- Email: [email protected]
Recreational drug use concurrent with abuse or dependence on other psychoactive substances
Journalof SubstanceAbuseTreatment,Vol.3, No. 6, pp. 499-504, 1996 Copyright© 1996Elsevier ScienceInc. Pnnted m the USA.Allrightsreserved 0740-5472/96 $15.00 + .00
PII S0740-5472(96)00190-0
ELSEVIER
PRELIMINAR Y INVESTIGATION
Recreational Drug Use Concurrent with Abuse or Dependence on Other Psychoactive Substances KERRY P. HEFFNER, PhD, H A R O L D
R O S E N B E R G , PhD, N A N ROTHROCK, BA,
K E L L Y K I M B E R - R I G G S , BA, AND CHRISTINE G O U L D , BA Bowling Green State University
Abstract-In an effort to identify instances of the non-problematic use of a drug concurrent with the problematic use of one or more other drugs, we used structured interviews to obtain comprehensive drug use histories from 48 clients admitted to an intensive outpatient program. We classified clients on the basis of whether they demonstrated evidence of concurrent problematic and non-problematic drug use (Index and Probable Index cases) or only problematic drug use patterns (Non-Index cases). Both Index and Non-Index drug use patterns were about equally common in our sample. Both Index and Probable Index cases used a variety of drugs in a non-problematic manner and were generally congruent in their self-labelling of their drug use relative to their DSM-IV status for each drug used. We discuss several limitations of the stud)', including our reliance on retrospective, self-report data; potential problems with generalization to other populations; and possible changes in drug use patterns over time. Copyright © 1996 Elsevier Science Inc.
Keywords-substance use patterns; recreational drug use; polydrug abuse.
INTRODUCTION
ment agencies, research attention to this phenomenon has increased dramatically in the past decade. For example, recent investigations have examined personality differences between individuals who abuse a single drug and those who abuse multiple drugs (O'Boyle, 1993; Schinka, Curtiss, & Mulloy, 1994) and several studies have looked at whether treatment outcome varies as a function o f single versus polydrug use (Brower, Blow, Hill, & Mudd, 1994; Brown, Seraganian, & Tremblay, 1993). Although the prevalence of polydrug abuse and dependence has increased in recent years, there are still many persons who abuse or become dependent on only one substance. Some proportion of single-drng abusing persons may never sample other drugs or may discontinue their use of other drugs as they come to prefer one substance over others. An alternate possibility is that some people who abuse or are dependent on one (or more) substances are able to use other drugs in a non-abusive and non-dependent manner.
THE PHENOMENON OF polydrug abuse (i.e., concurrent abuse or dependence on more than one type of psychoactive substance) has received increasing attention by researchers and clinicians. Data from the N I M H Epidemiologic Catchment Area (ECA) survey (Regier et al., 1990) indicated that among individuals with a lifetime prevalence for an alcohol use disorder, 22% were comorbid for other drug use disorders. The abuse of multiple drugs by a single user is not new but, as a result of both more comprehensive assessment and an apparently genuine increase in polydrug abuse by clients presenting to treat-
Kerry P. Heffner is now at the Department of Psychology. University of Indianapolis, IN. Reprint requests should be addressed to Kerry P. Heffner, Department of Psychology,University of Indianapolis, 1400 East Hanna Avenue. Indianapolis. IN 46227-3697.
Received April 2, 1996; Accepted November 15, 1996. 499
500
K.P. Heffner et al.
This latter implication is contradictory to the clinical lore that it is impossible for a drug-addicted person to moderate or control his or her use of any psychoactive substance. Typically expressed in the vernacular as, "A drug is a drug is a drug," this view is often based on the notion that the drug-addicted individual has a brain disease in which psychological factors (e.g., goal choice, self-efficacy) and social-environmental factors (e.g., peer group pressures, interpersonal consequences) are ancillary to the etiology and maintenance of polydrug addiction. We hypothesize that polydrug abuse and dependence are multi-determined and that addiction to any one drug does not automatically preclude the moderate or controlled use of all other psychoactive substances. This is not to say that all persons who abuse substances will be able to moderate their use of all substances, but rather that evidence for abuse or dependence with one drug does not mean that the individual will automatically be addicted to all substances he or she uses. As an initial, test of this hypothesis, we interviewed substance-abusing clients who were admitted to an intensive outpatient therapy program in an urban setting. The purpose of the interview was to take a comprehensive history of clients' use, abuse, and dependence on illicit and non-illicit drugs, including over-the-counter medications. We had no doubt that we would find clients for whom all substance use was problematic. The more interesting question for us was whether we could identify cases in which a client who was abusing or dependent on one or more drugs was also consuming other drugs in a nonabusive and non-dependent manner.
METHOD Participants The population from which we sampled comprised 80 clients admitted between October, 1994 and April, 1995 to an intensive outpatient substance abuse treatment program affiliated with an urban general medical hospital. Clients were recruited by their counselors to participate in the study during the second or third week of the 6-week program. Clients who participated were not compensated for their participation. Forty-eight clients (39 men; 9 women) volunteered to be interviewed. The mean age of respondents was 36.4 years (SD = 8.1 years; range = 19-60). The majority of clients identified themselves racially as White or Caucasian (81%); 8% classified themselves as Black or African American; the remaining 10% reported other ethnic backgrounds (mostly Hispanic). Most participants were either currently married (52%) or never married (27%). The mean years of education was 12.8 (SD = 2.4; range = 7-20). A broad range of occupations were reported, including manageriai, sales, professional, and blue-collar positions. Only two respondents reported being unemployed.
