Skip to Content

Psychodynamics of Drug Dependence - Overview

Edward J. Khantzian, M.D., and Catherine J. Treece, Ph.D.


Until recently, attempts to apply psychoanalytic theory to an understanding and treatment of drug dependence have been limited, and when attempted, have proven inconsequential.

Self-help and methadone maintenance approaches, which developed and proliferated in the 1960's, provided the mainstay of treatment for compulsive and chronic drug use.

The promise that these two new modalities seemed to offer, corresponding with rising alarm about a drug problem in this country of 'epidemic proportions,'' resulted in the development and expenditure of enormous resources at the Federal, State, and local levels to enlist large numbers of patients in treatment and rehabilitation.

Unfortunately, the hope that vast expenditures would result in speedy cures and expeditiously eliminate the problem has proven illusory. The development of this monograph represents a modest but significant departure from these recent trends.

In April of 1976, the Division of Research, Clinical- Behavioral Branch, under the leadership of Drs. Pollin, Renault, Julius and Blaine convened a group of psychoanalysts, psychiatrists and psychologists who had demonstrated an interest in understanding substance use from a psychodynamic point of view.

The participants were charged to reexamine psychoanalytic theory and its relevance for understanding opiate addiction. There was also an implied hope that some practical treatment applications could be garnered from such an exercise.

The preliminary results of this effort are represented by the content of this monograph. We believe it represents a good beginning. With the interest and support of NIDA, the participants of and contributors to the first Technical Review in 1976 and to this monograph, have, in our estimation, demonstrated the value and utility of developing and expanding on previous applications of psychoanalytic theory to problems of substance use.

The study of addiction as a human process begs for a depth psychology. We are convinced that psychoanalytic theory continues to be the most enabling and useful depth psychology at our disposal to understand the human mind and behavior, including addictive behavior.

Psychoanalysis does not invalidate other psychological methods of understanding, but rather attempts to account more adequately for the complexities of human behavior in terms of dynamic, economic, structural, developmental and adaptational factors.

For the unversed in psychoanalytic theory the contents of this monograph might seem too complex, obscure and far removed from an everyday understanding and management of drug-dependent individuals.

We believe that the theoretical formulations developed in this monograph reflect rather than obscure the complexities of the drug problems with which we work. We also believe that diligence in trying to comprehend this theoretical point of view will help to organize clinical observations and apply them more meaningfully and consistently in work with patients.

In turn, hopefully, our theory will more likely reflect data and observations obtained from what our patients tell us and what they experience. Clearly, the clinical observations about drug-dependent patients and the theoretical discussions in this monograph suggest a number of avenues for further exploration and research.

We believe that this beginning effort to involve psychoanalysts in national efforts to address problems of opiate addiction represents a promising vista for the development of a more dynamic clinical approach and a richer theoretical understanding of drug dependence. The intention of this chapter is to provide an updated review of the literature and to provide an overview of this monograph, stressing areas of agreement and complementarity, as well as unresolved differences, among the contributors.


Yorke (1970) and Khantzian (1974) have recently reviewed the early psychoanalytic literature on addiction. They have both concluded that the limitations of an excessively drive-oriented model prevented these early authors from fully developing and utilizing many important clinical observations.

Thus, there was an excessive emphasis on the libidinal gratification provided through drug use to account for the addict's involvement with his drug. Similarly, in their overemphasis on the symbolic meaning of the drug and how it was used, they failed to make distinctions between various classes of drugs and their distinctive psychopharmacologic effects.

Khantzian notes that Rado (1933, 1957), in particular, and Savit (1954), Fenichel (1945) and others, seemed to appreciate underlying depression and tension as motives for taking drugs.

However, in his review Khantzian concludes that these themes are not well developed and that too much emphasis is placed on pleasurable and regressive aspects of drug use to explain the compelling nature of addiction.

Although Glover (1956) also failed to appreciate the specific effects of different drugs and excessively stressed symbolic factors, Yorke and Khantzian have both noted that he better appreciated the adaptive, 'progressive'' use of drugs to cope with and defend against powerful, overwhelming and psychoticogenic rage and aggression.

The work of Chein et al. (1964) and the previous related work by Gerard and Kornetsky (1954) marks a significant shift in the psychoanalytic literature. Studying adolescent addiction in the ghetto rather than adult addiction in the analytic office provided new perspectives on which to build.

