Skip to Content

Self Medication Hypothesis, ADHD & Cocaine

A Literature Review      
By Daniela Plume B.A


Historically, drug and alcohol dependencies have been attributed to sin, disease, maladaptive, self destructive, and antisocial behavioral patterns. Addicted substance abusers have frequently been charged with poor motivation, lack of insight, escapism, and the development of self destructive tendencies.

Khantzian (1985a) asserts these allegations stem from archaic models of therapy and rigid attitudes as to which paradigm best justifies substance abuse and subsequent treatment modalities.

Many clinicians still hold that psychotropic medication for coexisting pathology should not be administered until the patient has attained abstinence for at least one year. However, coexisting disorders should be addressed concurrently, not ignored until abstinence has been consolidated (Weiss & Colins, 1992; Zweben & Smith, 1989). 

Various studies have indicated a high degree of coexisting psychopathology amongst both adolescent and adult substance abusers. The most frequently seen include affective disorders (Dilsaver, 1987; Gawin & Ellinwood, Jr., 1988; Weiss, Mirin, Michael, & Sollogub, 1986), borderline personality disorder (Blume, 1989; Bukstein et al., Brent & Kaminer, 1989), antisocial personality disorder (Kleinman, Miller, Millman, Woody, Todd, Kemp, & Lipton, 1990), depression (Dorus & Senay, 1984; Kleinman et al., 1990; Lemere & Smith, 1990), anxiety disorders (Bukstein, 1989; Weiss & Rosenburg, 1985), conduct disorder (Dimilio, 1989; Milin, Halikas, Meller, & Morse, 1991), and attention-deficit hyperactivity disorder (Carroll & Rounsaville, 1993; Dimilio, 1989; Milin et al., 1991; Wilens, Biederman, Mick, & Faraone, 1995). 

Given the prevalence of comorbid psychopathy in substance abusers observed in both clinical and treatment settings, Khantzian (1985b) postulated a theoretical model of self medication; a concept first raised by Freud (1884) upon noting anti-depressant properties of cocaine.

The Self Medication Hypothesis: A Brief Overview

On the basis of psychodynamic/psychiatric diagnostic findings and clinical observations, Khantzian (1985b) proposed a model of self medication as an etiological factor in substance abuse.

He suggested psychotropic drug effects interact with psychiatric disturbances and "painful affect states" to predispose some individuals to addictive disorders. The addict's choice of drug is thought to be the result of the interaction between the psychopharmacologic properties of the drug and the "primary feeling states" experienced.

In this way, the drug effect is thought to substitute for defective or non-existent ego mechanisms of defense (Khantzian, 1985b). A number of clinical findings have supported the hypothesis that the preference for a specific drug is not random, but rather, appears to be a process of "self selection" (Dorus & Senay, 1980; Khantzian & Treece, 1985; Rounsaville, Weissman, Crits-Cristoph, Wilbur, & Kleber, 1982; Weissman, Slobetz, & Prusoff, 1976; Wurmser, 1974; and others).

This course of self selection has also been referred to as "preferential drug use" (Milkman & Frosch, 1973) and "the drug of choice" phenomenon (Wieder & Kaplan, 1969). 

Cocaine Abuse: The Self Medication Hypothesis

Cocaine is thought to help overcome fatigue and alleviate depression in some depressed individuals (Freud, 1884; Schnoll, Daghestani, & Hansen, 1984; Khantzian, 1975), increase feelings of self esteem, assertiveness, and frustration tolerance (Weider & Kaplan, 1969), overcome boredom and emptiness (Wurmser, 1974), and alleviate impulsive/hyperactive states in attention-deficit disordered individuals (Weiss, & Mirin, 1986; Zweben & Smith, 1989, and others).

Cyclical mood disorders (manic depressive illness, cyclothymic disorders) have been shown to be more common in cocaine abusers than opiate addicts (20 % vs 1%), suggesting such persons may preferentially select stimulants over other illicit substances of abuse (Gawin & Ellinwood, Jr., 1988). Cocaine is a CNS stimulant with pharmacological properties similar to the stimulant medications Ritalin?, Cylert?, and Dexedrine? that are commonly used to treat attention-deficit hyperactivity disorder.

Thus, it is thought that individuals with untreated ADHD may be using cocaine to "self medicate" these disease symptoms (Hallowell & Ratey, 1994; Milin, 1995; Weiss, et al., 1986).

Sub-Types of Cocaine Abusers:

On the basis of clinical observations, Khantzian (1984; Khantzian & Khantzian, 1984) proposed four categories or "subtypes" to explain how psychiatric/psychological factors might predispose an individual to become and remain dependent on cocaine. These include pre-existent chronic depression (dysthymic disorder), cocaine abstinence depression, hyperactive/restless/emotional lability syndromes, or attention-deficit disorder, and cyclothymic or bipolar illness.