Structured Interviews Clients' drug use histories for over-the-counter medications and each of the drug classes included in the current Diagnostic and Statistical Manual o f Mental Disorders (DSM-IV; American Psychiatric Association, 1994) were
obtained using a structured interview designed for this study, j Research interviews were scheduled during the clients' regular treatment hours and conducted on the premises of the clinic. Clients were excused from group therapy for the 30 to 60 minutes required to conduct the interview. Participants were first asked to provide basic sociodemographic information (age, race, years of education, current relationship status, and occupation). They were then asked to identify which of the following classes of drugs they had ever used at any time in their life: alcohol, amphetamine, cannabis, cocaine, hallucinogens, inhalants, nicotine, opioids, PCP, sedative-hypnotics, tranquilizers or antianxiety drags, and over-the-counter medicines. 2 We used this procedure to establish a list of all drugs that a client had used in his or her lifetime before the interviewer initiated detailed inquiry about each drug class ever used. For each class of drugs that respondents reported using at least once in their lifetime, they were asked: (a) whether the drug was prescribed by a physician; (b) their age at their first and last use; (c) the typical route of administration (i.e., oral, smoked, intranasal, intravenous); (d) the approximate number of times they used the substance (i.e., 1-2 times, 3-4 times, 5-10 times, more than 10 times); and (e) the label they applied to characterize their use of a substance (i.e., medical, experimental, occasional, social/moderate, problem, addicted/dependent). Response options for several of these questions were provided to respondents on a piece of paper. Respondents were next asked a series of questions to assess whether their use of a particular class of substances met any of the four DSM-IV criteria for Substance Abuse or seven D S M - I V criteria for Substance Dependence. Responses to the structured interview questions were used by the research team to classify respondents' drug use for each drug class as either Abuse, Dependence, Nonproblem, Experimental, or Medical. One client did not meet the criteria for either Abuse or Dependence to any of the substances that he or she reported using. Information obtained from this client was not used in any subsequent analyses. Interviewers. Three female undergraduate psychology
majors were trained to conduct the structured interviews. Training of interviewers included instruction on D S M - I V criteria for Substance-Related Disorders, basic interviewing skills, and how to respond to potentially difficult
~Acopy of the structured interview is availablefrom the first author. -'Although sedative-hypnotics and tranquilizers are included in the same drug class in the DSM-IV, the two were separatedfor this study.
Concurrent Recreational and Abusive Drug Use
interview situations (e.g., a client making sexual advances, reports of continuing drug use, expression of suicidal or homicidal intent). This was followed by several practice interviews with the trainers who provided supervisory feedback.
Procedure. After a participant signed the consent form, one of the interviewers conducted the structured interview. Interviews were initiated with a review of the purpose of the study and the participants' rights. Respondents were informed that any drug-related information they reported during the interview would not be disclosed to treatment staff in a manner in which they could be individually identified. Additionally, participants were informed about the limits to confidentiality (i.e., reports of suicidal intention, child abuse, and threats of harm to others).
RESULTS Co-Occurrence of Abuse and/or Dependence with Nonproblem, Experimental, and/or Medical Use Based on DSM-IV criteria, we classified each respondent's use of each drug class they reported using as either Abuse, Dependence, or Other use. There were three further classification categories for Other use: (a) Nonproblem: use of a particular class of drugs five or more times without meeting the criteria for either Abuse or Dependence; (b) Experimental: use of a drug class less than five times along with a self-label of experimental or occasional use; and (c) Medical: use of a drug class that was prescribed by a physician and did not meet the criteria for either Abuse or Dependence. After classifying clients' drug use histories for each class of drugs that they reported using, we classified clients on the basis of whether or not they experienced Abuse and/or Dependence for one or more drug classes concurrent with Nonproblem, Experimental, or Medical use of other substances. Clients who met the criteria for Abuse or Dependence to one or more substances with no concurrent Nonproblem, Experimental, and/or Medical use of other substances were classified as Non-Index cases. Those who met the criteria for Abuse and/or Dependence to one or more substances, but who also demonstrated evidence of Nonproblem, Experimental, and/or Medical use of one or more other substances were classified as Index cases. Respondents who met the criteria for Abuse and/or Dependence to one or more substances and who had stopped their non-abusive or non-dependent use of a drug in the year prior to their current age were classified as Probable Index cases. This third group represents clients who demonstrated the likely co-occurrence of nonabusive or non-dependent drug use of one (or more) substances concurrently with the abusive or dependent use of other substances in the year prior to initiation of therapy. The major question in this study was whether we could identify cases in which an individual was abusing
501
or dependent on one or more drugs and also concurrently consuming other substances in a non-abusive and nondependent manner. That is, we were interested in the prevalence of Index and Probable Index cases. A total of 22 Index cases (18 male, 4 female), 5 Probable Index cases (4 male, 1 female), and 20 Non-Index cases (16 male, 4 female) were identified, accounting for 47%, 10%, and 43% of the sample, respectively. Thus, over half of the sample was classified as either Index or Probable Index cases.
Classes of Drugs Used in Nonproblem, Experimental, or Medical Ways. We next examined what classes of drugs Index and Probable Index cases reported using in nonabusive and non-dependent ways at the same time that they abused or were dependent on at least one other class of drugs. The substances used by the 27 Index and Probable Index cases in Nonproblem, Experimental, or Medical ways were nicotine (n -- 12), cannabis (n = 7), tranquilizers/anxiolytics (n = 7), opiods (n -- 4), sedativehypnotics (n = 3), hallucinogens (n = 1), inhalants (n = 1), PCP (n = 1), OTC diet pills (n = 1 ), and OTC cough medicine (n = 1). 3 No Index or Probable Index case reported the use of alcohol in Nonproblem, Experimental or Medical ways. Self-labels Used to Characterize Nonproblem, Experimental, or Medical Use. We were also interested in how these clients labeled their drug use and the correspondence between self-labels and DSM-IV categorization. The 12 nicotine users who did not meet the criteria for Abuse or Dependence labeled their use as addicted/ dependent (n = 6), problem (n = 1), social/moderate (n = 2), occasional (n = 2), and experimental (n = 1). The 7 users of cannabis who did not meet the criteria for Abuse or Dependence, labeled their use as social/moderate (n = 6) and occasional (n = 1). The 7 users of anxiolytic drugs who did not meet the criteria for Abuse or Dependence selected the social/moderate (n = 2), occasional (n = 2), and medical (n = 3) labels. The 4 opioid users who did not meet the criteria for Abuse or Dependence selected the problem (n = 1), occasional (n = 1), experimental (n = 1), and medical (n = 1) labels. The 3 users of sedative-hypnotics who did not meet the criteria for Abuse or Dependence selected the problem (n = 1), social/moderate (n = 1), and occasional (n = 1) labels. The single respondent using inhalants in a manner that did not meet the DSM-IV criteria for Abuse or Dependence labeled his use a problem. The single non-abusive and non-dependent PCP user selected the experimental label. Each of the single users of hallucinogens, OTC diet pills, and OTC cough medicine selected the occasional label to
~The reported ns indicate the number of Index and Probable Index cases reporting the non-abusive and non-dependent use of a particular drug class. Because respondents could use more than one class of drug in non-abusive and non-dependent ways, the total exceeds 27.
