be equated with drug dependency," recommend that a specific medical diagnosis be made in each case. W. H. Dobbs (1971) warned that some people applying for methadone maintenance ought to be rejected because they were only occasional users. Paul H. Blachly (1973) demonstrated that Naloxone, an opiate antagonist, could be used, unlike urinalysis, to distinguish a person who was physically dependent from an occasional user, and he suggested that Naloxone should be administered for this purpose to those who applied for methadone maintenance. At one methadone clinic, he found that one-third of the applicants, who had no prior documentation of withdrawal in an institution, "showed no evidence of physical dependence"; and he concluded that there was "a significant hazard" that the treatment would addict these occasional users to methadone.

When Charles P. O'Brien (1976) administered Naloxone to applicants for methadone maintenance, he found that 55% gave positive responses for opiate dependence, 18% were weakly positive, and 15% were negative. The remaining 12% were not given Naloxone because they were "found not to be using heroin daily." Frederick B. Glaser (1978) suggested that only those who gave positive responses (55%) in O'Brien's study were definitely addicted and argued that the remaining 45% should have been regarded as addicted.

William L. Minkowski, R. C. Weiss and G. A. Heidbreder (1972) found more occasional heroin users than daily users in a random sample of clients at a Los Angeles Health Department youth clinic. Replies to 300 questionnaires and interviews with 100 subjects indicated that 12% had used it once a month or less, 1.1% had used it two to four times a month, and 1.2% every weekend or more; only 1.7% had used daily.

John Newmeyer (1974) of the Haight-Ashbury Free Medical Clinic also reported that individuals who could be "characterized as `persons who sample heroin without becoming addicted"' self-refer for drug-related treatment.

In a study of 2,750 subjects aged eighteen or under who had received treatment in a drug program, Yoav Santo, E. C. Farley and A. S. Freedman (1980) reported that whereas 6.7% had "ever" used heroin within the threemonth period prior to admission, only 4.3% repeated "weekly or more frequent use for at least four weeks" during the same period. Use of other opiates followed a similar pattern.

Occasional users may suffer from the dangers to health that are usually associated with heroin addiction. Edward Kersh (1974) found that 70% of the patients treated for narcotic overdoses in a particular New York hospital emergency room were occasional users. R. W. Light and T. R. Dunham (1974) commented on vertebral osteomyelitis due to septic intravenous administration of heroin in two individuals who had not used heroin for at least eight weeks preceding the onset of symptoms and who were "definitely not addicts." Roger Lewis, S. Gorbach and P. Altner (1972) reported on five occasional users they had encountered in one year who had spinal chondroosteomyelitis. These patients had all used heroin intravenously within the week prior to their hospital admission, but they were not addicted.

Marco Systems, Inc. conducted a study (1975) of admissions to New York City treatment programs in 1971 and found that approximately one-third of the sample had been categorized as less than daily users for the two months preceding treatment.

S. B. Sells (1977) has studied national data based on 27,460 drug admissions to a variety of treatment programs between 1 June 1969 and 31 March 1974. During this more than four-year period, from 8% to 12% of the patient population reported that they had been less than daily opiate users during the two months preceding treatment.

These studies raise the question, "Why do occasional users enter treatment?" There are many possible answers, each of which may be valid in specific cases.

1. As some of these studies suggest (Lewis, Gorbach & Altner 1972; Kersh 1974; Light & Dunham 1974; Macro Systems 1975), occasional users who are suffering from health problems associated with their drug use may self-refer for treatment. However moderate the use, the impurities in street heroin and the injection of opiates pose a serious risk of infection. In addition there is a risk of overdose when drugs of uncertain potency are used.

2. Although many studies report "nonaddicted" and "less than daily" users among their subjects, it does not necessarily follow that such use is nonabusive. Use in amounts and frequencies below those required for addiction may still interfere with work, marriage, and general functioning and thus may require treatment.

3. Occasional users may apply for treatment to help with personal problems that do not necessarily result from their use of drugs. For example, of the 2,750 adolescents who had been in one drug treatment program, almost half stated that they had applied for admission only to get help for personal problems (Santo, Farley & Friedman 1980).