In addition to appreciating the specific psychotropic effects of opiates, their work more adequately focused on the addict's ego and superego pathology, problems with narcissism and other psychopathology.

They delineated some of the major difficulties that addicts have in engaging their environment, and stressed how their use of heroin was 'adaptive and functional,'' helping them to overcome crippling adolescent anxieties evoked by the prospect of facing adult role expectations with inadequate preparation, models, and prospects.

Subsequently, Hartmann (1969) developed similar themes through the study of other populations of adolescent drug users, focusing on the use of drugs, particularly by those with passive tendencies, to avoid active mastery of adolescent tasks in which narcotics were used to provide a passive regressive solution to intra-psychic conflicts associated with the adolescent phase of development.

Wieder and Kaplan (1969) elaborated further on this point of view, emphasizing that preadolescent developmental conflicts left certain individuals specifically vulnerable to problems of anxiety, depression, and physical discomfort during adolescence.

In such cases drugs seemed to provide the means to induce a desirable ego regression. Specific drugs were understood to be related to stage-specific developmental conflict.

Opiates, for instance, were said to produce a state reminiscent of a blissful closeness and union with the mother, which resulted in avoidance of separation anxieties aroused by the adolescent dependency crisis.

Despite a superficial resemblance to earlier formulations that stressed regressive pleasurable use of drugs, their work represents an important advance and elaboration of trends set in motion by Gerard and Kornetsky in the 1950's which utilized recent developments in ego theory enabling Wieder and Kaplan to appreciate that individuals self-select different drugs based on personality organization and ego impairments.

Their emphasis on the use of drugs as a 'prosthetic,'' and their focus on developmental considerations, adaptation and the ego, clearly sets their work apart from earlier simplistic formulations based on an id psychology.

Based on this and other recent work that considers ego and adaptational problems of addicts and following lines pursued by Wieder and Kaplan, Milkman and Frosch (1973) empirically tested the hypothesis that self-selection of specific drugs is related to preferred defensive style.

Using the Bellak and Hurvich Interview and Rating Scale for Ego Functioning, they compared heroin and amphetamine addicts in drugged and non-drugged conditions.

Their preliminary findings supported their hypothesis that heroin addicts preferred the calming and dampening effects of opiates and seemed to use this action of the drug to shore up tenuous defenses and reinforce a tendency toward withdrawal and isolation, while amphetamine users used the stimulating action of amphetamines to support an inflated sense of self-worth and a defensive style involving active confrontation with their environment.

Similarly, Hendin (1974), using psychological testing and interview data, concluded that heroin as well as barbiturates acted to assist in withdrawal and to avoid intimacy and thus defend against overwhelming destructive impulses.

In contradistinction, the works of Wurmser (1972, 1974) and Khantzian (1972, 1974, 1975) suggest that the excessive emphasis on the regressive effects of narcotics in these studies is unwarranted, and that in fact, the specific psychopharmacologic action of opiates has an opposite, 'progressive'' effect whereby regressed states may be reversed.

Wurmser believes that narcotics are used adaptively by narcotic addicts to compensate for defects in affect defense, particularly against feelings of 'rage, hurt, shame-and loneliness.'' Khantzian stresses drive defense and believes narcotics act to reverse regressed states by the direct antiaggression action of opiates, counteracting disorganizing influences of rage and aggression on the ego.

Both these formulations propose that the psychopharmacologic effects of the drug can substitute for defective or nonexistent ego mechanisms of defense. As with previously mentioned recent investigators, Wurmser and Khantzian also consider developmental impairments, the severe, predisposing psychopathology, and problems in adaptation as central issues in understanding addiction.

Radford, Wiseberg, and Yorke (1972) reported detailed case material that supports the findings of Wurmser and Khantzian that opiates can have an antiaggression and antiregression action or effect. They further observe that opiate use cannot be exclusively correlated with any particular patterns of internal conflict or phase-specific developmental impairment.

Despite differences in emphasis, the work of Gerard and Kornetsky, Hartmann, Wieder and Kaplan, Milkman and Frosch, and Hendin shares in common a stress on opiate use as an attempt to correct impaired or defective ego functions and thereby assist the individual to cope.