502
K.P. Heffner et al.
TABLE 1 Number of Index and Probable Index Cases Meeting Either Zero, One, or Two DSM-IV Criteria for Substance Dependence for Drugs Classified as Nonproblem, Experimental, or Medical Use N u m b e r of DSM-IV
Criteria Met Drug C l a s s Cannabis
Nicotine Opiods Sedative-Hypnotics Tranquilizers/Anxiolytics Hallucinogens Inhalants PCP O T C D i e t Pills OTC Cough Medicine
Zero
One
Two
5 4 3 2 4 1 1 1 1 1
-4 1 -2 ------
2 4 -1 1 ----
Note. PCP = phencyclidine, O T C = over-the-counter.
characterize their use. This listing demonstrates that, with the exception of the nicotine users (who often rated their use more severely than their DSM-IV status), the majority of Index and Probable Index cases labeled their nonabusive and non-dependent use of substances as either occasional or social/moderate. Thus, self-selected labels closely matched respondents' DSM-IV categorization.
The Relationship of Medical to Nonproblem and ExperiMental Use. Given that we included among the Index and Probable Index cases those respondents who were taking physician-prescribed medications, we assessed the extent to which these cases might have "artificially" inflated the number of Index and Probable Index cases we found. In only 4 of the 27 Index or Probable Index cases was the non-abusive and non-dependent use of a particular drug class classified as Medical. Furthermore, because respondents could use more than one class of drug in nonabusive and non-dependent ways, 3 of these 4 cases also reported the non-problem use of additional drug classes. Thus, only 1 Index case was found in which the non-abusive or non-dependent use of another substance was solely Medical.
The Cusp of Dependence. We were also interested in the extent to which Nonproblem, Experimental, or Medical use of a particular substance was on the cusp of becoming dependent use. Specifically, we were interested in how many of the clients whose drug use histories did not meet the full DSM-IV Dependence requirement of three symptoms but did display either zero, one, or two of the diagnostic criteria. Table 1 lists the number of Index and Probable Index cases meeting either zero, one, or two criteria for each of the substances classified as Nonproblem, Experimental, or Medical use.
As Table 1 shows, the majority of respondents who used at least one class of drugs in a non-abusive and nondependent manner did not meet any of the DSM-IV criteria for Substance Dependence. In 7 of the 8 instances where two DSM-IV criteria for Dependence were met, and 3 of the 7 instances in which only one DSM-IV indicator was present, symptoms associated with physiological dependence on a substance (i.e., tolerance or withdrawal) were reported. Further, 8 of the 15 respondents who met one or two of the Dependence criteria labeled their own use of a substance as either a problem or addicted/dependent, 4 selected the social/moderate label, and 3 reported physician-prescribed drug use. These findings suggest that some clients whose drug use did not meet the full DSM-IV criteria for Dependence were having (and recognized they were having) problems. If we reclassified those respondents whose use of a particular class of drugs did not meet the criteria for Dependence, but whose use was associated with signs of tolerance or withdrawal, four current Index cases and one current Probable Index case would be classified as Non-Index cases. This more conservative classification would result in a total of 18 Index cases, 4 Probable Index cases, and 25 Non-Index cases. Thus, reclassifying non-DSM-Dependent drug use associated with any report of physiological dependence still produces a meaningful proportion of Index and probable Index cases (n = 22; 47% of the total sample).
Patterns of Co-Occurrence of Substance Abuse and/or Dependence to One or More Drugs Our data allowed us to examine the question of the prevalence of single-drug versus polydrug abuse or dependence. Table 2 summarizes the findings regarding the prevalence of single versus polydrug abuse or dependence. As Table 2 shows, 10 of the 47 respondents met the criteria for Abuse or Dependence to only one drug. In all 10 cases the drug was alcohol, and all 10 cases met the criteria for Dependence. Six of the 10 cases were also classified as Index cases and one as a Probable Index case. In six of these seven Index or Probable Index cases among the single-drug dependent users, nicotine was the drug used in a non-abusive or non-dependent manner; the remaining respondent used physician-prescribed opioids non-abusively. The remaining 37 cases met the criteria for Abuse and/ or Dependence to two or more substances. Although there were idiosyncratic polydrug use patterns within the sample, there are a few drugs that regularly co-occur throughout the sample. In addition to alcohol, these substances include nicotine, cannabis, and cocaine (see Table 2). Past research on polydrug use has tended to focus on the use of alcohol and other illicit drugs. Nicotine has been largely overlooked. In our sample 79% (n = 37) of the respondents were using nicotine at the time they entered treatment. Of these 37, the majority met criteria for
Concurrent Recreational and Abusive Drug Use
503
TABLE 2 Patterns of Co-Occurrence of Abuse and/or Dependence
Total Cases Abuse/Dependence to Only One Drug Alcohol Abuse/Dependence to Two Drugs Alcohol-nicotine Alcohol-cannabis Alcohol-cocaine AlcohoI-OTC cough medicne Abuse/Dependence to Three Drugs Alcohol-cannabis-nicotine Alcohol-cocaine-nicotine Alcohol-cocaine-cannibis Alcohol-cannabis-hallucinogens Alcohol-opioids-OTC cough medicne Alcohol-amphetamine-tranquilizers Abuse/Dependence to four drugs Alcohol-cocaine-cannabis-nicotine Alcohol-cocaine-cannabis-OTC cough medicine Alcohol-cocaine-opioids-nicotine Alcohol-cociane-nicotine-OTC sleep aids Abuse/Dependence to Five or More Drugs Note.
10 10 14 8 3 2 1 13 4 5 1 1 1 1 6 3 1 1 1 4
OTC = over-the-counter.
