Nomenclature and classification of drug- and alcohol-related problems: a WHO Memorandum*

Memoranda are state- Les Mimorandums ments concerning the exposent les conclu- M emoranda conclusions or recom- sions et recomman- mendations of certain dations de certaines WHO scientific meet- re'unions scientifiques Aflveimorandums ings; they are signed de l'OMS; ils sontby the participants in signes par les partici- the meeting. pants d ces reunions. Bulletin of the World Health Organization, 59 (2): 225-242 (1981) Nomenclature and classification of drug- and alcohol-related problems: a WHO Memorandum* Earlier work in thisfield is reviewed andpresent concepts and terminologies are exam- ined in detail. A revised way of dealing with ideas implicit in the terms "drug abuse" or "drug misuse" isproposed; the term "neuroadaptive state" issuggested as an alternative to "physical dependence"; aprofile is given ofthe elements that constitutea "drug dependence syndrome"; and the need to differentiate conceptually between "dependence" and "drug related disability" is stressed. A model of dependence is outlined in which dependence is considered as a psycho-physiological-social syndrome determined and kept going by a complexsystem ofreinforcements. The association between substance useandmental illness is discussed with emphasis on possible two-way relationships. The need for population studies and in particularfor longitudinal studies is emphasized. The bearing ofall thefore- going considerations on work towards the revision of relevant classification systems is considered, and, in thefinal section, several recommendations on nomenclature are brought togetherand suggestions are madefor research that might lead to refinement ofclassification and diagnostic systems. Drug and alcohol problems are becoming increas- ingly recognized as a major public health problem and advances in classification and nomenclature will facilitate not only epidemiological research and analysis, but also a greater understanding of disease etiology, and improved management and decision- making in the health care system. However, in this field there are special intrinsic difficulties for classification which are compounded by social and psychological variables that complicate description and application. In this Memorandum an attempt is made to: (a) review the present situation in relation to classi- fication and diagnosis in the whole field of drug- and alcohol-related problems; * This Memorandum was prepared by Professor G. Edwards, Dr A. Arif, and Dr R. Hodgson following a meeting held in Washing- ton, DC, on 25-29 August 1980. Substantial contributions to the thinking embodied in this article were made by all the participants in the meeting and their names are listed on page 242. The three authors wish particularly to acknowledge the contributions made by Dr Drew, Dr Jaffe, Dr Kalant, Dr Schuster, and Dr Smith, who provided material that has been incorporated into the text, Requests for reprints should be addressed to: Director, Division of Mental Health, World Health Organization, 1211 Geneva 27, Switzerland. A French translation will appear in a future edition of the Bulletin. (b) identify gaps in present knowledge, and short- comings in present systems of classification; (c) make recommendations for future work to resolve identified problems with special emphasis on practical proposals for field testing of any ideas put forward. This work on alcohol- and drug-related problems forms part of a much larger international programme on the diagnosis and classification of mental disorders and obviously any recommendations put forward must relate to the deliberations of the other eight specialized groups within the programme. It is also essential to bear in mind that the present study is only part of a continuing and much longer-term endeav- our. The aim is, therefore, more to propose some good ideas for further debate and testing, than to attempt to reach positions of premature finality. EARLIER WORK As long ago as 1952, the WHO Expert Committee on Drugs Liable to Produce Addiction gave a defi- 4050 -225- MEMORANDUM nition of addiction that could be applied to drugs then under international control (1). In 1957, the WHO Expert Committee on Addiction-Producing Drugs introduced a terminology that distinguished between "drug addiction" on the one hand, and "drug habitu- ation" on the other (2). In retrospect, this was a useful attempt to come to terms with the fact that different drugs give rise to different patterns of use, but in practice this terminology led to a good deal of con- fusion. The next step came in 1964, when the same WHO Expert Committee recommended substitution of the term "drug dependence" for both of the previous terms "addiction" and "habituation" (3). Drug dependence was defined as a state arising from repeated administration of a drug on a periodic or continuous basis, and with characteristics that varied according to the agent involved. With this concept, it was also necessary to designate the particular type of drug dependence under consideration, and to build up descriptions for different drug types. This formulation also implied an important distinc- tion between psychic and physical dependence and these conditions were defined by Eddy et al. (4) in the following terms: Psychic dependence. A condition in which a drug produces "... a feeling of satisfaction and a psychic drive that require periodic or continuous adminis- tration of the drug to produce pleasure or to avoid discomfort. " Physical dependence. "... an adaptive state that manifests itself by intensive physical disturbances when the administration of the drug is suspended... These disturbances, i.e., the withdrawal or abstinence syndromes, are made up of specific arrays of symp- toms and signs of psychic and physical nature that are characteristic for each drug type." Another element in what might be termed the "lexi- con", is the concept of "drug abuse" (or drug misuse). Although this term is often used, there is a widespread feeling that problems are set by its imprecision. The first attempt by WHO to define "alcoholism" is to be found in the report of an Expert Committee published in 1951 (5). When the dependence formul- ation was promulgated, alcohol took its place as one of the family of drugs capable of producing a depen- dence picture. But in 1977 a WHO Group of Investi- gators suggested a framework which in some respects took thinking in a new direction. In the report of this Group (6) the idea of a dependence syndrome was developed as follows: "The alcohol dependence syndrome is manifested by alterations at the behavioural, subjective, and psychobio- logical levels with, as a leading symptom, an impaired control over intake of the drug ethyl alcohol. The alcohol dependence syndrome exists in degrees. Its varied manifes- tations are influenced by modifying personal and environ- mental factors so as to give many different presentations... Not all people manifesting alcohol-related disabilities are alcohol dependent,..." This same report (6), as well as developing the idea of a dependence syndrome, emphasized the need to consider the importance of the whole range of physi- cal, psychological, and social disabilities that might result from excessive drinking. The dependence syn- drome is an important disability in its own right, but it should not dominate health concerns. Many people suffer serious alcohol problems who are in no sense dependent. This brief review of previousWHO endeavours sug- gests therefore that progress has been a matter of trial and error. The questions that arise include the need to examine the usefulness of the central idea of depen- dence, the conceptual basis for the distinction between physical and psychic dependence, the meaning to be given to the term drug abuse, the lead that may be given to the general drug field by development of the idea of dependence syndromes, and the relative weighting to be given to dependence and disability (and the relationship between these two categories). FUNDAMENTAL CONCEPTS AND TERMINOLOGY Relationship between research technologies, concepts, and definitions In the area under discussion, a scientific definition cannot be fixed for all time but will be influenced by the state of our knowledge and particularly by advances in research technology. A few examples of technological changes that have affected drug and alcohol studies, selected from among numerous poss- ible examples, will illustrate the impact of such changes on scientific concepts and, in turn, on terminology in this field. For many years it was believed that physical depen- dence on morphine-like drugs required weeks or at least days of regular use. In the 1950s, the use of the opioid antagonist nalorphine permitted researchers at Lexington, Kentucky, to demonstrate a "withdrawal effect" after as little as two days of administration of morphine at therapeutic doses. With the development of the pure antagonist naloxone, it became possible to demonstrate that in normal volunteers a single dose of morphine produces an altered biological state; an injection of naloxone produced a relatively intense dysphoria as well as some other signs and symptoms typical of opioid withdrawal. In short, in the technical sense, physical dependence begins with the first dose of an opioid. Consequently, the emphasis on this phenomenon as a major distin- guishing feature in syndromes characterized by drug- seeking behaviour requires revision. 226 DRUG- AND ALCOHOL-RELATED PROBLEMS We do not have any comparable antagonist for immediately reversing the effects of alcohol or bar- biturate-like drugs, but other technical and con- ceptual advances make it necessary to assume that physical dependence on the latter drugs occurs far sooner than the days or weeks that were previously believed necessary. For example, measurements of sleeping patterns using analyses of all-night electroen- cephalogram recordings suggest that use of short- acting sedative drugs for a few nights may be followed by a "rebound insomnia" when the drug is discon- tinued. Thus, our concepts of physical dependence and its relationship to drug-seeking behaviour must recognize the rapidity with which low levels of the phenomenon develop. Similarly, the distinctions between socially accep- table levels of drug intake (e.g., alcohol and tobacco) and levels that are medically hazardous are highly dependent on the sensitivity of methods for measuring biological damage and risk. For years it was assumed that only heavy and prolonged alcohol use was associ- ated with significant damage to the structure and function of the brain, but with more sensitive psycho- logical tests and computerized axial tomography (CAT) scanners, it is becoming apparent that central nervous system impairment occurs even in those who have no history of malnutrition and that some subtle impairments may develop even in "social drinkers". Many other examples could be given but these few illustrations are sufficient to make the point that new technologies may force radical reappraisal of old assumptions, and in the process call for considerable revision of concepts and terminologies. What is a drug? The exact definition of a "drug" is a very old problem in