Other researchers have also speculated that individuals with chronic depression may value the euphorigenic effects of cocaine, whereas cyclothymic and bipolar disordered patients may use cocaine to maintain a hypomanic state and fend off depression (Lemere & Smith, 1990).

Interestingly, individuals with ADHD frequently report a paradoxically placid response to cocaine as well as temporary relief from hyperactive symptoms(Cocores, Davies, Mueller, & Gold, 1987; Gawin & Kleber, 1986; Hallowell & Ratey, 1994; Khantzian, 1984; Weiss & Mirin, 1986; Zweben & Smith 1989). 

Psychiatric disorders, particularly the affective disorders, are believed to increase susceptibility to stimulant abuse (Gawin & Kleber, 1986). In two unrelated studies, Weiss, et al., (1986) and Gawin and Kleber (1986), each reported 50% of inpatient cocaine abusers to have met the diagnostic criteria for mood disorders.

A number of researchers have also noted patients with attention deficit disorders to be over-represented among those undergoing treatment for cocaine abuse(Cavanagh, Clifford, & Gregory., 1989; Khantzian, 1985b; Milin et al., 1991). 

Attention Deficit Disorder (ADHD) & Cocaine Abuse:

Attention-Deficit Hyperactivity Disorder: (ADHD)


Incidences of ADHD in adults have not been recognized until recently. In 1978, Leopold Bellack chaired a conference which focused on adult forms of ADHD, known then as "minimal brain dysfunction". It would take more than 10 years before the clinical significance of ADHD in adulthood would be established (Hallowell & Ratey, 1994).

It had previously been assumed that children outgrew ADHD symptoms by adolescence. Recent studies, however, have reported that upwards to 50-60% of children with ADHD continue to experience residual or full blown manifestations of this disorder in later life (Biederman, Faraone, Spencer, Wilens, Norman, et al., 1993; Gittleman, Mannuzza, Shenker, & Bonagur, 1985; Weiss & Hechtman, 1986).

In one such study, over 70% of children with ADHD were found to have met criteria for the disorder in adolescence (Barkley et al., 1990). ADHD has also been shown to be more prevalent in males than females by ratios ranging from 2:1 for general populations, to 9:1 for clinics (Biederman et al., 1993; Kaplan & Sadock, 1991).

ADHD is seen across a wide range of cultures, although the incidence rates vary. This is thought to reflect differing diagnostic practices (American Psychiatric Association, 1994). There is also some evidence that ADHD occurs more frequently in lower socioeconomic groups (Biederman et al., 1993). 

Diagnostic Criteria:

In accordance with the DSM-IV diagnostic criteria (APA, 1994), behavioral manifestations of ADHD must be seen before the age of 7 and must clearly interfere with social and academic functioning (and in later life, occupational functioning).

Previously, ADHD was considered to be a single disorder comprised of three main components: short attention span, impulsivity, and hyperactivity. A diagnosis of ADHD required meeting 8 of 14 criteria, such as, fails to listen, interrupts frequently, and fidgets or moves excessively (Kaplan & Sadock, 1991).

The terms "minimal brain dysfunction", "learning disabilities", "hyperkinetic syndrome", and "hyperactivity" have all been used to reflect this pattern of overactivity, short attention span, and learning problems (Ralph & Barr, 1989).

The symptoms described in the DSM-III-R represented behaviors believed by many to be common to children in general, thus, the DSM-IV has returned to an earlier subgrouping system for ADHD diagnosis; separating those children with both attentional deficits and hyperactivity (Davison & Neale, 1994 p.429). 

The DSM-IV presently recognizes three distinct classifications of ADHD, with a clear delineation between ADHD combined type (requiring 6 of 9 criterion to be met in both the inattention and the hyperactivity/impulsivity categories), ADHD-Predominantly Inattentive Type (which requires 6 of 9 criterion to be met in the inattentive category and less than 6 for hyperactivity), and ADHD-Predominantly Hyperactive Type (which requires 6 of 9 criterion to be met for the hyperactivity category and less than 6 for inattention). As a requisite to diagnosis for all three of these categories, behavior must persist for at least 6 months (APA, 1994). 

Genetic Factors:

ADHD appears to have a strong genetic component and is seen more frequently in first degree biological relatives (APA, 1994; Kaplan & Sadock, 1991). Goodman and Stevenson (1989) found concordance for clinically diagnosed hyperactivity in 51% of identical twins and 33% of fraternal twins. Adoption studies have also shown strong support for a genetic constituent (Morrison & Stewart, 1973; Wender, Reimherr, &, Wood, 1981). 
Back To: Table of Contents

Diathesis-Stress Theory & Outcome Studies:

Weiss, Minde, and Werry (1971) conducted a 5 year prospective follow-up study of 91 subjects aged 10-18 years. They found adolescents with ADHD tended to have lower self esteem, and most continued to be distractible, impulsive, and emotionally immature when compared with controls matched for age, sex, IQ, and social class.