Dependence (n = 23; 62%) or Abuse (n = 2; 5%). As noted in Table l, 8 of the 12 remaining respondents who used nicotine met either one or two of the D S M - I V criteria for Dependence, with most of these reporting either tolerance or withdrawal. DISCUSSION One of the most interesting questions in the treatment of persons who abuse substances is whether all drug-addicted individuals are destined to become addicted to any and all drugs they consume, or whether some addicts are able to use drugs other than the ones to which they are addicted in a controlled, moderate, or harm-free manner. This question has implications for both the treatment and conceptualization of addictive behavior. For example, when conducting therapy, will we be accurate if we inform clients that they are constitutionally incapable of recreational use of alcohol or marijuana if they have been addicted to cocaine or opiates? With regard to theory, are addicts' drug use patterns consistent with a model of addiction as an intrapersonal disease process which will manifest itself regardless of the drug consumed; or, are drug use patterns more consistent with a biopsychosocial model that postulates that addiction to any particular substance is a behavior influenced by three interacting factors: (a) a drug's pharmacologic effects, (b) an individual user's psychology, and (c) the environmental context in which the behavior occurs? In an effort to shed light on these questions, we used structured interviews to obtain comprehensive drug use
histories with a subset of client volunteers admitted to an intensive outpatient therapy program. We first classified clients' use of each drug as abusive, dependent, nonproblematic (i.e., recreational), experimental, or medical, and then classified clients as Non-Index cases if their drug use at the time they entered treatment demonstrated evidence of abuse or dependence on all drugs they reported using. We classified clients as Index or Probable Index cases if their histories demonstrated evidence of abuse or dependence on one or more drugs concurrent with recreational, experimental, or medical use of other drugs at the time of entry into treatment or within the past year. We found that the majority of our sample were abusing multiple drugs (37 of 47 cases) and that both Index and Non-Index drug use patterns were about equally common in our sample. Further analysis of the Index and Probable Index cases found that: (a) these clients used a wide variety of drugs in a non-abusive and non-dependent manner, with nicotine and cannabis being the most frequently reported; (b) these clients' self-labelling and their D S M - I V status for each drug they used were generally congruent; (c) when there was a discrepancy between self-labels and D S M - I V classification, it was usually in the direction of clients selecting the "problem" or "addicted" label to describe their use even though they did not meet three required D S M - I V criteria for Dependence; (d) few of the drugs used in non-problem ways by Index and Probable Index cases were medical prescriptions; and (e) in a few cases, participants' use of a drug did not meet enough D S M - I V criteria to be classified as dependence, although they did report signs of tolerance, withdrawal, or other consequences indicative of a problem. Although we found numerous cases demonstrating the phenomenon of addiction concurrent with recreational and experimental use of other drugs, there are some methodological limitations of our study that may temper the reliability and validity of our results. For one, we used a retrospective method and do not know to what degree respondents may have either over- or under-reported their drug use. However, it seems somewhat unlikely that respondents would disclose a wide range of drug use and report numerous negative consequences associated with their use of one class of drugs but not another. Further, past research (Skinner, 1984) has established that alcohol dependent persons' self-reports are relatively valid when the individual is drug-free at the time of the assessment; when structured interviews are used; when the respondent has no clear motives to dissimulate or distort reports of alcohol use; when confidentiality of information is assured to the individual; and when the assessment is completed in a clinical or research setting. A second limitation of our study is that we do not know what changes in drug use, if any, will occur in the future. It is possible that the current recreational use of a drug (or several drugs) by an Index case will become abusive or dependent in time, with that case being classified as a Non-Index case at some later point. For example,
504
a cocaine-addicted individual might currently use tranquilizers in a non-abusive manner but later develop a physiological dependence on them or experience negative consequences associated with his or her use. Similarly, it is possible that a current Non-Index case might begin and continue to use another drug in a recreational manner, and be classified as an Index case at some later point. For example, a heroin-addicted person might come to drink alcohol in a non-abusive and non-dependent manner. It is also possible that the Index and Probable Index cases simply misattributed negative consequences actually associated with one class of drugs to the primary drug that they were abusing or dependent on. That is, respondents' main drug of abuse may be the most salient focus of their awareness and may be part of the motivating influence for entering treatment. As a result, they may associate the negative effects of any drug use to their drug of choice, overlooking the actual connection with other drugs used. However, it is apparent from examination of Table 1 that the Index and Probable Index cases did attribute at least some negative consequences to the drugs they used in a recreational or nonproblematic manner. Finally, our results may not generalize across the broader population of drug-addicted clients. Our sample was limited to clients admitted to an intensive outpatient treatment program who volunteered to participate. Therefore, the present study is best seen as but one investigation that sheds some light on these questions. Research-
K.P. Heffner et al.
ers will need to conduct prospective follow-up studies of a broad cross-section of drug users--both those in residential, outpatient, and informal treatments and those who never participate in any form of therapy or self-help g r o u p - - t o assess more fully the prevalence and stability of both polydrug use per se and recreational drug use in the context of addiction.
REFERENCES American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.), Washington, DC' Author. Brower, K. J., Blow, F. C.. Hill, E. M., & Mudd, S. A. (1994). Treatment outcome of alcoholics with and without cocaine disorders. Alcoholism: Clinical and Experimental Research, 18, 734-739. Brown, T. G., Seraganian, R., & Tremblay, J. (1993). Alcoholics also dependent on cocaine in treatment: Do they differ from "'pure" alcoholics? Addictive Behaviors, 19, 105-112. O'Boyle, M. (1993). Personality disorder and multiple substance dependence. Journal of Personality Disorders, 7, 342-347. Regier, D. A., Farmer, M. E., Rae, D. S., Locke, B. Z.. Keith, S. J., Judd, L. L., & Goodwin, F. K. (1990). Comorbidity of mental disorders with alcohol and other drug abuse: Results from the Epidemiologic Catchment Area (ECA) study. Journal of the American Medical Association, 264, 2511-2518. Schinka. J. A., Curtiss, G., & Mulloy, J. M. (1994). Personality variables and self-medication in substance abuse. Journal ofPersonali~' Assessment, 63, 413-422. Skinner, H. A. (1984). Assessing alcohol use by patients in treatment. In R. G. Smart, H. D. Cappell, F. B. Glaser, Y. Isreal, H. Kalant, W. Schmidt, & E. Sellers (Eds.), Research advances in alcohol and drug problems (pp. 183-207). New York: Plenum Press.