pharmacology, and any working party on drug-related problems must first consider ways in which this term is used. In 1969, the WHO Expert Committee on Drug Dependence (7) defined a drug as "A substance that, when taken into the living organism, may modify one or more of its functions. " This definition is clearly too broad since both air and water would then be labelled "drugs". Some of the difficulty in defining the term might be overcome by adopting a hierarchical or taxonomic definition. If successive layers of differentiation and specificity were provided, the user would be able to stop at any level that was compatible with the needs of the particular circumstances. For example, "drug" in the broadest sense is any chemical entity or mixture of entities, other than those required for the maintenance of normal health, the administration of which alters biological function and possibly structure. This definition would exclude food, water, oxygen, endocrine substances, etc., in amounts required for the maintenance of normal health. A second level of definition could then specify the uses to which these entities (or "drugs") are put. They may be used for the treatment or alleviation of disease (in which case they are also designated as medications, therapeutic agents, remedies, etc.) or for non-thera- peutic purposes. In this case, while other languages employ different terms, e.g. drogue (French), Rausch- mittel (German), estupefaciente (Spanish), English falls back on such terms as non-medical use of drugs, drug abuse, etc. At a third level of specification, one may differenti- ate chemical entities on the basis of the bodily system or function on which they exert their most marked or obvious effects. "Psychoactive drugs" are those that alter mood, cognition, and behaviour. However, there are no sharp boundaries between "psychoactive drugs" and those not so designated, since substances used primarily for other effects may have psycho- active effects at certain doses or under certain con- ditions. Nevertheless, broad areas of agreement and consensus exist, and underlie the current schemes of pharmacological classification. Psychoactive drugs may be, at the second level, either therapeutic or non- therapeutic agents and some psychoactive substances may be used for both purposes. At a fourth level of definition, our concern is with possible adverse effects on health or social function. Not all psychoactive drugs are equally likely to be self- administered (i.e., to have an equally high reinforce- ment potential in man) or equally likely to induce biological adaptive states in users. For example, tri- cyclic antidepressants, some antihistamines, and some dopamine-receptor blockers are clearly psychoactive, and can cause performance difficulties in people operating complex equipment, yet seldom give rise to self-administration. In the present context, our concern is primarily with those entities that we can define as non-medically- used psychoactive drugs that are likely to be self- administered. For most, but not all, of these, such use can also give rise to other phenomena (such as toler- ance, drug-seeking behaviour under various condi- tions, biological adaptation, response generalization, etc.). There are at least 15 groups of such substances that cause difficulties of sufficient magnitude and with sufficient frequency to warrant their being placed under this heading. Use, abuse, and misuse "Abuse" and "misuse" are unsatisfactory concepts within a scientific approach. Because the terms involve value judgements they are impossible to define in such a way that they are appropriate for 227 MEMORANDUM different drugs in different contexts. For example, in 1969 a WHO Expert Committee (7) defined drug abuse as "... persistent or sporadic excessive drug use inconsistent with or unrelated to acceptable medical practice". Since alcohol is not usually prescribed by the medical profession then any excessive use would be abuse. But then what is excessive? Further attempts to achieve a comprehensive definition of abuse have con- sidered medical need, individual and social harm, and degree of tolerance by the community. Clinicians and research workers can often manage without the terms abuse and misuse, since they are concerned only with the various ways in which a drug is used, and the beneficial and harmful consequences of particular patterns of use. However, where for purposes of any particular agency or administration there has previously been a need for some such terms, a dissection of the ideas behind "abuse" and "misuse" suggests that the following concepts might be clearer and more useful: (a) Unsanctioned use: Use of a drug that is not approved by a society, or a group within that society. When the term is used, it should be made clear who is responsible for the disapproval. The term implies that we accept disapproval as a fact in its own right, with- out having to determine or justify the basis of the disapproval. (b) Hazardous use: Use of a drug that will probably lead to harmful consequences for the user- either to dysfunction or to harm. This concept is similar to the idea of risky behaviour. For instance, smoking twenty cigarettes each day may not be accompanied by any present or actual harm but we know it to be hazardous. (c) Dysfunctional use: Use of a drug that is leading to impaired psychological or social functioning (e.g., loss of job or marital problems). (d) Harmful use: Use of a drug that is known to have caused tissue damage or mental illness in the particular person. Scientific studies are needed to investigate the relationships between patterns of use and the degree of disapproval expressed in particular cultures, as well as the probability of dysfunctional or harmful consequences. Drug tolerance and its relation to drug dependence Tolerance is of interest because it is one of the most prominent changes accompanying chronic adminis- tration of those drugs that can give rise to problems of dependence. However, it does not necessarily follow that an explanation of tolerance will shed much light on the nature of dependence. The relationship be- tween tolerance and dependence is still far from completely understood. Tolerance to drugs (for present purposes, specific- ally those of dependence), is traditionally defined in terms of the dose-response curve, which describes or predicts the intensity of effect produced by a given dose of the drug in question. The term may be further specified as initial tolerance (i.e., the tolerance or sensitivity to the drug in an individual not previously exposed to it), or acquired tolerance (i.e., the increase in tolerance as a result of previous exposure to the drug). When no specification is made, the term " toler- ance" is generally used synonymously with "acquired tolerance". There are many important unanswered questions about tolerance, including the relationship between pharmacological and behavioural or environmental factors in the development of tolerance; the degree of interrelation between tolerance and other different effects of the same drug; the extent and mechanism(s) of cross-tolerance, especially between drugs of differ- ent pharmacological categories; and the role of specific biochemical factors or processes in the development of tolerance, including various neuro- peptides, neurotransmitters, receptors, and cell membranes. The measurement of tolerance is of critical importance to these and other questions, because entirely erroneous conclusions may be drawn on the basis of inappropriate methods. The degree of tolerance acquired, the rate of its development, and its duration after the last drug administration, have been shown to be affected by at least two major types of learning process. The first type occurs when the subject is required to repeat a task on several occasions while under the influence of the drug. Under these conditions, tolerance develops much more rapidly than it does if the same dosage is given after, rather than before, the performance of the task. To some investigators this suggests that toler- ance is a learned compensation for the drug effect, while others conclude that the increased arousal caused by the performance requirement facilitates the development of tolerance. In either case, the effect of learning is undoubted. The second type of tolerance- learning interaction is that involving classical Pavlovian conditioning. It has been reported by many investigators that tolerance develops more rapidly and lasts much longer if drug administration and tolerance testing are carried out in the same environment than if they occur in different environments. These findings make it quite clear that tolerance cannot be explained solely in terms of exposure of the nervous system to the drug. The response to the drug, and hence the development of tolerance, are also pro- foundly affected by the functional state of the organ- ism during the period of drug exposure. Conclusions based on only one set of circumstances are bound to lead to erroneous or over-simple conclusions. Most studies have shown that for a wide variety of drugs, species, and test procedures, tolerance, and 228 DRUG- AND ALCOHOL-RELATED PROBLEMS what has been termed physical dependence, appear to develop more or less in parallel, with respect to rate and degree. It has also been observed that acceleration of tolerance development, by means of behavioural manipulations as mentioned earlier, similarly acceler- ates the development of signs of "physical depen- dence" (i.e., of withdrawal reactions). Exposure to repeated cycles of drug administration and with- drawal accelerates the reacquisition of both tolerance and "physical dependence". Drug withdrawal signs are usually mirror images of the corresponding acute effects of the drug in question. For all these reasons, it is generally believed that tolerance and physical dependence are manifestations of the same adaptive process. Yet there is not total unanimity on this point. Many researchers have observed tolerance in the absence of overt "physical dependence", or vice versa. Though one cannot be categorical about it, the most probable explanation of these discrepancies is the use of inappropriate methods. Numerous hypotheses have been advanced to explain drug tolerance in terms of neuronal mechan- isms. Altered biochemical, biophysical, or neuro- physiological properties of neurones, synapses, and neuronal circuits have all been proposed. Enzyme induction, drug-receptor induction, altered coupling of receptors to intracellular effectors such as adenylate cyclase, altered membrane composition and structure, altered biosynthesis and release of various neurotransmitters, and altered modulation of primary neuronal pathways by peptidergic feedback loops have all been proposed. For all of these ideas there is some evidence, but at the same time all are rooted in rather too simple models of brain function. In the past, interest has been directed mainly toward the consequences of intensive exposure to a drug, including tolerance and physical dependence. If research is to yield insights into the causes of heavy drug use, and the possible relation of this behaviour to other aberrations of behaviour, motivation, and per- ception, it is important to redirect the