It is also recurrently evident in the work of these investigators that this attempt is only partially successful at best. Their findings repeatedly seem to suggest that adopting passive solutions through opiates induces a self-perpetuating tendency for maladaptive and pathologic ego and drive regression.

This apparent contradiction between the adaptive and maladaptive effects of opiate use implied in these formulations awaits further clarification. Krystal and Raskin (1970) are somewhat less precise about the specific effects of different drugs, but allow that they may be used either to permit or prevent regression.

However, their work does focus much more precisely on the relationship between the affects of pain, depression, and anxiety, and drug and placebo effects. Addicts' difficulties in recognizing and tolerating painful affects are explored and greatly clarified.

The tendency for the affects of depression and anxiety to remain somatized, unverbalized and undifferentiated in addicts, results in a defective stimulus barrier and thus leaves such individuals ill-equipped to deal with their feelings, and predisposes them to drug use.

Their work also focuses in greater depth on the major problems that addicts have in relation to positive and negative feelings about themselves and in relation to other people. Krystal and Raskin believe that addicts have major difficulties in being good to themselves and in dealing with their positive and negative feelings toward others because of rigid and massive defenses such as splitting and denial.

They maintain that drug users take drugs not only to assist in defending against their feelings, but also briefly and therefore 'safely'' to enable the experience of feelings like fusion (oneness) with loved objects, which are normally prevented by the rigid defenses against aggression.

The problem of ego regression among narcotic addicts has been addressed by Zinberg (1975) from a rather different perspective than most of the other authors cited here.

Zinberg minimizes factors of psychopathology and proposes an alternative explanation for the uniformly regressed appearance and behavior of addicts. Zinberg develops his thesis around the concept of relative autonomy of the ego.

He suggests that the enforced social isolation and deviant status of illicit narcotics users result in a sufficient reduction of balanced input from external reality as to undermine the ego's relative autonomy from the id, and simultaneously affect superego structures which are maintained by social supports.

The cyclical nature of addiction in which the user continuously cycles from high to low and back again serves to keep drive tension high which results in increased dependence on the environment for obtaining drugs and for whatever is left of coherent social relations, thus weakening relative autonomy from the environment.

Under such conditions, Zinberg notes, the ego could be expected to undergo a regressive process resulting in the 'typical impulse-ridden, psychopathic junkie'' who is the subject of most clinical studies.


Dr. Wieder's paper begins the series of papers in this monograph with a proposal that we proceed from an historical perspective. He takes us back to an early period at Lexington where Federal efforts were first made to unravel the enigma of addiction and drug dependence through collaborative efforts of behavioral and biological scientists and psychiatrists.

He recounts his experience as a member of these pioneering researchers, noting their difficulties even then in unifying a body of knowledge generated by researchers and clinicians who had yet to agree upon basic definitions of the phenomenon, the nature of the problem, or the constructs to be used in their study.

Dr. Wieder reminds us of the interdependency of theory and data in scientific investigation, and challenges the technical review group to learn from what has gone before them.

Before generating yet more voluminous data, he pleads for systematic evaluation of what has already been studied and learned. Wieder observes that despite the accumulation of massive, valuable data at Lexington, they were unable to develop a valid psychological definition or theory of addiction because of a restrictive medical model that was prevalent.

Psychoanalysis was similarly unequal to the challenge as a result of its own definitional problems, wherein addiction was too narrowly defined as 'an impulse problem'' and therefore not within the domain of psychoanalysis.

Because of exclusion of addicts from treatment based on inadequacies of earlier theory, plus the low socioeconomic background of most opiate addicts, few such patients were seen by analysts. Based on these trends, Wieder laments that the Lexington experience suffered from an absence of analysts, and that when analysts attempted to formulate an understanding of addiction, their formulation suffered because of an absence of comprehensive, objective clinical data.

Wieder believes that Rado made a good beginning and that his insight, 'not the toxic agent, but the impulse to use it,'' could have been an important cornerstone for a psychoanalytic theory of addiction. Wieder is skeptical that the application of any particular formulation by itself will greatly improve our treatment capabilities; however, he is firm in his conviction that psychoanalysis represents the most meaningful method of understanding human behavior and development, and applied to addiction, will make its greatest contributions in the areas of formulating and understanding the problem and in theory development.