PII S0740-5472(96)00190-0
ELSEVIER
PRELIMINAR Y INVESTIGATION
Recreational Drug Use Concurrent with Abuse or Dependence on Other Psychoactive Substances KERRY P. HEFFNER, PhD, H A R O L D
R O S E N B E R G , PhD, N A N ROTHROCK, BA,
K E L L Y K I M B E R - R I G G S , BA, AND CHRISTINE G O U L D , BA Bowling Green State University
Abstract-In an effort to identify instances of the non-problematic use of a drug concurrent with the problematic use of one or more other drugs, we used structured interviews to obtain comprehensive drug use histories from 48 clients admitted to an intensive outpatient program. We classified clients on the basis of whether they demonstrated evidence of concurrent problematic and non-problematic drug use (Index and Probable Index cases) or only problematic drug use patterns (Non-Index cases). Both Index and Non-Index drug use patterns were about equally common in our sample. Both Index and Probable Index cases used a variety of drugs in a non-problematic manner and were generally congruent in their self-labelling of their drug use relative to their DSM-IV status for each drug used. We discuss several limitations of the stud)', including our reliance on retrospective, self-report data; potential problems with generalization to other populations; and possible changes in drug use patterns over time. Copyright © 1996 Elsevier Science Inc.
Keywords-substance use patterns; recreational drug use; polydrug abuse.
INTRODUCTION
ment agencies, research attention to this phenomenon has increased dramatically in the past decade. For example, recent investigations have examined personality differences between individuals who abuse a single drug and those who abuse multiple drugs (O'Boyle, 1993; Schinka, Curtiss, & Mulloy, 1994) and several studies have looked at whether treatment outcome varies as a function o f single versus polydrug use (Brower, Blow, Hill, & Mudd, 1994; Brown, Seraganian, & Tremblay, 1993). Although the prevalence of polydrug abuse and dependence has increased in recent years, there are still many persons who abuse or become dependent on only one substance. Some proportion of single-drng abusing persons may never sample other drugs or may discontinue their use of other drugs as they come to prefer one substance over others. An alternate possibility is that some people who abuse or are dependent on one (or more) substances are able to use other drugs in a non-abusive and non-dependent manner.
THE PHENOMENON OF polydrug abuse (i.e., concurrent abuse or dependence on more than one type of psychoactive substance) has received increasing attention by researchers and clinicians. Data from the N I M H Epidemiologic Catchment Area (ECA) survey (Regier et al., 1990) indicated that among individuals with a lifetime prevalence for an alcohol use disorder, 22% were comorbid for other drug use disorders. The abuse of multiple drugs by a single user is not new but, as a result of both more comprehensive assessment and an apparently genuine increase in polydrug abuse by clients presenting to treat-
Kerry P. Heffner is now at the Department of Psychology. University of Indianapolis, IN. Reprint requests should be addressed to Kerry P. Heffner, Department of Psychology,University of Indianapolis, 1400 East Hanna Avenue. Indianapolis. IN 46227-3697.
Received April 2, 1996; Accepted November 15, 1996. 499
500
K.P. Heffner et al.
This latter implication is contradictory to the clinical lore that it is impossible for a drug-addicted person to moderate or control his or her use of any psychoactive substance. Typically expressed in the vernacular as, "A drug is a drug is a drug," this view is often based on the notion that the drug-addicted individual has a brain disease in which psychological factors (e.g., goal choice, self-efficacy) and social-environmental factors (e.g., peer group pressures, interpersonal consequences) are ancillary to the etiology and maintenance of polydrug addiction. We hypothesize that polydrug abuse and dependence are multi-determined and that addiction to any one drug does not automatically preclude the moderate or controlled use of all other psychoactive substances. This is not to say that all persons who abuse substances will be able to moderate their use of all substances, but rather that evidence for abuse or dependence with one drug does not mean that the individual will automatically be addicted to all substances he or she uses. As an initial, test of this hypothesis, we interviewed substance-abusing clients who were admitted to an intensive outpatient therapy program in an urban setting. The purpose of the interview was to take a comprehensive history of clients' use, abuse, and dependence on illicit and non-illicit drugs, including over-the-counter medications. We had no doubt that we would find clients for whom all substance use was problematic. The more interesting question for us was whether we could identify cases in which a client who was abusing or dependent on one or more drugs was also consuming other drugs in a nonabusive and non-dependent manner.
METHOD Participants The population from which we sampled comprised 80 clients admitted between October, 1994 and April, 1995 to an intensive outpatient substance abuse treatment program affiliated with an urban general medical hospital. Clients were recruited by their counselors to participate in the study during the second or third week of the 6-week program. Clients who participated were not compensated for their participation. Forty-eight clients (39 men; 9 women) volunteered to be interviewed. The mean age of respondents was 36.4 years (SD = 8.1 years; range = 19-60). The majority of clients identified themselves racially as White or Caucasian (81%); 8% classified themselves as Black or African American; the remaining 10% reported other ethnic backgrounds (mostly Hispanic). Most participants were either currently married (52%) or never married (27%). The mean years of education was 12.8 (SD = 2.4; range = 7-20). A broad range of occupations were reported, including manageriai, sales, professional, and blue-collar positions. Only two respondents reported being unemployed.