research speci- fically towards the relation between tolerance and the acquisition rather than the effects of drug-taking. Neuroadaptation The term "physical dependence" continues to cause confusion. Dependence is not dependence if drug-taking, or at least the desire to take drugs, is absent. For example, surgical patients given opiate substances to relieve pain may sometimes experience withdrawal symptoms but have no desire to continue taking drugs. Should the biological changes induced in these patients be called physical dependence or should we consider using another term for this state? The term neuroadaptation is one possible alternative. The nature of this neuroadaptation is such that if drug administration is stopped or the drug displaced from its site of action, a syndrome is not necessarily aversive, although in some cases (e.g., opioids) it can be aversive and dysphoric in the extreme. Further- more, it does not necessarily give rise to drug-seeking behaviour, even when its manifestations include a considerable degree of subjective distress. For example, an individual may appreciate that opioid withdrawal is the cause of some internal distress, but may elect to forgo additional drug use even when it is available. Whether a withdrawal syndrome associated with a given drug leads to drug-seeking depends on many factors besides its characteristics (the degree to which it is experienced as aversive), its intensity, its expected duration, and the characteristics and past experiences of the individual. Several points need to be emphasized. At present, the degree of neuroadaptation is inferred from the appearance of the withdrawal syndrome. It may in the future be possible to measure directly the biological changes underlying the neuroadaptive state for differ- ent drugs, and thereby quantify the state without requiring the individual to experience the withdrawal syndrome. A neuroadaptive state is not necessarily a major disability. Thus, the caffeine withdrawal syn- drome can be annoying, but it is not life-threatening. Like tolerance, the withdrawal syndrome can be influenced by processes of learning. Once developed and linked by learning processes, the withdrawal phenomena can be intensified or minimized by emotional states and can be evoked by environmental or emotional stimuli long after the drug has been with- drawn. These stimulus-evoked withdrawal phenom- ena may be associated with thoughts about the drug and by subjective increases in craving for the drug. Finally, it has been shown that when withdrawal symptoms are precipitated in dependent monkeys, drug-seeking behaviour actually occurs before clear signs of withdrawal. A similar study in human subjects has shown that the subjective report of "craving" is much more sensitive than physiological or behavioural measures. Neuroadaptation refers to the neuronal changes associated with both "physical dependence" and tolerance. It is sometimes possible to exhibit neuro- adaptation without exhibiting the dependence syn- drome and to exhibit dependence with no evidence of neuroadaptation or at a time when neuroadaptation is no longer present. Drug dependence Drug dependence is a syndrome manifested by a behavioural pattern in which the use of a given psycho- active drug, or class of drugs, is given a much higher priority than other behaviours that once had higher 229 MEMORANDUM value. The term syndrome is taken to mean no more than a clustering of phenomena so that not all the components need always be present, or not always present with the same intensity. The dependence syndrome is not absolute, but is a quantitative phenomenon that exists in different degrees. The intensity of the syndrome is measured by the behaviours that are elicited in relation to using the drug and by the other behaviours that are secondary to drug use. On the basis of current knowledge, no sharp cut-off point can be identified for distinguishing drug dependence from non-dependent but recurrent drug use. At the extreme, the dependence syndrome is associated with "compulsive drug-using behaviour". Dependence on drugs is not always a major dis- ability. The drug may cause little tissue damage, little impairment of function, and be relatively inexpensive. At a time when neither the drug-using behaviour nor the neuroadaptive syndrome is manifest (drug dependence in partial remission) the continued pres- ence of the dependence syndrome may be manifest in the degree to which drug-related cues may elicit a desire for the drug and the effort that must be expended in order to resist the urge to continue drug use. The view that dependence is a clustering of phenom- ena (cognitive, behavioural, and physiological) implies that multiple criteria are necessary for its assessment. Evidence of neuroadaptation is just one of these criteria and should not be considered to be the most important. The phenomena or dimensions that together identify the syndrome(s) must be intensively researched in order to determine what they are, as well as the relative weights that must be given to each. They will probably include some of the following: - a subjective awareness of compulsion to use a drug or drugs, usually during attempts to stop or moderate drug use; - a desire to stop drug use in the face of continued use; - a relatively stereotyped drug-taking habit, i.e., a narrowing in the repertoire of drug-taking behaviour; - evidence of neuroadaptation (tolerance and withdrawal symptoms); -use of the drug to relieve or avoid withdrawal symptoms; - the salience of drug-seeking behaviour relative to other important priorities; - rapid reinstatement of the syndrome after a period of abstinence. We have noted previously that evidence of neuroadaptation is not sufficient to define drug- taking behaviour as dependent and yet we cannot reject this as one important criterion. Similarly, a desire to stop drug use, and difficulty in stopping are neither necessary nor sufficient, but when suitably weighted can give some indication of dependence. For example, many cigarette smokers who do not want to stop are actually dependent and would show other signs of dependence if they were encouraged to resist smoking for a short time. Dependence is discussed further in the section "Towards a useful model". Dependence and disabilities Not every individual who experiences impairment or disability related to drug consumption is suffering from drug dependence. There is no conceptually satisfactory cut-off point to differentiate persons exhibiting drug-related, but not syndrome-related, disabilities from the remainder of the population; a term such as "problem drinker" gives only a spurious concreteness. The drug-related disabilities of any indi- vidual are susceptible to the same multifactorial anal- ysis whether that person is or is not drug dependent. The complete description of an individual's drug- related pathology therefore comprises: (a) a statement as to whether he is or is not suffering from the dependence syndrome (with an added de- scription of degree of dependence and any modifi- cation of the picture); (b) a statement as to the kinds and degrees of disability; (c) a description of important personal and en- vironmental interacting factors that exacerbate or ameliorate the dependence or other drug-related disabilities. In this section, we have discussed fundamental ideas on the way a problem is viewed, the way it is defined, and the classification systems that have been devel- oped and which are closely bound up with an under- lying scientific framework. In the next section we shall move one step further and attempt to create a model of drug-seeking behaviour, giving special consider- ation to such questions as whether one overriding model (with variations) is applicable to all drugs, whether drug-seeking behaviour exists in a continuum of degrees or with a cut-off point defining a compul- sive syndrome, whether a compulsive syndrome itself exists in degrees, and to the relationship between biological, psychological, and social levels of explanation. TOWARDS A USEFUL MODEL A model to explain the initiation, continuation, and discontinuation of drug and alcohol use is fundamen- tally concerned with use of these substances, but it also helps to explain the meaning of the term dependence. Investment in the development of a model is con- sidered worthwhile not only for scientific purposes but 230 DRUG- AND ALCOHOL-RELATED PROBLEMS also because of the bearing that a more precise and integrated model must have on a variety of practical concerns. Much has been written on model building and on the meaning given to this term. These matters have been much discussed in terms of the philosophy of science, and as regards the possibilities of computer simulation. The degree of precedence given to model building varies among different scientific disciplines. In economics and various social sciences, for instance, such endeavours have been given much emphasis. In the sense in which we are using the term, there is no contradiction between models and common sense, and the designing of a model results in the organiz- ation of complex ideas which cannot be properly conceived or dealt with unless the relationships between sets of variables are clarified. A model is different from a map. A model must represent the dynamic relationships between these sets of variables and their interaction (often in terms of feedback loops) and not be merely a descriptive display of the variables potentially involved. A model involves the linking of postulates. The proposed model A schematic representation of this model is given below. 1. One set of stimuli bears on the individual and evokes the initial drug-taking behaviour. (a) This set of stimuli includes both external stimuli (taking due account of a wide range of social influ- ences, as well as more immediate situational cues), and internal stimuli, such as mood states. (b) The description of the individual must be multi- dimensional and take account of biological and psy- chological attributes. Genetic influences may have to be taken into account, as well as previous learning and previous social influences. (c) Drug-seeking behaviour is also to be viewed as multi-dimensional, for example in terms of dose and frequency, type, and route of administration. 