The results of this study suggest three main "outcomes" of childhood ADHD: individuals with ADHD who function normally in adulthood, individuals who as adults continue to have problems with concentration, irrationality, anxiety, and who experience general difficulties in work and personal life (most fall into this group), and those who develop serious psychiatric and/or antisocial pathology, and may experience extreme depression, suicidal tendencies, become heavily involved with drugs and/or alcohol, and exhibit antisocial behavior.

Outcome studies may aid in explaining why many individuals with ADHD are successful in later adult life and conversely, why all individuals with ADHD do not develop substance abuse problems. Research has strongly suggested that children with both attentional deficits and hyperactivity (combined type) are most at risk to develop conduct problems, oppositional behavior, and other severe problems in later life (Barkely, DuPaul, & McMurray, 1990). 

Bettleheim (1973) proposed a diathesis-stress theory to explain the development of ADHD; suggesting that certain critical factors impinge on a child's life, which may in turn become a catalyst for the development of ADHD in those genetically predisposed. Studies employing multivariate stepwise regression techniques have identified a number of factors thought to predict adult outcome of children with ADHD (Hechtman, 1991; Hechtman, Weiss, Perlman, & Amsel, 1984).

These potential predictors include: factors specific to the individual child (health, temperament, intelligence, and psychological factors), characteristics of the family (socioeconomic status, emotional/psychological, and family composition), and the larger social environment. All three of these areas have been shown to contribute significantly to a child's resiliency/vulnerability. 

Children with fewer health problems either during pregnancy, perinatelly, or during infancy, are shown to be less likely to develop ADHD (Hechtman et al., 1984). Individual characteristics of the child may also influence outcome, as IQ and temperament contribute to the development of quality relationships with others. Bettleheim (1973) posited temperamental differences between child and parents as one possible stressor which may promote the development of the disorder in a predisposed child.

A hyperactive child may also elicit negative reactions from his/her parents and in turn become more disruptive. Weiss et al., (1971) found children with higher IQs and lower scores of hyperactivity to be more adaptable, socially responsive, and able to elicit positive responses from their environment. Having an internal locus of control, a good sense of autonomy, and positive self esteem have also been shown to contribute to resiliency.

Generally, the better the ego strengths, the less likelihood of developing ADHD (Ralph & Barr, 1989). Khantzian (1984, 1985a) emphasized the importance of ego development as a major contributor in predisposing individuals to self medicating for uncomfortable or painful feeling states. 

Werner and Smith (1982) found resilient children to come from homes that were more cohesive and supportive, with more structure, regularity, supervision, and clearly defined rules, as well as realistic expectations of the child. Socioeconomic status appears to be another strong predictor of ADHD in adolescents (Loney, Kramer, & Milich, 1981). Higher family social status enables greater physical, social, and educational benefits; whereas lower status may place undue stress on both the child and the family.

Finally, the larger social and physical environment can provide a beneficial extra-familial support system through school and church as extended family (Werner & Smith, 1982; Rutter, 1979). In one long term prospective follow-up study of young adults with ADHD, when asked what had been most beneficial to them while growing up, the most common answer was having someone in their lives who believed in them (Weiss & Hechtman, 1986). 

Comorbid Disorders:

ADHD is usually characterized by impulsivity, lack of emotional control, attentional deficits, and learning disabilities, however, there may be no single critical attribute of ADHD (Wender, 1979). In fact, many researchers have identified the most frequently seen characteristics of individuals with ADHD to be irritability, emotional lability, explosive personality, violent dyscontrol, depression, low self esteem, anxiety, and aggression (Hallowell & Ratey, 1994; Kaplan & Sadock, 1991; Ralph & Barr, 1989; Turnquist, Frances, Rosenfeld, & Mobrak, 1983; Wender et al., 1981).

In one study, depression was seen to be the most common symptom associated with ADHD (Heussy, Cohen, Blair, & Rood, 1979), while Weiss and Mirin (1986) identified frequent occurrences of borderline personality and antisocial personality disorders. 

ADHD in childhood is associated with an increased frequency of psychopathology in later life (Wender, et al., 1981). Adults are seen to exhibit the same patterns as children with respect to psychiatric and cognitive features, as well as psychiatric comorbidity. In childhood, ADHD frequently occurs with conduct disorder, antisocial personality disorder, oppositional defiance disorder, and Tourette's syndrome (APA, 1994).