4. At least some users report for treatment who believe they are addicted but in fact are not. O'Brien (1976) found that such users "became insulted if they were told they were not really dependent on drugs." These users may subscribe to the cultural mythology that even infrequent use of heroin will result in addiction. Alternatively, as articles by Frederick B. Glaser (1974) and by B. J. Primm and P. E. Bath (1973) suggest, some applicants who have used heroin frequently have not become physically dependent because of the low potency of the street drug.

5. For some occasional users (probably a small percentage), participation in treatment may provide the entree to a deviant drug-using subculture. Association with addicts in treatment offers the occasional user a new means of obtaining illicit drugs; in addition, a sense of belonging and the status associated with membership in a deviant group may be considered important rewards (Gay, Senay & Newmeyer 1974). A relevant case is reported in Zinberg and Lewis (1964): A twenty-four year-old man who had been hospitalized for hepatitis and who claimed to be an addict finally confessed to a house officer that he had taken heroin only twice, but he cautioned the physician "not to tell his wife because he pretended to her and to their `cool' friends that he was 'hooked."'

6. Many occasional users are coerced into treatment. Not infrequently, lawyers and "sympathetic" probation officers propose that the user enroll in a treatment program prior to criminal prosecution as a means of getting a lighter sentence. Dan Waldorf (1973) believed that as much as 27% of his treatment sample had not been addicted prior to treatment because "most of the sample had been committed to treatment under New York State civil commitment laws (89%), and actual physical addiction is not a necessary condition for treatment."

7. Finally, admission of nonaddicts may be encouraged by treatment program recruiters. Treatment programs have an obvious interest in maintaining high enrollments because levels of funding are usually linked to the size of their patient population.

Treatment Data as a Measure of Occasional Use

If it could be assumed that all heroin users eventually seek treatment, data like those just considered would obviously be representative of the entire population of users and so would provide workable estimates of the number of occasional users. But in fact there is considerable evidence that many users apply for treatment.

A survey of heroin users in Wyoming identified a significant number of addicts who were unknown to the police and to the drug treatment centers of the community health facilities (Bourne, Hunt & Vogt 1975; Hunt 1977). Stuart Nightingale (1977) has also reported that according to national estimates, there are more opiate users out of treatment than in treatment: at any particular time, while 170,000 persons are in treatment for opiate addiction and 100,000 addicts are in jail, "another 300,000 to 400,000 are not in treatment," and "the majority . . . have never been in treatment."

In a smaller area Carl D. Chambers and John A. Inciardi (1972) conducted a study of active heroin addicts in Bedford-Stuyvesant, Brooklyn (New York), who had not been in treatment or in jail for six months prior to the interview. Analysis of the demographic characteristics of the ninety-five respondents "revealed nothing which would suggest that they were significantly different from the universe of `heroin street addicts' in New York City." Seventy percent of them were males, and of these, 40% were black, 32% white, and 22% Puerto Rican. Their ages ranged from seventeen to forty-six (median age, twenty-four). All were reported to be addicted; they had used a median of four "bags" a day for periods from at least nine months to as long as twenty-three years. Yet almost half of the sample had never been treated for drug use.

In an epidemiological Chicago street study, Patrick H. Hughes et al. (1971) found that addicts occupying dealership roles were especially difficult to "involve in treatment and if involved, were unlikely to remain over time." By contrast, addicts who retained jobs were likely to participate and remain in treatment, usually for at least six months. A third category, "hustling addicts," fell between these two groups in terms of the likelihood of entering and remaining in treatment.

Andrew T. Weil (1977) discussed "stable addicts" he had encountered in San Francisco who were "hidden" from the attention of the social authorities. He described them as suburban working-class whites who were steadily employed and who purchased drugs with money they had earned. Their life-style was quite different from that of the stereotypic heroin addict. Some injected heroin once a day, others did it both morning and evening, and most "kept up these patterns for years."