2. Drug-seeking behaviour or drug use will lead to a range of consequences that may act as either deter- rents or reinforcers of further drug-seeking or drug use. These influences operate in at least three ways: (a) By direct impact on the individual or on the learning processes that are set up, and may be rein- forcing effects or aversive effects. (b) The individual's drug-taking will itself affect the range of external forces bearing on him. His behav- iour may excite disapproval and other negative res- ponses or may win social approval from society or a certain sector of society. SOCIAL AND INDIVIDUAL CONSEQUENCES 231 SOCIAL AND INDIVIDUAL ANTECEDENTS MEMORANDUM (c) Certain drug-related biological consequences will then bear either directly or indirectly on the way in which the experience of taking the drug influences the subsequent effect of further drug exposure. Here, for instance, we must consider the development of toler- ance, which may occur at a different rate for the positive and aversive immediate effects of the drug. The development of neuronal adaptation and with- drawal effects may also have both aversive and rein- forcing consequences; the withdrawal state and the associated distress may be immediately aversive in their impact, while repeated relief of withdrawal by further drug-taking may set up a conditioning process that further strengthens the drug-seeking habit. Toxic effects of the drug on the central nervous system may alter reactivity to the drug. 3. The potential importance of a number of other learning processes must also be taken into consider- ation. The following are given as examples: (a) By a process of generalization, the number of cues in the environment that may stimulate drug- taking is likely to be increased because these cues have been associated with relief of withdrawal symptoms, or are similar to other drug-taking cues. For instance, if anxiety is a component of drug withdrawal, then the general experience of anxiety may tend to excite drug- taking. If a particular setting or circumstance (even a particular room or geographical location) has been associated with drug-taking, these situations may sub- sequently become potential stimuli to a drug-taking response. (b) Because drug-taking becomes a strongly rein- forced response to a particular stimulus, this may lead to the lessening or extinction of other responses to that stimulus. For instance, the individual may be left with few responses to anxiety other than drug-taking. (c) It is likely that as drug-taking becomes less exploratory and more determined by a complex set of cues, the pattern of drug-taking will become increasingly narrowed and stereotyped. 4. Cessation of drug-taking on a permanent or temporary basis may be encouraged by a number of possible factors: (a) Lack of reinforcement might be related to a number of factors, including such factors as biological changes that alter the immediate impact of the drug, psychological changes that alter the value placed on drug effect, or, at the therapeutic level, the use of drugs that block the drug effect. (b) Aversive consequences such as, for example, the growing weight of negative responses from the environment, the increasing distress of withdrawal experiences, or an increase in other aspects of suffer- ing occasioned by drug-taking, may serve to suppress or extinguish drug-taking behaviour. (c) What may be termed "competitive reinforce- ment" can alter the balance of likely behaviours, for example when behaviours other than drug-taking are preferentially rewarded. This may reflect the impact of ordinary social circumstances, or of therapeutic strategies, and would include the development of coping or self-control strategies. (d) Lastly, if drugs are not available, they cannot be taken. Reduced availability would be associated with both fewer cues and an increased cost of response. 5. If drug-taking is restarted after a period of cess- ation, a variety of processes may affect the probability that the system we have described above is more or less rapidly reinstated. Such reinstatement will probably reach its prior level of organization in a much shorter time than in the naive subject. This may be due partly to biological processes that determine the rates of re- acquisition of tolerance of the neuroadaptive state, and withdrawal experiences; learned responses may also be rather rapidly reinstated and the responses of the external environment may also perhaps swing back to reinforce drug-taking (e.g., the person is again involved in a drug-taking subculture). But the process of reinstatement again involves an interreacting system rather than any one event. Where is dependence located in this model? Different people might read different meanings into this model, and therefore answer this question in different ways. We prefer to argue that dependence is essentially located within a system, and in this context we regard a syndrome as a rather simple translation of the concept of a system. The syndrome is to be understood as the various phenomena listed in the preceding section (pages 229-230) and the various relationships between these phenomena that we have outlined. Other people might prefer to see dependence in terms of what is going on within the individual, either physiologically or psychologically, or strictly in terms of behaviour alone, or in terms of the social role that the drug-user assumes. We believe that a system or syndrome model that seeks to take account of the interraction between drug, person, and environment, is much to be prefer- red. Any interpretation that places too much emphasis on only one part of the whole system is imperfect and misleading. Limitations of the model 1. The model does not attempt to define the means by which dependence is to be distinguished from drug- taking. We believe that in the present state of know- ledge that distinction is difficult. It may eventually be possible to make meaningful statements either in terms of the degree to which a certain phenomenon 232 DRUG- AND ALCOHOL-RELATED PROBLEMS has developed, the strength of the relationship between certain phenomena, or the introduction of new phenomena or new processes. In commonsense terms, what the present model would seem to propose is that a clinical or an operational definition of dependence must be multidimensional and, in terms of measurements, related to a number of phenomena within the syndrome. 2. Although the model displays what we believe to be many of the relevant phenomena and their many connexions, we are not able to state a degree of signifi- cance or weighting to be given to any particular factor or connexion. 3. The model is presented in general terms, without exploration of its specific application to particular drugs. 4. We have not attempted detailed analysis of the processes that may be going on in terms of any vari- able, or the nature of the relationships between variables. In many instances we are obviously drawing a circle around relative ignorance, and leaving the question unanswered. 5. Obviously there are many important aspects of drug-taking that we have not related to the model at all, or hardly at all. For instance, our model does not attempt to encompass ideas on the diffusion of drug- taking within a population. The usefulness ofsuch a model Whether or not such a model proves to be useful can only be determined by practical testing and by the responses and criticisms of many informed people. Our hope is that it may prove to be of some use in at least the following regards: 1. It may serve to clarify what we know and to reveal more clearly what we do not know. At a certain stage in the development of scientific understanding it may be profitable to put forward a model even if that model is wrong-it will serve as a basis for debate. 2. The model may help to organize strategies for scientific research. It may become evident that one question cannot properly be answered until another question is tackled, or that the efforts being devoted to understanding the different parts of the system are disproportionate. 3. By more accurately displaying the processes that may initiate drug-taking and serve to build the dependence system, we may be able to see more clearly the range of interventions that might logically be useful in prevention. 4. We believe that this model should also be exam- ined in terms of the leads it may give to multiple and alternative treatment strategies. Such examination may reveal which strategies might attack a certain system, and where, and with what likely impact on the whole system. 5. As regards classification, we have already dis- cussed the relationship indicated in the model between drug-taking and dependence, and the difficulties that are encountered in practice as regards cut-off points. We believe though that the model does help towards more accurate delineation of dependence syndromes, and that this is very important in the long term as regards classification. Although the model is fundamentally a scientific tool, and a tool to promote the application of scien- tific understanding to such concerns as prevention, treatment, and classification, we believe that what it proposes is also relevant to the needs of the commun- ity. For such popular purposes, the model obviously needs "translation". But, the way in which experts and professionals see these problems may influence the images that guide popular understanding, even as scientific thinking is influenced by the wider culture. We believe it is very important for us all to accept a model that admits the complexities and interactions, rather than to view the problem as being entirely due to the drug, or residing entirely within the individual, or as being concerned only with sociology. We hope that this model, despite its complexities and short- comings, will in some way lead to a more humane understanding of a human condition. ASSOCIATIONS BETWEEN DRUG AND ALCOHOL USE AND MENTAL DISORDERS Types of association People presenting for treatment of problems related to drug use often also have mental disorders. The associations are complex and variable, and can be: - mental disorders that cause drug-taking - mental disorders that are a result of drug-taking - chance associations. It should be stressed that etiological implications should be deduced only with great caution. Temporal relationships alone are no guarantee of a cause-effect relationship. For example, people with early schizo- phrenia may use cannabis for a variety of reasons- to expand their view of the world or to reduce anxiety feelings. At a later date it may be mistakenly con- cluded that cannabis induced the schizophrenia. However, in other cases the use of cannabis may be associated with an acute toxic psychosis of some hours' duration. In that case there would be grounds for attributing a causative role to the use of the drug. In East Africa, cases of more chronic toxic psychosis -lasting a number of weeks-have been observed with apparent abrupt onset following cannabis use. In this situation the role played by cannabis must remain 233 MEMORANDUM unclear until more is known about whether, and for how long, cannabinoids remain circulating at signifi- cant levels in the blood, the mechanisms involved in regulating the blood concentration, and any other factors that might be acting. Depression and alcohol use are often found in com- bination. The depression may be clearly primary, with drinking being seen as symptomatic and sequential, or the depression may be secondary (and sequential) to the drinking. In other cases the interrelationship may be two-way- someone drinks because he is depressed, and the drinking makes him more depressed. Multiple interactions Many factors influence the impact of drug-taking on mental health. The nature of the particular drug is obviously important, as well as the characteristics of the individual and the type of illness under consider- ation. Drug use may have both beneficial and harmful effects and in any individual the balance may vary at different times. Thus, in cases of schizophrenia, illicit drug use may at first provide the support of a drug- using group that might to some extent be protective and therapeutic. On the other hand, a psychotic experience induced