The next four papers plunge directly into the rich complexity of psychoanalytic formulation. These papers exemplify the fact that psychoanalysis is not one theory, but many part theories, each addressing a different metapsychological level or a different explanatory perspective. This will be demanding reading for non-psychoanalytic readers, but well worth their while.

As Dr. Wieder notes in closing, reminding us of the mandate of our work and the promise that it brings to this diverse field, psychoanalytic theory is the 'most comprehensive view of human behavior'' currently available.

Dr. Wurmser begins by reminding us of the importance of appreciating the severe psychopathology in compulsive drug users. He cautions the reader against the impulse to seek simple or expedient solutions to an immensely complex problem.

In elaborating and expanding upon his own earlier observations (1974), Dr. Wurmser challenges us greatly to apply already complex concepts of narcissism, affect defense, compulsivity, splitting, psychological boundaries, and externalization to the analysis and understanding of the etiology of compulsive drug use.

However, in pursuing the challenge and complexity of Wurmser's thinking, new and enriching vistas of understanding as well as valuable hints as to where we should be looking further are opened up for us. His meticulously developed model for what he considers the direct and specific psychological antecedents of compulsive drug use is an object lesson in the rewards of patience and perseverance in the face of such complexity.

In this paper the knowledgeable clinician will find many occasions to nod his or her head in recognition as Dr. Wurmser unravels the threads of behavior and motivation, and puts them into a model which manages to allow for specific application within the framework of highly abstract concepts. And by reconsidering compulsive drug use in the light of more recent conceptual developments in psychoanalytic theory, Dr. Wurmser enriches and contemporizes both the drug field and the theories themselves. 

Dr. Greenspan in his paper focuses primarily on the developmental and adaptational perspective within a psychoanalytic model of learning. He summarizes the major developmental challenges of early life, starting with the earliest phases of achieving homeostasis and need satisfying attachment, through subsequent phases of separation, individuation and capacity for mental representation (as developed by Mahler and Piaget).

He stresses how optimal nurturante and encouragement from the environment (as primarily represented by the mother) fosters adequate mastery of these phases by the infant, and leads to the development of stable ego structures and capacities to manage drives and object relations.

To the extent that the individual is overly deprived or indulged in his/her development, varying degrees of ego impairment occur and drugs then come to substitute and compensate for the developmental defects and impairments.

Greenspan specifically delineates how the lack of integrated ego structures and differentiated drive organization in certain individuals leave them particularly susceptible to environmental reinforcers and influences.

In the case of addicts, drugs as well as many other environmental influences become powerful determinants of behavior in the absence of adequate ego structures and drive organization. Notwithstanding a difference in terminology, this formulation is consistent with and bears important resemblance to

Krystal's notions of developmental impairments in the stimulus barrier and the failure in differentiation of affects. It also parallels Wurmser's thinking, wherein he emphasizes the addict's need to externalize and to act in the absence of adequate ego defenses.

In the next paper, Krystal clarifies the defensive function of the 'splitting'' mechanism in drug-dependent individuals, and traces the enormous problems that addicts have with ambivalence, particularly aggression.

He discusses the nature of childhood trauma which can lead to undifferentiated or regressed affects and poor affect tolerance. Such trauma, in later life, may result in a massive 'walling off'' of self- and object-representations and an inability to provide comforting and self-care maternal modes for the self, or to tolerate aggression toward significant objects.

Krystal cogently describes the viscissitudes of these trends in the transference when long-term psychotherapy is undertaken with such patients. The work of therapy centers on helping the individual to overcome the fear of closeness with the therapist (i.e., vis-a-vis rekindled childhood longings and fears of aggressive impulses), learning to grieve effectively, owning up to one's destructive feelings, and finally, overcoming the barriers that prevent effective comforting and care of oneself.

Krystal suggests that success in this process invalidates the need for the placebo and pharmacologic action of the drug to effect access to these parts and functions of the self.

Dr. Khantzian's paper serves to provide some perspective on what has gone before, by reviewing and pulling together the major themes embodied in the foregoing papers and in other recent literature. He then explores aspects of ego function related to drive defense and 'self-care.''