It appears also that underrepresentation of the using population in treatment samples applies to both younger and older addicts. James V. DeLong (1977) has noted that the current approaches to treatment fail to attract a large number of the younger addicts. The reason for this, he says, is that most young addicts either are not "sufficiently disenchanted with either the lifestyle or the effect of heroin" to apply for treatment or "are as yet unconvinced that they are truly hooked." Similar conclusions have been drawn with respect to older users. Charles Winick (1974) reports that a "substantial proportion" of the 1,300 Metropolitan New York heroin users who came to the attention of correctional or treatment authorities "gave every indication of indefinitely continuing drug use." Earlier, W. C. Capel et al. (1977) had interviewed thirty-eight opiate users in New Orleans between the ages of forty-five and seventy-five, many of whom "have not stopped using opiates, but whose lifestyle has exhibited a protective adaptation that diverts attention from their illicit activity." Fewer than half (47%) of these older subjects used daily; the remainder reported various weekly patterns of use. Many had eluded arrest for years (some had never been arrested), and none were currently in treatment. Again, the authors concluded that neither methadone maintenance nor drug-free treatment groups appeared "to be attractive to men of this age group.

These studies support the commonsense view that treatment populations, which include fewer occasional users than addicts, underrepresent the at-large, active drug population. Although several reasons were given in the last section to explain why occasional users might come to treatment, it can be deduced from the data presented here that the proportion of occasional users who actually do enter treatment is much smaller than the proportion of heavy users (addicts) who enter treatment. To counteract this underrepresentation of occasional users in the treatment data, it is necessary to turn to another type of source material-data drawn from nontreatment (noninstitutionalized) samples.

Occasional Noninstitutionalized Users

Much of the data from studies of noninstitutionalized opiate users show that the proportion of occasional users in the community at large is greater than that in treatment populations. With one exception, these studies also reveal a higher proportion of occasional users than of addicts among users who are not in treatment. The study that constitutes the exception is one of the earliest surveys of drug use in a normal population, a sample of young Negro males in St. Louis. In that study, published in 1967, Lee N. Robins and G. E. Murphy found that of the 13% who had tried heroin, 1o% had become addicted. In fact, no one who had used the drug more than six times had avoided addiction. Thus the St. Louis study indicated that nonaddictive use among blacks was infrequent between the 1930s and the 1950s. But since then the same investigator (Robins) found nonaddictive use to be far more commonplace (Robins 1979), and several other studies and surveys of noninstitutionalized users have supported this view.

In 1971 and 1972 Albert Levengood, P. Lowinger and K. Schooff (1973) interviewed sixty single, white male heroin users, aged fifteen to twenty-four, living in a Detroit suburb. Of these, twenty-two (37%) used "regularly or on a daily basis"; twenty-four (40%) used occasionally, with the quantity and frequency of use varying widely; and fourteen (23%) were considered "former users," having taken heroin within the previous year but not during the month preceding interview. The group of "former users" included some subjects who were recent daily users, but no one in that group had received any type of treatment. Levengood, Lowinger and Schooff mentioned that many occasional users had maintained "the same frequency and dosage" for at least one year, but they emphasized that all patterns of use were subject to change.

David S. Nurco et al. (1975) conducted a study of 267 noninstitutionalized males who had been known to the Baltimore police department as narcotic addicts. Eighty-six percent of the sample acknowledged at least one period of addiction. Data on frequency of use indicated that for any one of several time periods, subjects who had used less than daily within the previous month had always outnumbered daily users, constituting at least 56% of all users.

David B. Graeven and Andrew Jones (1977) examined adolescent heroin use in a suburban San Francisco high school between 1966 and 1974. Of the 294 heroin users identified during the study, the proportion not addicted ("experimenters") was almost equal to the proportion who had become addicted: 49% versus 51%. During the period of the study 38% of the experimenters had used heroin only once or twice, 33% had used it 3 to 33 times, 19% 31 to 100 times, and the remaining 10% more than l00 times. In the final year of the study, however, most of the original experimenters (60%) reported no use, 33% were using less than twice a month, and only 7% were using once or twice a week. The addict group had also significantly reduced its heroin use by the end of the study period: 20% reported no use, "8 percent reported modal use of one or two times per month, i6 percent had had modal use of one or two times per week, and 47 percent modal use of three or more times per week."