by LSD or cannabis may precipi- tate a schizophrenic breakdown. Again, social use of alcohol might be of benefit to the person with mild depression, while heavy use would later make depression worse. Other influences besides the drug should also be considered. The individual's nutritional state is, for instance, often of importance. Hypoglycaemic coma in association with heavy drinking of alcohol is relatively common in countries such as Kenya, presum- ably because of coexisting nutritional deficiencies. Difficulties in establishing the relationship between drug-taking and mental illness Establishing the relationship between drug use and mental disorder may be complicated not only by the practical difficulties of determining whether the person has been taking a drug, but also by difficulties in deciding whether that person is now free from the effects of the previously taken drug or its active meta- bolites, or free from withdrawal effects. Difficulties also arise because the sampling of drug users that commonly come to medical attention is sub-ject to bias. Systematic collection of information on both drug use patterns among psychiatric patients and psychiatric disorders among drug-user populations would be most useful, and the value of such data would be increased if international comparisons could be made. Such information would be of assistance in deciding which treatment programme would best suit the needs of the individual. Effect of social approval on the harmful effects of drug-taking The use of a drug that is both widely used and socially accepted in a given culture is unlikely in that setting to be significantly associated with mental disorder, as long as the use remains within customary bounds. For instance, in the hill tribes of Thailand many children start using opium regularly from an early age, while in some parts of Central America chewing coca leaf (containing cocaine) is almost the norm. But in other instances, high rates of drug use may be symptomatic of social disruption, which may itself be conducive to problems in mental adjustment. Glue- sniffing is common among certain disaffiliated groups of young children in both Mexico and Kenya. These children are often orphaned and homeless and seem to be seriously disturbed emotionally. Where a drug is prohibited and unaccepted it tends to be used by those who indulge in other deviant behaviours as well, indicating a more general under- lying maladjustment. The influence of these different factors is illustrated in the contrast between the frequency of psychopathological troubles among Americans using heroin in the USA, and among those in the United States armed forces who started using heroin in Viet Nam. Drug use and personality disorders The use of drugs such as opium, cannabis, "coca", alcohol, or tobacco can be an integrated part of the life of some communities. Even the use of such drugs at levels producing a significant degree of neuroadap- tation may not be indicative of a psychopathological disorder. In such a context drug use may help the indi- vidual in various ways. For instance, kat may be used by the shepherd to help mantain his vigilance while protecting his flocks at night, and opium or coca may help the rural peasant to tolerate the privations of a hard environment. The positive aspects ofpersonality and situations In this field, previous classifications have always catalogued abnormalities of personality but have not taken into account the individual's positive attributes. Since disabilities related to drug-taking and prog- nosis may be as much modified by strengths and sup- ports as by weaknesses and disadvantage, it is essential for both clinical practice and research that the positive elements are given appropriate emphasis. Eventually this question will have to be addressed in any truly comprehensive, multiaxial approach to classification. 234 DRUG- AND ALCOHOL-RELATED PROBLEMS POPULATION STUDIES The development of a rational approach to classifi- cation of drug-related problems must rely in part on data derived from cross-sectional and longitudinal studies of populations of drug users. Cross-sectional studies help to define the current relationships between drug-taking, drug-related problems, depen- dence symptoms, personal attributes, and environ- mental variables; the clustering of variables and any patterns that are demonstrated may suggest new approaches to classification, or confirm or reject classifications that have been derived from clinical impressions. Life history studies of drug users are potentially very important: we can often learn how a syndrome regresses or develops over time, and under- stand the influences that are likely to bring about change. Indeed, it is essential to take a long-term view, in particular as regards the drug user's career, rather than to restrict enquiries to current aspects of the individual's situation. Population studies are in turn dependent upon standardized, reliable and valid methods of measurement and classification. The questions that are addressed in population studies are important not only for the formulation and validation of classifications, but also for the design of preventive and treatment policies. For example, one needs to know the levels of drug use that are associated with certain levels of prevalence and certain types of harmful consequences, and also what changes in behaviour take place in the absence of organized attempts to intervene. Do particular methods of detec- tion, early intervention, or treatment significantly alter these changes? How many people need, or would make use of, treatment services? What is the likely impact of a given preventive policy? However, even an apparently simple question, such as "How many alcoholics are there in a particular society?" is beset by difficulties as to what exactly constitutes an alcoholic, and how alcohol dependence should be measured. The assessment of problems caused by drug use is also clearly a function of our definitions. It is difficult to make any cross-cultural comparisons until the terminology has been clarified. The ultimate challenge is of course greater than just how to define a case or how to measure the extent of the problem. Our ultimate concern, when considering population studies and the drug user's life history, should be to build a framework of understanding which links the extent and distribution of drug use, drug dependence, and drug problems to various ante- cedent, or current, psychological and socioeconomic factors. The complexity of such an endeavour should not be underestimated. Facing the complexities As we have indicated in the section that begins on page 230, the focus should not be just on the person or on the drug-taking but on an interacting system. Thinking on drug epidemiology has seldom faced up to the real conceptual difficulties. The complexity is evident when one attempts to consider the relationship between the immediate and distant antecedents of drug use and the immediate and distant consequences. Antecedents and consequences can be measured with- in various domains-for example, psychological, sociological, and biological-and these domains are not independent. Antecedents and consequences may be global, national, cultural, or individual, and depend upon the nature of the population under investigation. Drug availability or licensing laws, international controls, cultural attitudes, personality attributes, genetic endowment, parental relationships and early life experience, employment, and demographic charac- teristics have all been implicated as significant antecedents of drug use. Harmful consequences or disabilities may be physical or mental disorders, impairments of certain skills, and many other social and psychological dysfunctions. The beneficial conse- quences of drug use must also be included in the final equation. In addition to the many antecedents and conse- quences, a further complication is the diversity of drug-taking behaviour. People may use many different drugs and drug combinations at one time. This highly complicated system would be difficult to investigate if it were stable over time, but it is not. It is not easy to study a complex and continually changing set of relationships, and yet this is the problem that has to be faced when we attempt to understand human behaviour in its social and cultural context, rather than dealing only with discrete disease entities. The best models in this area will always express relationships in terms of probabilities rather than certainties. In the area of drug use, the state of knowledge is such that the categories we employ are often arbi- trarily and empirically based. This is easily forgotten, and categories are then mistaken for true entities. Within a complicated system, any one significant factor will account for only a small percentage of the total variance. Progress will be made in this area only if the limits of our knowledge are freely admitted. Of course, any dynamic system involves two-way interactions. For example, dependence, withdrawal symptoms, and marital breakdown can all be conse- quences of drug use as well as antecedents for further drug use. Similarly, pattern of use can be both the dependent variable under investigation or a predictive index of future consequences. 235 MEMORANDUM Measurement of antecedents Surveys of drug use have tended to focus upon social and demographic antecedents. Drug use has been related to age, sex, marital status, social class, urbanization, occupation, ethno-religious grouping, and personality variables. Most of these antecedents are associated to some extent with some types of drug use in some societies and there is a need for standard- ized, reliable, and valid methods of assessing these and other antecedent variables that are simple and easy to administer. More demanding techniques will often be required for the sub-studies that are sometimes carried out at the same time as a general survey. Past drug use is an important antecedent variable when attempting to predict future use and future consequences so that various methods of assessing past use and even life- time consumption patterns need to be investigated, and are discussed below. Measurement and categorization of drug use and dependence Discussions concerning methods for measuring and categorizing drug use often seem to end up with the plea that the scientific community should agree on standardized approaches and get down to the business of designing a set of reliable and valid instruments for these purposes. Only on that basis, it is argued, will results be comparable over time and between different cultures, and only a disciplined commitment to uni- form methods will offer any real hope of advance. The temptation is to hector one's fellow scientists for their recalcitrance. It might be better to stand back for a minute and ask dispassionately and objectively why there has been such chronic difficulty in securing any common rubric on drug-use categories. We might do well to ask whether we have been trying to force the pace too quickly and to examine some of the real difficulties that stand in the way of any neat and immediate solution. The two most basic difficulties are those already emphasized in preceding paragraphs. Firstly, there