He proposes a gap and/or impairment in the ego function of self-care. He relates this to failures in internalization of vital functions which have left the individual vulnerable to a whole range of hazardous behavior and involvement, but in particular, to addiction.

He then goes on to review how certain narcissistic processes and resultant defenses are related to unique characteristics and traits of addicts that impair such individuals in obtaining satisfaction in their involvement with people, work and play.

The next three papers offer a sampling of clinical case material and research which follows from the preceding formulations.

Dr. Davidson's paper brings a psychoanalytic understanding to an evaluation of methadone maintenance clinics as a treatment modality. She has provided valuable flesh and substance to our often made generalizations about addicts' prominent reactions of splitting, projection, impaired reality sense, other primitive defenses and narcissistic rage.

She accomplishes this through her compelling observations and descriptions of the very frequent and troublesome transference distortions that occur between patients and staff in methadone clinics. Her clarification that the often observed extreme, intense and labile outbursts by patients in methadone clinics have irrational and overdetermined origins in the patient's past (and are not simply a function of clinic setting or social background), is a helpful reminder to the most seasoned staff member, and may also act as a helpful guide to the novice staff worker in a methadone clinic.

Her paper should be required reading for all staff and administrators working in a methadone clinic. Kaplan presents an in-depth case report about a heroin addict and her family. The detailed clinical description of this woman highlights some of the severe narcissistic disturbances and pathological regression and fixation present in so many addicts.

The pathological disturbances of the parents are also well spelled out and give us clear understanding of this patient's developmental failures and troubled identification with her parents.

In the absence of mature defenses and an enabling ego (ideal), such patients have few choices but to adopt regressive oral satisfactions. Finally, Frosch and Milkman in their paper discuss their research findings which appear to support the observations of Wieder and Kaplan, Khantzian and others, that selection of specific 'drug of choice'' is determined by particular ego vulnerabilities, dispositions and drive strengths: The drugs are used by such individuals syntonically to either augment or bolster certain modes and styles of adaptation, or to compensate for certain ego deficiencies.

Frosch and Milkman's findings and their conclusions further lead them to concur with Wieder and Kaplan that the drugs are used to induce ego states reminiscent of a similar state in earlier developmental phases (along lines proposed by Mahler).

Whereas the heroin addict uses opiates to achieve a 'narcissistic blissful'' state of union dating back to a very early phase of development, the amphetamine user takes advantage of the stimulating action of the drug to bolster a grandiose sense of omnipotence through movement and activity akin to the 'practicing period'' at a slightly later period of development (around one and one-half years old).


In the following collection of papers the reader will be rewarded by the richness of insight which a psychodynamic understanding can bring to bear on the drug-problems field.

He/she will no doubt also encounter some of the difficulties inherent in an approach which attempts to do justice to human complexity and individuality. As a beginning, one need only note the plethora of terminologies and definitions to conclude, as Dr. Wieder points out in the opening paper, that there is an urgent need to integrate and clarify our definitions and knowledge to date before attempting to open further vistas.

Greenspan's paper, for instance, addresses the generic construct of 'substance abuse,'' while Khantzian and Davidson are writing specifically about heroin addicts; Krystal refers to 'other drug-dependencies'' but appears to center his observations around the problems of alcoholism.

This difference in focus leads to the apparently contradictory emphasis on the adaptive advantage of long-acting drugs by Wurmser and Khantzian on the one hand, and of short-acting drugs by Krystal and Raskin on the other.

Wieder specifically redefines drug addiction around a concept of compulsiveness in distinction to the traditional criterion of physical dependence; yet Frosch and Milkman in their experimental work based on theoretical conceptualizations similar to Wieder's, define drug dependency operationally in terms of specific amount and frequency of drug use.

1 As indicated in chapter 1, this paper was written before the second meeting which is reported by Woody in chapter 11. 

Admittedly, some of these difficulties are only semantic, and other apparent differences stem from preferences in focus.

But a close look suggests some basic areas which require fuller consideration before a unitary conceptualization of the problems and issues can be derived.

Diagnostic considerations constitute a case in point. Both Wieder and Kaplan consider drug use as symptomatic but not pathognomic, thus underscoring observed diagnostic heterogeneity of drug-using patients, even within a specific type such as narcotic addicts.