Perhaps the most prominent research dealing with occasional use is the series of studies of Vietnam veterans by Lee N. Robins (Robins 1973, 1.974; Robins, Davis & Goodwin 1974; Robins, Helzer & Davis 1975; Robins et al. 1979). She interviewed 900 enlisted men twelve to eighteen months before their return from Vietnam in 1971 and then reinterviewed 617 (more than two-thirds) of the same men in. 1974. In 1974 she also collected data on a comparable population of 284 nonveterans who were matched to the veterans in age, eligibility for service, education, and place of residence as of the verterans' dates of induction. Robins found that of all the men who had been addicted to opiates while in Vietnam, only 12% had relapsed into addiction within three years after leaving Vietnam. Still more relevant and surprising is her finding that although half of the veterans who had been addicted in Vietnam had used heroin since their return home, only 12% of them had become readdicted. "Even when heroin was used frequently, that is, more than once a week for a considerable period of time, only half of those who used it frequently became readdicted" (Robins et al. 1979). Data obtained from the matched sample of nonveterans led Robins to conclude also that the low level of addiction observed for veterans after leaving the battle area was consistent with the patterns of use prevailing in the larger U.S. population.

In a survey of noninstitutionalized users of illicit drugs in Ohio (Abt Associates, Inc. 1975), 1.4% of the respondents admitted that they had used heroin, a percentage much higher than any previous estimate. Of these, 8.7% used "several times per week or more," 18.7% "a few times per month," 17.4% "a few times per year," and 65.2% "less frequently."

In a national representative sample of twenty- to thirty-year-olds drawn in 1974 (N = 2,510), John A. O'Donnell et al. (1976) found that only 6% had ever used" heroin (of these most had used on only a few occasions, that is, less than 10 times) and 76% had used less than 99 times (light use). The use of other opiates also tended to be occasional. Although only 31% of the sample had "ever used" opiates, 11% were using them "in a manner not far removed from legitimate medical use." The investigators found that only 20% of the sample used opiates "in a way that could reasonably be seen as abuse" and that only 8% exceeded the level of "experimental" use. Among all nonmedical opiate users, 61% were classified as having been experimenters (fewer than 10 occasions of use), 29% as light users (10 to 99 occasions), and only 10% as heavy users (100 or more occasions).

Annual survey data drawn from large samples of high-school seniors more than 15,000 students in the classes of 1975 through 1979-also suggest that occasional opiate use predominated over heavy use throughout the entire five-year period (Johnston, Bachman & O'Malley 1982). For example, in 1979 0.5% of seniors used heroin in a thirty-day period, but the percentage who used daily was less than 0.1. Data on the number of using occasions, whether for the previous thirty days, twelve months, or a lifetime, also indicated that occasional use was more common than intensive use. For example, lifetime data for seniors in 1975, 1976, and 1977 show that approximately 1.6% of the sample had used opiates on from 1 to 19 occasions compared with 0.2% (at the most) who had used on more than 19 occasions.

In 1971 and 1972 two national drug surveys were completed by the Response Analysis Corporation for the National Commission on Marihuana and Drug Abuse. Since 1974, similar surveys have been made jointly by the Response Analysis Corporation and the Social Research Group at George Washington University, under the sponsorship of the National Institute on Drug Abuse (NIDA) (Cisin, Miller & Harrell 1979; Fishburne, Abelson & Cisin 198o; Miller & Associates 1983). All these surveys have also yielded data showing that occasional opiate use is more common than heavy use. The pre-1974 surveys, for instance, indicated that among students who had used opiates, most had used less than once a week (75% of high school users and 98% of college users) (National Commission on Marihuana and Drug Abuse 1973). According to a 1979 study of young adults aged eighteen to twenty-five, unselected for drug use, 2.5% reported on interview that they had used heroin less than ii times, compared with 1.3% who had used ii or more times (Fishburne, Abelson & Cisin 1980).