is the sheer complexity of the issues involved and the multiplicity of variables to be encompassed. If one tries to design guidelines for survey work, they easily become so complicated and all-embracing as to be quite impractical, or one becomes overwhelmed with doubt as to the worth of the whole exercise and with the imperfections of any proposed solution. Secondly, there is the lack of any comprehensive model or scien- tific rationale for the setting of priorities, or for determining what really is or is not worth measuring. In such circumstances it is hardly surprising that the scientific community fails to agree on any rules. The whole argument in this Memorandum is that classifi- cations must be based on a proper scientific rationale, and that much work remains to be done. As far as population studies and the classification of drug-use patterns are concerned, we consider that the way ahead can be summarized as follows: (a) There would be immediate advantage if every- one conducting population studies would ensure that their categories for description of drug use and their operational definition of dependence were made explicit, whatever the approach employed. (b) In terms of the model proposed on page 231, population studies would be much helped by better definition and description of the dependence syn- dromes that are associated with the use of different drugs. Intensive investment in the development of standardized instruments for use in survey work would then be very fruitful, and would lead towards a theory-based measurement of important variables, rather than the gathering of numerous data which may in any case not be very illuminating. The design of schedules to cope with dependence arising from multiple substance use represents a further challenge. (c) It is more difficult to see the way ahead in relation to the categorization of actual drug-use pat- terns. As regards the consequences, it may be neces- sary to categorize the quantity of drug used in the recent past or over longer periods. Route of adminis- tration must also be recorded. To link in with theor- etical concepts on dependence, emphasis should also be put on precise observations on schedules of self- administration, such as the degree of variability in drug-use patterns, the spacing of administration during a 24-hour period, parameters of frequency as well as quantity, and the peak levels of intoxication probably achieved. Before finalizing a decision it would be valuable to design and test such schedules on a national and international basis, and explore their reliability and validity. (d) Advances are also being made in the measure- ment of cigarette-smoking and nicotine consumption. The number and type of cigarettes consumed per day is a useful measure, but depth of inhalation is also important. The carbon monoxide levels in expired air can be easily and accurately measured with relatively cheap equipment. This provides an index of depth of inhalation and correlates with blood nicotine levels. (e) In accordance with the emphasis on a longi- tudinal perspective, when basic schedules are available for the study of dependence syndrome concepts and patterns of use, it is vital that these instruments should be employed not only in cross-sectional but also in longitudinal studies. Population studies and the consequences of drug use Reaching agreement on measurement of the conse- quences of drug use has proved as difficult as the 236 DRUG- AND ALCOHOL-RELATED PROBLEMS search for consensus and uniformity in approaches to measuring antecedents, dependence symptoms, and drug-use patterns. Here, any cut-off points as to "what counts as harm" are empirical. However, some suggestions can be made as to matters in this area that need testing in population studies, and the approaches that might be used: (a) There should be a willingness to record the population occurrence of drug-related problems in an accurate and straightforward descriptive fashion, with subsequent analysis to determine the magnitude, clustering, and ranking of such problems in different populations. (b) There is a particular need to understand how social, cultural, and family processes may protect against, exacerbate, or modify the consequences of drug use. (c) Empirical approaches should take into consider- ation the problems experienced by families, friends, employers, and society in general as a consequence of an individual's drug use. In this sense, we should try to obtain population estimates of the problems that society is incurring, and to develop methods of categorizing and costing them. (d) Although many of the problems that result from drug use are social and psychological, the need for physical measurements in population studies should not be forgotten. Enzyme tests of liver func- tion in alcoholism provide an example of the methods that might usefully be employed. Research strategies We believe that population studies provide a particular opportunity for collaborative research, and that such studies can be valuable in many ways. It will often be necessary to make a preliminary and costly investment in the development of suitable instru- ments, the testing and development being carried out collaboratively between different centres. We would emphasize particularly the need to relate the precise scientific endeavour involved in instrument design to the realities of local experience. Instruments must derive from informed sensitivity to many cul- tural realities and needs, and schedules designed in one country and exported heedlessly to another area are of little use in drug studies. CLASSIFICATION SYSTEMS The previous sections of this Memorandum have analysed the ideas that should form the basis of any classification in this area. In this section we shall consider the implications of the ideas developed elsewhere, and examine some general principles that must influence any classification. We have not pro- ceeded, however, to a detailed examination of the ninth revision of the International classification of diseases (or any national approach), since we believe that this would be premature. Special considerations relating to classification of drug- and alcohol-related problems We believe that a number of special difficulties con- front any system of classification in this particular area. We list them as follows: 1. Drug dependence is not an all-or-none phenom- enon. In many ways, we are dealing with continuities rather than categories. 2. Attempts to categorize the use patterns for even one drug soon demonstrate the very great complexity of the problem. It is difficult to know how to describe in simple terms the possible combinations of vari- ations in quantity of drug and frequency of use. 3. In practice the individual is often taking more than one drug. It is very difficult to know whether such individuals should be classified primarily as takers of one drug rather than another-patterns are almost infinitely variable. When the evolution of drug-taking patterns over time has to be considered, the situation becomes even more complex. 4. Defining the route ofuse is also an important and difficult problem. Are tobacco-chewing, snuff- taking, cigarette-smoking, and the smoking of cigars or pipes to be regarded as the same for the purposes of classification? 5. The focus of the health concern is frequently on the disabilities that may result from excessive use of drugs or alcohol. The many ways in which these dis- abilities may cluster, and the different degrees which any one of them may exhibit, produce a myriad of possible clusters. One is not dealing with linear scales or adding like to like. It is therefore rather arbitrary to set up classifications based on "degree of disability" -"problem drinking" is, for instance, very difficult to define. How can the magnitude of a problem or of the sum of separate problems, be indicated in a work- able, problem-based category? 6. Many pictures of mental disorders are much coloured by personality and cultural presentation. This is particularly true of syndromes that result from excessive use of psychoactive substances. 7. Although in the field of general psychiatry there is still considerable debate as to how presentation should be subclassified (e.g., subtypes of depression or schizophrenia), in the field of drug and alcohol abuse the debate is still often at a stage of uncertainty as to whetherparticular syndromes actually exist. The status of cannabis psychosis provides an example. 8. Concern with drug problems inevitably raises 237 MEMORANDUM issues in the physical and psychological domains, and often in particular the social domain. This leads to suggestions for multi-axial classification, but in this case the number of possible axes is very large. 9. Although many aspects of mental disorders are covered by national legislation, drug problems are probably more enmeshed in national and inter- national legal frameworks than any other aspect. There is therefore the additional difficulty of develop- ing a scientific and medical classification that also meets the needs of legal provisions. Multi-axial classification We have already discussed the factors that indicate the need for a multi-axial classification of drug and alcohol problems. In practical terms there must be a limit to such developments-there are limits to what the diagnostician can cope with or the diagnostic record sheet can handle. Greater complexity in the diagnostic system would be particularly unwelcome in developing countries, where diagnosis is often under- taken by medical assistants rather than by fully trained medical practitioners. There is, perhaps, a danger of trying to devise all-embracing classification systems, rather than leaving the description of the case to what might alternatively be called the clinical work of " for- mulation". However, whether the framework being discussed is that of classification or of formulation, the dimensions must include at least the following: the drug or drugs used, duration of use, route of adminis- tration, drug-use pattern, presence or absence of dependence and its degrees, drug- or alcohol-related disabilities (and their clustering and degree), any accompanying mental pathology, an assessment of the strengths and weaknesses of the individual's person- ality, and the degree of subcultural involvement in drug-taking. This is obviously a very considerable list, and what is needed is further examination of minimal criteria, and the relevance of such criteria to particular needs. Strategies for advance The following strategies are proposed for advance in this area: (a) Analysis of the scientific bases of classification. This is, of course, what has been attempted in pages 226-237. (b) A willingness to apply precise case observation and analysis of small groups of cases. We believe that this traditional method still has much to offer in this particular field. (c) Population studies that will give epidemiological and statistical information on each type of presen- tation, its frequency of occurrence, and its develop- ment through time. (d) The further development of laboratory and special investigation techniques to examine such matters as the causal relationship between drug-taking and particular presentations (e.g., in cannabis psycho- sis), or