Structured Interviews Clients' drug use histories for over-the-counter medications and each of the drug classes included in the current Diagnostic and Statistical Manual o f Mental Disorders (DSM-IV; American Psychiatric Association, 1994) were
obtained using a structured interview designed for this study, j Research interviews were scheduled during the clients' regular treatment hours and conducted on the premises of the clinic. Clients were excused from group therapy for the 30 to 60 minutes required to conduct the interview. Participants were first asked to provide basic sociodemographic information (age, race, years of education, current relationship status, and occupation). They were then asked to identify which of the following classes of drugs they had ever used at any time in their life: alcohol, amphetamine, cannabis, cocaine, hallucinogens, inhalants, nicotine, opioids, PCP, sedative-hypnotics, tranquilizers or antianxiety drags, and over-the-counter medicines. 2 We used this procedure to establish a list of all drugs that a client had used in his or her lifetime before the interviewer initiated detailed inquiry about each drug class ever used. For each class of drugs that respondents reported using at least once in their lifetime, they were asked: (a) whether the drug was prescribed by a physician; (b) their age at their first and last use; (c) the typical route of administration (i.e., oral, smoked, intranasal, intravenous); (d) the approximate number of times they used the substance (i.e., 1-2 times, 3-4 times, 5-10 times, more than 10 times); and (e) the label they applied to characterize their use of a substance (i.e., medical, experimental, occasional, social/moderate, problem, addicted/dependent). Response options for several of these questions were provided to respondents on a piece of paper. Respondents were next asked a series of questions to assess whether their use of a particular class of substances met any of the four DSM-IV criteria for Substance Abuse or seven D S M - I V criteria for Substance Dependence. Responses to the structured interview questions were used by the research team to classify respondents' drug use for each drug class as either Abuse, Dependence, Nonproblem, Experimental, or Medical. One client did not meet the criteria for either Abuse or Dependence to any of the substances that he or she reported using. Information obtained from this client was not used in any subsequent analyses. Interviewers. Three female undergraduate psychology
majors were trained to conduct the structured interviews. Training of interviewers included instruction on D S M - I V criteria for Substance-Related Disorders, basic interviewing skills, and how to respond to potentially difficult
~Acopy of the structured interview is availablefrom the first author. -'Although sedative-hypnotics and tranquilizers are included in the same drug class in the DSM-IV, the two were separatedfor this study.
Concurrent Recreational and Abusive Drug Use
interview situations (e.g., a client making sexual advances, reports of continuing drug use, expression of suicidal or homicidal intent). This was followed by several practice interviews with the trainers who provided supervisory feedback.
Procedure. After a participant signed the consent form, one of the interviewers conducted the structured interview. Interviews were initiated with a review of the purpose of the study and the participants' rights. Respondents were informed that any drug-related information they reported during the interview would not be disclosed to treatment staff in a manner in which they could be individually identified. Additionally, participants were informed about the limits to confidentiality (i.e., reports of suicidal intention, child abuse, and threats of harm to others).
RESULTS Co-Occurrence of Abuse and/or Dependence with Nonproblem, Experimental, and/or Medical Use Based on DSM-IV criteria, we classified each respondent's use of each drug class they reported using as either Abuse, Dependence, or Other use. There were three further classification categories for Other use: (a) Nonproblem: use of a particular class of drugs five or more times without meeting the criteria for either Abuse or Dependence; (b) Experimental: use of a drug class less than five times along with a self-label of experimental or occasional use; and (c) Medical: use of a drug class that was prescribed by a physician and did not meet the criteria for either Abuse or Dependence. After classifying clients' drug use histories for each class of drugs that they reported using, we classified clients on the basis of whether or not they experienced Abuse and/or Dependence for one or more drug classes concurrent with Nonproblem, Experimental, or Medical use of other substances. Clients who met the criteria for Abuse or Dependence to one or more substances with no concurrent Nonproblem, Experimental, and/or Medical use of other substances were classified as Non-Index cases. Those who met the criteria for Abuse and/or Dependence to one or more substances, but who also demonstrated evidence of Nonproblem, Experimental, and/or Medical use of one or more other substances were classified as Index cases. Respondents who met the criteria for Abuse and/or Dependence to one or more substances and who had stopped their non-abusive or non-dependent use of a drug in the year prior to their current age were classified as Probable Index cases. This third group represents clients who demonstrated the likely co-occurrence of nonabusive or non-dependent drug use of one (or more) substances concurrently with the abusive or dependent use of other substances in the year prior to initiation of therapy. The major question in this study was whether we could identify cases in which an individual was abusing
501
or dependent on one or more drugs and also concurrently consuming other substances in a non-abusive and nondependent manner. That is, we were interested in the prevalence of Index and Probable Index cases. A total of 22 Index cases (18 male, 4 female), 5 Probable Index cases (4 male, 1 female), and 20 Non-Index cases (16 male, 4 female) were identified, accounting for 47%, 10%, and 43% of the sample, respectively. Thus, over half of the sample was classified as either Index or Probable Index cases.
Classes of Drugs Used in Nonproblem, Experimental, or Medical Ways. We next examined what classes of drugs Index and Probable Index cases reported using in nonabusive and non-dependent ways at the same time that they abused or were dependent on at least one other class of drugs. The substances used by the 27 Index and Probable Index cases in Nonproblem, Experimental, or Medical ways were nicotine (n -- 12), cannabis (n = 7), tranquilizers/anxiolytics (n = 7), opiods (n -- 4), sedativehypnotics (n = 3), hallucinogens (n = 1), inhalants (n = 1), PCP (n = 1), OTC diet pills (n = 1 ), and OTC cough medicine (n = 1). 3 No Index or Probable Index case reported the use of alcohol in Nonproblem, Experimental or Medical ways. Self-labels Used to Characterize Nonproblem, Experimental, or Medical Use. We were also interested in how these clients labeled their drug use and the correspondence between self-labels and DSM-IV categorization. The 12 nicotine users who did not meet the criteria for Abuse or Dependence labeled their use as addicted/ dependent (n = 6), problem (n = 1), social/moderate (n = 2), occasional (n = 2), and experimental (n = 1). The 7 users of cannabis who did not meet the criteria for Abuse or Dependence, labeled their use as social/moderate (n = 6) and occasional (n = 1). The 7 users of anxiolytic drugs who did not meet the criteria for Abuse or Dependence selected the social/moderate (n = 2), occasional (n = 2), and medical (n = 3) labels. The 4 opioid users who did not meet the criteria for Abuse or Dependence selected the problem (n = 1), occasional (n = 1), experimental (n = 1), and medical (n = 1) labels. The 3 users of sedative-hypnotics who did not meet the criteria for Abuse or Dependence selected the problem (n = 1), social/moderate (n = 1), and occasional (n = 1) labels. The single respondent using inhalants in a manner that did not meet the DSM-IV criteria for Abuse or Dependence labeled his use a problem. The single non-abusive and non-dependent PCP user selected the experimental label. Each of the single users of hallucinogens, OTC diet pills, and OTC cough medicine selected the occasional label to
~The reported ns indicate the number of Index and Probable Index cases reporting the non-abusive and non-dependent use of a particular drug class. Because respondents could use more than one class of drug in non-abusive and non-dependent ways, the total exceeds 27.