In 1975 survey data gathered on a regional basis from 1,426 youths living in Detroit, Baltimore, Cincinnati, and Providence also indicated that occasional use was relatively common (Watkins & McCoy 198o). This sample was divided into two groups for the purpose of analyzing ethnic differences in drug use: 31% of the subjects had been born in Appalachia or had "a relative no further back than a grandparent who had lived in rural Appalachia"; the other 69% were designated "non-Appalachian." Of the 17.1% of Appalachian youths who had used heroin and other opiates in the two months before being interviewed, almost 50% had used only once a week. Of the 8% in the nonAppalachian group who were current opiate users, an even larger proportion (59%) used only once a week.

In contrast to the one early survey that found a high rate of addiction among young Negroes in St. Louis (Robins & Murphy 1967), all the later data on the general (noninstitutionalized) population suggest that at any one time those who use opiates occasionally may constitute 40% or more of the total opiate-using population. One estimate based on recapture techniques goes far beyond that figure, claiming that nonaddicts outnumber addicts by as much as ten to one (Hunt & Chambers 1976; Hunt 1977, 1982). Even on the basis of combined data for populations in treatment and out of treatment, it is reasonable to conclude that the number of nonaddicted opiate users both in and out of institutions is roughly equal to or greater than the number of addicts. Although not all researchers would agree with this view, many of those who would not agree would admit that occasional use is not so rare as they once supposed. For example, in reviewing papers dealing with the epidemiology of heroin and other narcotics, Jane D. Rittenhouse (1977) has stated that "increasing evidence has shown that self-reported mild use without social and health problems is not uncommon." Similarly, Bruce D. Johnson (1977) concluded that the typical addict "is involved in relatively brief addiction episodes . . . interspersed with voluntary abstinence, irregular consumption, institutionalization, or admission to treatment."

Unresolved Questions about Occasional Use

Although in the past decade a great mass of data has been gathered about occasional opiate users, critical questions about use versus abuse among such users remain unanswered.

First, virtually nothing is known about the way in which occasional opiate consumption is patterned over time. Prior to 1970 most surveys did not go "beyond the simple fact of use to obtain data pertaining to patterns of use" (Rittenhouse 1977). In more recent work, data on the frequency of use are often reported in the form of the number of occasions of use within the previous month, year, or lifetime of the user but without indicating how such occasions are distributed. For example, an occasional user classified as having used fewer than ninety-nine times in one year may, hypothetically, have used three times daily for about one month, or he may have used once every three or four days throughout the year. Although these two hypothetical examples indicate two very different patterns of opiate use or levels of control, such differences are not revealed by the data.

Second, little is known about how stable the various patterns of occasional use are. There is some evidence (Powell 1973) that users may continue their nonaddictive use for years, but data are not yet available as to the proportions of occasional users who maintain this style for various periods of time. If occasional use is not a stable pattern-that is, if the majority of such users either quickly give up opiate use or eventually become addicted-it would seem that the apparent pattern of occasional use has little significance for policy formulation.

Third, none of the studies just reviewed consider a subject's level of nonopiate drug use to be relevant to his or her classification as an occasional opiate user. Until recently the existence of occasional opiate users who might also be alcoholics or barbiturate addicts could hardly have been seen as having relevance for policy regarding opiate use.

Fourth, most of the available data do not clearly relate frequency of use to quality of use. (Robins's work is a notable exception.) Although it is tempting to assume that less frequent use is less abusive, this is not always the case. Except at the extremes (use once a year versus multiple daily use), frequency is not a reliable predictor' of drug-related difficulties (Zinberg, Harding & Apsler 1978). Problem drinking, for example, occurs across a broad range of frequencies of use (Smart 1974; Pattison 1979). And as Jerome H. Jaffe has pointed out (1975), although the risk of adverse effects tends to increase with the level of drug involvement, there are important exceptions to this rule, such as the increased risk of overdose or of psychological discomfort among neophyte users. Because researchers have not consistently studied opiate related problems in conjunction with the quality of opiate use (Smart 1974), no light has been shed on the numbers and characteristics of opiate users as opposed to the numbers and characteristics of opiate abusers.

Fifth, little is known about the demographic characteristics, social background, personality structure, and other pertinent factors that tend to be associated with different patterns of opiate use.