the presence of minor pathologies that are not always easily clinically identifiable (e.g., minor degrees of alcohol-related brain damage). Operational studies ofsystems of classification Given the many difficulties that beset any attempt to establish a valid basis for classification in this area, and the added difficulties that confront any attempt to use the proposed terms or make them workable in a different cultural setting, it is unrealistic to put enor- mous effort into abstract discussion of classification systems without examining how any such system is likely to be employed in practice. The utility of any system for different groups and interests within any one country needs to be explored, as well as its cross- national value. Studies of reliability are urgently needed. To what extent does any particular system of classification have social consequences? How does the proposed system modify our awareness and our responses? RECOMMENDATIONS In this section, three main ways of developing and improving the approaches to classification systems in the field of drug- and alcohol-related problems are emphasized: recommendations concerning nomencla- ture, research relevant to drug- and alcohol-related problems, and studies of the nature, variety, and uses of classification systems in various countries, and in particular their use in diagnosis. The research recommendations are pertinent to the model of drug dependence proposed on pages 230-233, and were formulated following the discussions of the complex interactions between mental disorders and drug- and alcohol-related problems (pages 233-234). The section on classification clearly established the need for studies on the nature and uses of existing systems of classification. Although priorities in this field must clearly be closely related to community and country problems, we realize that some of the ideas addressed below may not be of high priority in some countries, and that questions of more exact classification may at the present seem absurdly sophisticated in countries where medical services are struggling with the first stages of development. Nevertheless, we believe that everyone will find in these recommendations some issues that capture their attention and some proposals which in the long- or short-term must bear on their needs. There is enormous opportunity here for inter- 238 DRUG- AND ALCOHOL-RELATED PROBLEMS national cooperation. 1. Nomenclature (a) Drug. A "drug", in the broadest sense, is any chemical entity or mixture of entities used for treat- ment or alleviation of disease, or for non-therapeutic purposes. Psychoactive drugs are those that alter mood, cognition, or behaviour. There is no sharp boundary between "psychoactive drugs" and those not so designated, since drugs used primarily for other purposes may have "psychoactive effects" at certain doses or under certain conditions. Our concern here is primarily with those psycho- active drugs that are likely to be self-administered and have high reinforcement properties (see below), some of which are used non-therapeutically and others of which have important medical uses. We can character- ize certain general properties of this class of sub- stances, but there is no unique feature that can be used to identify them all. Most, but not all, of them can give rise to tolerance, drug seeking-behaviour under various conditions, neuro-adaptation, and response generalization (see below). (b) The term used for non-therapeutic purposes is meant to convey that the drug is self-administered by the user, and that the decision to use it is unsanctioned by local medical custom or the assumptions of folk- medicine. Because this term is often awkward, we suggest that consideration be given to using inter- changeably the term self-administered (recognizing of course that even prescribed therapeutic agents are generally self-administered). Self-administered has the advantage of being a term that is used in labora- tory studies of psychoactive drugs. Proposals for a term shorter than "used for non-therapeutic purposes" and less subject to misinterpretation than "self-administered" should, however, be considered. (c) Reinforcement potential is a property of a drug that determines the likelihood of its being self- administered under a variety of environmental con- ditions and schedules of administration. The species, route, conditions, and schedule need to be specified, since some drugs may be self-administered by one species but not another. Reinforcement potential is not an absolute property. There are powerful re- inforcers which under a variety of conditions will evoke prodigious efforts by the organism to self- administer the drug, and weak reinforcers self- administered only under special conditions and only when the amount of work or effort required to res- pond is relatively small. The correlation between the reinforcement potential, as measured in laboratory experiments with animals, and the behaviours seen in human populations, is high enough to be useful-but there are important exceptions. (d) Neuroadaptation. We believe that the term "physical dependence" should be abandoned and replaced with the term neuroadaptation or neuro- adaptive state. The nature of this neuroadaptation is such that if drug administration is stopped or the drug displaced from its site of action a withdrawal syn- drome, a constellation ofsigns and symptoms develops that is characteristic for the particular drug (or category of drugs) and for the specific biological system, or species. Neuroadaption is almost always accompanied by some degree of tolerance to some of the effects of the drug (see below), and it is therefore assumed that the basic processes underlying these two phenomena are related. It is sometimes possible to exhibit a neuro- adaptive state induced by a given drug without exhibit- ing the corresponding dependence syndrome (see below), or alternatively to exhibit the dependence syn- drome at a time when neuroadaptation to the given drug is no longer present to any significant degree. (e) Reciprocal neuroadaptation is a relationship among chemical entities in which one drug has the capacity (a) to prevent the withdrawal manifestations that would emerge upon discontinuation or removal of another and (b) to maintain a similar type of neuroadaptive state. (f) Some drugs have the capacity to suppress clinically significant withdrawal phenomena, but do not maintain similar neuroadaptive states (e.g., clonidine suppresses some aspects of opioid with- drawal but does not maintain the opioid-type neuro- adaptive state). At present, there is no generally accepted term for this relationship. We provisionally suggest the terms non-reciprocal suppressant or disjunctive suppressant or disjunctive neuro- adaptation. (g) Tolerance is of several varieties. Pharmaco- dynamic tolerance is present when, after exposure to a drug, higher levels of the drug are required at its site of action to produce a given response. Metabolic toler- ance is the increased capacity to metabolize a drug; it can be induced by the substance itself or by some other agent. Acute pharmacodynamic tolerance may develop after repeated exposure to the drug. There is not a close parallelism between the capacity of a drug to produce tolerance and its reinforcement potential. (h) Loss of tolerance to alcohol and other drugs may be related to neuronal damage or to altered meta- bolic clearance. It may not occur with all drugs and is a phenomenon requiring investigation. (i) Cross-tolerance is the phenomenon that occurs when an organ system that is tolerant to one chemical entity exhibits tolerance to another chemical entity. This is subsumed under the idea of reciprocal neuro- adaptation (paragraph (e) above). (j) Psychological dependence is a term that requires further examination. Since we are very much 239 MEMORANDUM seeking to emphasize the importance of a holistic dependence syndrome, we considered whether it might be better to substitute some term such as " learnt drug-seeking" but a final decision was not reached. (k) Drug dependence- the dependence syndrome. We wish to avoid the dualism inherent in the use of the terms psychological dependence and physical depen- dence, and promote the concept of a syndrome that is not bound by these old ways of thinking. Drug dependence is a socio-psycho-biological syn- drome manifested by a behavioural pattern in which the use of a given psychoactive drug (or class of drugs) is given a sharply higher priority over other behaviours which once had significantly greater value, (i.e., drug use comes to have a greater relative value). Given the complexity of the interactions that contribute to the genesis, perpetuation, or "decay" of the system that underlies this behaviour, a description of the syn- drome is easier than a brief definition, but we would suggest that a key descriptive element is the priority given to drug-seeking over other behaviours. It is not clear whether there are fundamentally distinct types of dependence syndrome corresponding to the various drug categories (as appears to be the case with the variety of neuroadaptive states) or whether it will prove useful in the long run to continue to limit our descriptions of dependence syndromes to "depen- dence of the morphine, alcohol type", etc. (l) Abuse and misuse. We recommend replacement of these terms by the concepts of unsanctioned use, hazardous use, harmful use and dysfunctional use. 2. Research (a) Drug dependence and its constituent phenomena (i) Neuroadaptive syndrome. Experimental studies in animals can investigate the reinforcing properties of drugs only by examining patterns of self-adminis- tration. It is recommended that priority be given to studies of the brain neural mechanism(s) of drug- reinforcement, to investigate whether there is a single common mechanism or multiple drug-specific mech- anisms. Such studies should include investigation of the possible relation of tolerance to the gradual increase in self-administration of drugs and alcohol seen in both human and experimental animals during the development of dependence. Studies of the adaptive syndrome itself are needed. Efforts should be made to identify the biochemical or physiological manifestations of this syndrome per se, rather than the withdrawal reactions that it may pro- duce. This would clarify in a more direct way the nature of the syndrome itself without relying so much on the characteristics of the withdrawal reaction. It would also enable this phenomenon to be studied without necessarily subjecting people to the withdrawal reaction. (ii) Dependence. We recommend the development of reliable objective measures of the severity of dependence. Withdrawal. It is recognized that considerable variation occurs in the pattern, time of onset, and duration of the withdrawal reaction, even among indi- viduals using the same drug. Therefore, we recom- mend that systematic observations be undertaken in clinical and natural settings to identify variation in the pattern of the withdrawal reaction as a function of drug, personal, immediate environmental, and social variables. Further observations should be carried out on patients receiving prescribed psychoactive medication (e.g., tricyclic antidepressants) to