502
K.P. Heffner et al.
TABLE 1 Number of Index and Probable Index Cases Meeting Either Zero, One, or Two DSM-IV Criteria for Substance Dependence for Drugs Classified as Nonproblem, Experimental, or Medical Use N u m b e r of DSM-IV
Criteria Met Drug C l a s s Cannabis
Nicotine Opiods Sedative-Hypnotics Tranquilizers/Anxiolytics Hallucinogens Inhalants PCP O T C D i e t Pills OTC Cough Medicine
Zero
One
Two
5 4 3 2 4 1 1 1 1 1
-4 1 -2 ------
2 4 -1 1 ----
Note. PCP = phencyclidine, O T C = over-the-counter.
characterize their use. This listing demonstrates that, with the exception of the nicotine users (who often rated their use more severely than their DSM-IV status), the majority of Index and Probable Index cases labeled their nonabusive and non-dependent use of substances as either occasional or social/moderate. Thus, self-selected labels closely matched respondents' DSM-IV categorization.
The Relationship of Medical to Nonproblem and ExperiMental Use. Given that we included among the Index and Probable Index cases those respondents who were taking physician-prescribed medications, we assessed the extent to which these cases might have "artificially" inflated the number of Index and Probable Index cases we found. In only 4 of the 27 Index or Probable Index cases was the non-abusive and non-dependent use of a particular drug class classified as Medical. Furthermore, because respondents could use more than one class of drug in nonabusive and non-dependent ways, 3 of these 4 cases also reported the non-problem use of additional drug classes. Thus, only 1 Index case was found in which the non-abusive or non-dependent use of another substance was solely Medical.
The Cusp of Dependence. We were also interested in the extent to which Nonproblem, Experimental, or Medical use of a particular substance was on the cusp of becoming dependent use. Specifically, we were interested in how many of the clients whose drug use histories did not meet the full DSM-IV Dependence requirement of three symptoms but did display either zero, one, or two of the diagnostic criteria. Table 1 lists the number of Index and Probable Index cases meeting either zero, one, or two criteria for each of the substances classified as Nonproblem, Experimental, or Medical use.
As Table 1 shows, the majority of respondents who used at least one class of drugs in a non-abusive and nondependent manner did not meet any of the DSM-IV criteria for Substance Dependence. In 7 of the 8 instances where two DSM-IV criteria for Dependence were met, and 3 of the 7 instances in which only one DSM-IV indicator was present, symptoms associated with physiological dependence on a substance (i.e., tolerance or withdrawal) were reported. Further, 8 of the 15 respondents who met one or two of the Dependence criteria labeled their own use of a substance as either a problem or addicted/dependent, 4 selected the social/moderate label, and 3 reported physician-prescribed drug use. These findings suggest that some clients whose drug use did not meet the full DSM-IV criteria for Dependence were having (and recognized they were having) problems. If we reclassified those respondents whose use of a particular class of drugs did not meet the criteria for Dependence, but whose use was associated with signs of tolerance or withdrawal, four current Index cases and one current Probable Index case would be classified as Non-Index cases. This more conservative classification would result in a total of 18 Index cases, 4 Probable Index cases, and 25 Non-Index cases. Thus, reclassifying non-DSM-Dependent drug use associated with any report of physiological dependence still produces a meaningful proportion of Index and probable Index cases (n = 22; 47% of the total sample).
Patterns of Co-Occurrence of Substance Abuse and/or Dependence to One or More Drugs Our data allowed us to examine the question of the prevalence of single-drug versus polydrug abuse or dependence. Table 2 summarizes the findings regarding the prevalence of single versus polydrug abuse or dependence. As Table 2 shows, 10 of the 47 respondents met the criteria for Abuse or Dependence to only one drug. In all 10 cases the drug was alcohol, and all 10 cases met the criteria for Dependence. Six of the 10 cases were also classified as Index cases and one as a Probable Index case. In six of these seven Index or Probable Index cases among the single-drug dependent users, nicotine was the drug used in a non-abusive or non-dependent manner; the remaining respondent used physician-prescribed opioids non-abusively. The remaining 37 cases met the criteria for Abuse and/ or Dependence to two or more substances. Although there were idiosyncratic polydrug use patterns within the sample, there are a few drugs that regularly co-occur throughout the sample. In addition to alcohol, these substances include nicotine, cannabis, and cocaine (see Table 2). Past research on polydrug use has tended to focus on the use of alcohol and other illicit drugs. Nicotine has been largely overlooked. In our sample 79% (n = 37) of the respondents were using nicotine at the time they entered treatment. Of these 37, the majority met criteria for
Concurrent Recreational and Abusive Drug Use
503
TABLE 2 Patterns of Co-Occurrence of Abuse and/or Dependence
Total Cases Abuse/Dependence to Only One Drug Alcohol Abuse/Dependence to Two Drugs Alcohol-nicotine Alcohol-cannabis Alcohol-cocaine AlcohoI-OTC cough medicne Abuse/Dependence to Three Drugs Alcohol-cannabis-nicotine Alcohol-cocaine-nicotine Alcohol-cocaine-cannibis Alcohol-cannabis-hallucinogens Alcohol-opioids-OTC cough medicne Alcohol-amphetamine-tranquilizers Abuse/Dependence to four drugs Alcohol-cocaine-cannabis-nicotine Alcohol-cocaine-cannabis-OTC cough medicine Alcohol-cocaine-opioids-nicotine Alcohol-cociane-nicotine-OTC sleep aids Abuse/Dependence to Five or More Drugs Note.
10 10 14 8 3 2 1 13 4 5 1 1 1 1 6 3 1 1 1 4
OTC = over-the-counter.
Dependence (n = 23; 62%) or Abuse (n = 2; 5%). As noted in Table l, 8 of the 12 remaining respondents who used nicotine met either one or two of the D S M - I V criteria for Dependence, with most of these reporting either tolerance or withdrawal. DISCUSSION One of the most interesting questions in the treatment of persons who abuse substances is whether all drug-addicted individuals are destined to become addicted to any and all drugs they consume, or whether some addicts are able to use drugs other than the ones to which they are addicted in a controlled, moderate, or harm-free manner. This question has implications for both the treatment and conceptualization of addictive behavior. For example, when conducting therapy, will we be accurate if we inform clients that they are constitutionally incapable of recreational use of alcohol or marijuana if they have been addicted to cocaine or opiates? With regard to theory, are addicts' drug use patterns consistent with a model of addiction as an intrapersonal disease process which will manifest itself regardless of the drug consumed; or, are drug use patterns more consistent with a biopsychosocial model that postulates that addiction to any particular substance is a behavior influenced by three interacting factors: (a) a drug's pharmacologic effects, (b) an individual user's psychology, and (c) the environmental context in which the behavior occurs? In an effort to shed light on these questions, we used structured interviews to obtain comprehensive drug use
histories with a subset of client volunteers admitted to an intensive outpatient therapy program. We first classified clients' use of each drug as abusive, dependent, nonproblematic (i.e., recreational), experimental, or medical, and then classified clients as Non-Index cases if their drug use at the time they entered treatment demonstrated evidence of abuse or dependence on all drugs they reported using. We classified clients as Index or Probable Index cases if their histories demonstrated evidence of abuse or dependence on one or more drugs concurrent with recreational, experimental, or medical use of other drugs at the time of entry into treatment or within the past year. We found that the majority of our sample were abusing multiple drugs (37 of 47 cases) and that both Index and Non-Index drug use patterns were about equally common in our sample. Further analysis of the Index and Probable Index cases found that: (a) these clients used a wide variety of drugs in a non-abusive and non-dependent manner, with nicotine and cannabis being the most frequently reported; (b) these clients' self-labelling and their D S M - I V status for each drug they used were generally congruent; (c) when there was a discrepancy between self-labels and D S M - I V classification, it was usually in the direction of clients selecting the "problem" or "addicted" label to describe their use even though they did not meet three required D S M - I V criteria for Dependence; (d) few of the drugs used in non-problem ways by Index and Probable Index cases were medical prescriptions; and (e) in a few cases, participants' use of a drug did not meet enough D S M - I V criteria to be classified as dependence, although they did report signs of tolerance, withdrawal, or other consequences indicative of a problem. Although we found numerous cases demonstrating the phenomenon of addiction concurrent with recreational and experimental use of other drugs, there are some methodological limitations of our study that may temper the reliability and validity of our results. For one, we used a retrospective method and do not know to what degree respondents may have either over- or under-reported their drug use. However, it seems somewhat unlikely that respondents would disclose a wide range of drug use and report numerous negative consequences associated with their use of one class of drugs but not another. Further, past research (Skinner, 1984) has established that alcohol dependent persons' self-reports are relatively valid when the individual is drug-free at the time of the assessment; when structured interviews are used; when the respondent has no clear motives to dissimulate or distort reports of alcohol use; when confidentiality of information is assured to the individual; and when the assessment is completed in a clinical or research setting. A second limitation of our study is that we do not know what changes in drug use, if any, will occur in the future. It is possible that the current recreational use of a drug (or several drugs) by an Index case will become abusive or dependent in time, with that case being classified as a Non-Index case at some later point. For example,
504
a cocaine-addicted individual might currently use tranquilizers in a non-abusive manner but later develop a physiological dependence on them or experience negative consequences associated with his or her use. Similarly, it is possible that a current Non-Index case might begin and continue to use another drug in a recreational manner, and be classified as an Index case at some later point. For example, a heroin-addicted person might come to drink alcohol in a non-abusive and non-dependent manner. It is also possible that the Index and Probable Index cases simply misattributed negative consequences actually associated with one class of drugs to the primary drug that they were abusing or dependent on. That is, respondents' main drug of abuse may be the most salient focus of their awareness and may be part of the motivating influence for entering treatment. As a result, they may associate the negative effects of any drug use to their drug of choice, overlooking the actual connection with other drugs used. However, it is apparent from examination of Table 1 that the Index and Probable Index cases did attribute at least some negative consequences to the drugs they used in a recreational or nonproblematic manner. Finally, our results may not generalize across the broader population of drug-addicted clients. Our sample was limited to clients admitted to an intensive outpatient treatment program who volunteered to participate. Therefore, the present study is best seen as but one investigation that sheds some light on these questions. Research-
K.P. Heffner et al.
ers will need to conduct prospective follow-up studies of a broad cross-section of drug users--both those in residential, outpatient, and informal treatments and those who never participate in any form of therapy or self-help g r o u p - - t o assess more fully the prevalence and stability of both polydrug use per se and recreational drug use in the context of addiction.
REFERENCES American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.), Washington, DC' Author. Brower, K. J., Blow, F. C.. Hill, E. M., & Mudd, S. A. (1994). Treatment outcome of alcoholics with and without cocaine disorders. Alcoholism: Clinical and Experimental Research, 18, 734-739. Brown, T. G., Seraganian, R., & Tremblay, J. (1993). Alcoholics also dependent on cocaine in treatment: Do they differ from "'pure" alcoholics? Addictive Behaviors, 19, 105-112. O'Boyle, M. (1993). Personality disorder and multiple substance dependence. Journal of Personality Disorders, 7, 342-347. Regier, D. A., Farmer, M. E., Rae, D. S., Locke, B. Z.. Keith, S. J., Judd, L. L., & Goodwin, F. K. (1990). Comorbidity of mental disorders with alcohol and other drug abuse: Results from the Epidemiologic Catchment Area (ECA) study. Journal of the American Medical Association, 264, 2511-2518. Schinka. J. A., Curtiss, G., & Mulloy, J. M. (1994). Personality variables and self-medication in substance abuse. Journal ofPersonali~' Assessment, 63, 413-422. Skinner, H. A. (1984). Assessing alcohol use by patients in treatment. In R. G. Smart, H. D. Cappell, F. B. Glaser, Y. Isreal, H. Kalant, W. Schmidt, & E. Sellers (Eds.), Research advances in alcohol and drug problems (pp. 183-207). New York: Plenum Press.