see whether a with- drawal reaction can be detected that is distinct from a return of the disturbance for which the medication was originally prescribed. It would be of particular importance to determine whether a withdrawal reac- tion, if present, affects the future course of drug use and the occurrence of drug-related problems. (iii) Studies of alcohol and morphine. Recent experimental work has revealed a number of pre- viously unrecognized areas of limited cross-tolerance between alcohol and morphine. It is recommended that priority be given to experimental studies aimed at careful delineation and explanation of cross- tolerance and possible common mechanisms of rein- forcement between these and other drug classes. Such studies should not be focused on any single current biochemical hypothesis, but should draw on all available knowledge on the mechanisms of drug- action adaptation. (b) Drug use and related damage (i) In the field of alcohol research, a number of biochemical markers have recently been identified, which may make it possible to recognize the intake of ethanol during the preceding few days or weeks. It is recommended that analogous markers, preferably quantitative, be sought for use in screening surveys or monitoring, to test for recent use of psychoactive drugs other than ethanol. Such methods would be invaluable in attempting to correlate health problems with level of drug use in both individuals and populations. (ii) It is recommended that, with or without the advantage of such screening methods, epidemio- logical studies be carried out to correlate patterns and amounts of non-medical drug use with the frequency and severity of various types of health disorder. (iii) Both clinical and experimental studies have indicated that relatively moderate intake of alcohol, and possibly of cannabis, may be associated with minimal degrees of brain damage. It is recommended ')An DRUG- AND ALCOHOL-RELATED PROBLEMS that this relationship be investigated more thoroughly, and for other drugs, with respect to frequency and level of drug use, reversibility, mechanisms of damage, and possible relation to loss of tolerance occurring later in the course of drug exposure. (iv) It is widely believed that most of the diagnostic entities related to alcohol and drug use and depen- dence are complex, containing multiple subtypes with different causal factors, different prognoses, and possibly different optimal therapies. Therefore, it is recommended that detailed prospective clinical case studies be carried out, to validate clinical impressions and concepts (e.g., the existence of "alcoholic jeal- ousy"). Advantage should be taken, whenever poss- ible, of populations on which extensive data are already in hand, with specific subtypes of drug and alcohol users chosen from these groups. (c) Drugs and alcohol use and psychopathology (i) Research is urgently needed to determine the extent and the nature of the relationship between various forms of drug use and dependence, and psy- chiatric disorders. Studies should compare the fre- quency of occurrence of each form of substance use in the general population with that in clinical popula- tions whose primary diagnosis is schizophrenia, affec- tive disorder, or personality disorders. Similarly, efforts should be made to determine the relative frequency of occurrence of each psychiatric disorder in populations with a primary diagnosis of a drug- related disorder. These studies should be done both across cultures and in various populations within a given culture. Comparison of the results in males and females is of particular importance. Where possible such studies should be extended to non-patient or non- institutionalized populations. Such investigations will often require collaborative studies with carefully standardized instruments to assess both the drug-use disorder and the psychopathology. (ii) Cross-sectional and longitudinal comparisons should be made of groups with alcohol- and other drug-use disorders, with various demographic charac- teristics, including sex, age, socioeconomic status, residence, etc., at different ages, in order to answer the following questions: - Are there common characteristics of dependence regardless of the drug and, if so, what are they? - What are the differences between cultures, and what are the common characteristics? We stress the importance of taking a long-term view. (iii) Investigations are needed, to determine whether individuals who show "compulsive" drug or alcohol use are qualitatively different from those who use these substances heavily but apparently not com- pulsively, or whose drug use differs only quantitat- ively. If a qualitative difference were found, it would perhaps be fruitful to explore the common charac- teristics of compulsive disorders by comparing com- compulsive drug users with individuals suffering from other disorders of control of behaviour, including compulsive neurosis, over-eating, and compulsive gambling. (iv) Most clinicians believe that there is an entity that warrants diagnosis as multiple drug use. How- ever, the criteria have not been clearly established. Groups of individuals using multiple drugs in various ways should be studied by a range of epidemiological, laboratory, and individual case-study techniques. (d) Linguistics It would be of great interest to assemble compre- hensive information on the words used in different cultures to describe such concepts as drug, medicine, "drug abuse", addiction, intoxication, and so on. (e) General considerations The model that was outlined in the section starting on page 230 generates a number of suggestions for research. For example, there is a need to develop reli- able and valid methods of assessing the antecedents, the patterns of use, and the consequences of drug use. Also classifications could be devised that are based upon the clustering of antecedents or consequences or both. 3. Classification and diagnosis A study is recommended of the nature, variety, and uses of classification systems in various countries. This project would address the following questions: how important are classification systems in practice, how are they actually employed, what is expected of them, and what results are obtained from their use. This project should have an initial feasibility stage to determine the prospects for use in developing coun- tries and to examine the systems in relation to the paramedical reporting of diagnostic information. The ways in which various classification systems are used in international legal frameworks and in international organizations, should also be investigated. The existing classification systems used for clinical diagnosis are not entirely adequate in the areas of drug- and alcohol-related diagnoses or substance- related mental or physical disorders. Therefore, we recommend that a small working group of experts on the methodology of diagnosis and classification should examine the existing systems in terms of classifying a given case according to several different systems. We believe that we are far from the stage of attempting to revise the classification, but it would 241 242 MEMORANDUM nonetheless be timely to initiate a series of studies to look at such issues as the combined or separate axes, and the dimensions that might be considered. Empir- ical field tests should then be conducted to see how various approaches would work in practice in differ- ent setting, and how well they might meet the many different potential purposes and requirements of a classification system. S. W. Acuda, Associate Professor, Department of Psychiatry, University of Nairobi, Kenya Hector R. Acufia, Director, Pan American Sanitary Bureau/WHO Regional Office for the Americas, Washington, DC, USA A. Arif, Senior Medical Officer in Charge of Drug Dependence Programme, Division of Mental Health, World Health Organization, Geneva, Switzerland (Secretary) E. A. Babayan, Head, Department of Evaluation of New Drugs and Medical Equipment, Ministry of Health, Moscow, USSR S. B. Blume, Director, New York State Division of Alcoholism and Alcohol Abuse, Albany, NY, USA L. R. Drew, Senior Medical Adviser on Alcohol and Drug Dependence, Department of Health, Can- berra, Australia (Rapporteur) J. G. Edwards, Professor of Addiction Behaviour, Institute of Psychiatry, Addiction Research Unit, London, England (Chairman) R. de la Fuente, Mental Health Director, Secretaria de Salubridad, Mexico, DF, Mexico (Vice-Chairman) R. Gonzalez, Regional Adviser in Mental Health, Pan American Sanitary Bureau/WHO Regional Office for the Americas, Washington, DC, USA R. Hodgson, Institute of Psychiatry, Addiction Research Unit, London, England (Rapporteur) J. Jaffe, Veterans Administration Medical School, Newington, CT, USA (Temporary Adviser) H. Kalant, Director of Behavioral Studies, Addiction Research Foundation, Toronto, Ontario, Canada M. Katatsky, Regional Adviser in Alcoholism and Drug Dependence, Pan American Sanitary Bureau/ WHO Regional Office for the Americas, Washing- ton, DC, USA G. L. Klerman, Administrator, Alcohol, Drug Abuse and Mental Health Administration (ADAMHA), United States Public Health Service, Rockville, MD, USA V. D. Kusevic, Tomasiceva 7, 41000 Zagreb, Yugo- slavia P. R. Ladewig, Head, Addiction Research and Treat- ment Unit, University of Basle, Basle, Switzerland P. O'Brien, Philadelphia Veterans Administration Hospital, Philadelphia, PA, USA V. Poshyachinda, Chief, Drug Dependence Research Centre, Institute of Health Research, Bangkok, Thailand R. Ryback, Department of Laboratory, ADAMHA, Rockville, MD, USA C. R. Schuster, Director, Drug Abuse Research Center, Department of Psychiatry, University of Chicago, Chicago, IL, USA J. P. Smith, Assistant Director for International Activities, ADAMHA, Rockville, MD, USA M. I. Soueif, Psychiatry Department, University of Cairo, Egypt H. Suwaki, Neuropsychiatric Department, Kochi Medical School, Kochi, Japan E. Zarifian, Centre hospitalier Ste Anne, Paris, France ACKNOWLEDGEMENTS The participants in the meeting wish to acknowledge the part played by the following members of staff of the Alcohol, Drug Abuse and Mental Health Administration, United States Public Health Service, who were present at the meeting as observers and made valuable contributions to the discussion: Dr J. Blaine, Dr B. C. Hatbone, Dr R. Hirshfeld, Mr T. Lawrence, Dr G. C. Salmoiraghi, and Mr L. Towle. REFERENCES 1. WHO Technical Report Series, No. 57, 1952 (Third report of the WHO Expert Committee on Drugs Liable to Produce Addiction). 2. WHO Technical Report Series, No. 116, 1957 (Seventh report of the WHO Expert Committee on Addiction- Producing Drugs). 3. WHO Technical Report Series, No. 273, 1964 (Thirteenth report of WHO Expert Committee on Addiction-Producing Drugs). 4. EDDY, N. B. ET AL. Drug dependence: its significance and characteristics, Bulletin of the World Health Organization, 32: 721-733 (1965). 5. WHO Technical Report Series, No. 42, 1951 (Mental Health. Report on the First Session of the Alcoholism Subcommittee). 6. EDWARDS, G. ET AL. ed., Alcohol-related disabilities, Geneva, World Health Organization, 1977 (WHO Offset Publication No. 32). 7. WHO Technical Report Series, No. 407, 1969. (Sixteenth report of the WHO Expert Committee on Drug Dependence).