Researchers have consistently found higher rates of antisocial personality, conduct, oppositional defiant, substance use, and anxiety disorders in adults with ADHD when compared to non-ADHD adults (Biederman et al., 1993; Carroll & Rounsaville, 1993). In one study, Gittleman et al., (1985) found conduct disorders in 48% of adolescents with ADHD, in 13% of adolescents who had outgrown ADHD, and in only 8% of controls without the disorder.

In another study, Hinshaw (1987) reported a 30%-90% overlap between ADHD and conduct disorder. Incidences of major depression and anxiety disorders in childhood (which often persist into adulthood) have been documented as well (Hechtman et al., 1984). 

Adoption studies have indicated genetic origins associated with an increased risk of substance use, antisocial personality, and somatoform disorders in later adult life (APA, 1994; Morrison & Stewart, 1973; Cantwell, 1975). Higher rates of affective disorders have been noted in first degree relatives of cocaine abusers (Weiss & Mirin, 1986).

Some studies have observed as much as 25% of children with family pathology to have significantly higher ratings of antisocial and aggressive behaviors (Hinshaw, 1987; Weiss, 1986). The existence of psychopathology in the family of origin then, appears to be a significant risk factor for substance abuse. 

A significant number of attentional disordered individuals have shown serious delinquent and psychological outcomes, and have been shown to be at risk for chemical dependence (Clopton, Weddige, Contreras, Fliszar & Arrendondo, 1993; DeMilio, 1989, Gittelman, Mannuzza, Shenker, & Bonagur, 1985; Milin et al., 1991; Wilens, et al., 1995).

One study reported a lifetime prevalence of between 15% - 18% for the substance use disorders, making them the "most common mental disorders in the general public, especially amongst males" (Robins, Helzer, Weissman, Orvaschel, Gruenberg, Burke, & Reiger, 1984). Finally, Milin et al., (1991) noted the severity of substance abusing behavior to be greater in the presence of a coexistent psychiatric disorder. 

Hechtman, Weiss, and Perlman (1984) compared a clinical group of 75 subjects (male and female) who had been diagnosed in childhood as hyperactive, with 44 matched controls in a ten year prospective follow-up study. They found a tendency for adolescents with ADHD to have greater drug use (75% vs 5%), and were more likely to have experienced a period of dependency or abuse during the five years preceding evaluation.

This difference was seen to level out over the year following the study, perhaps indicating the attainment of similar levels of moral development. Gittleman and colleagues (1985) studied 101 adolescent males aged 16-23 years. They found substance abuse disorders in 28% of patients with ADHD, 8% of ADHD children who no longer showed symptoms in adolescence, and in only 3% of controls who had never exhibited ADHD symptoms.

Cocaine Abuse:

A Brief History and Epidemiology:

In the 1890's cocaine was considered safe. Use escalated but then abated as serious problems were noted. This pattern was repeated in the 1920's, early 1950's, and again in the late 1960's. Believing cocaine to be non-addictive, millions of people tried it and abuse exploded. In fact, the Diagnostic and Statistical Manual of Mental Disorders did not recognize cocaine as an addictive substance until the DSM-III-R was released in 1980 (Kaplan & Sadock, 1992).

In 1974 it was estimated that 5.4 million Americans had tried cocaine; in 1982 this figured had risen to 21.6 million. By 1985, the National Institute on Drug Abuse estimated 5.8 million Americans abused cocaine regularly. This figure dropped in 1988, to an estimated 2.9 million abusers and a reported 1.6 million in 1990, with males out-numbering female users 2:1 (Kaplan & Sadock, 1992).

ADHD in Cocaine Abusing Populations:

The relationship between cocaine dependence and attention-deficit hyperactivity disorder was first considered by Khantzian (1979) to be the possible extension or "augmentation" of a hyperactive, restless lifestyle by a select group of cocaine users. In recent years, ADHD has been frequently reported in cocaine abusing populations (Cavanagh, et al., 1989; Gawin, Riordan, & Kleber, 1985; Khantzian, 1983; Rounsaville et al., 1982; Weiss, Pope, & Mirin, 1985).

Carroll and Rounsaville (1993) found 103 of 298 (35%) treatment seeking cocaine addicts to have met the DSM-III-R criteria for ADHD. In an assessment of 111 juvenile delinquents aged 11-17 years, Milin and colleagues (1991) found attention-deficit disorder with hyperactivity in 23% of the substance abusers and in no cases of the non-substance abusing sample; with 50% of the adolescents with ADHD indicating a preference for cocaine.

Characteristics of Substance Abusers with ADHD:

Cocaine abusers with ADHD tend to be younger at the time of first treatment, and report more severe and frequent substance use, earlier onset of cocaine abuse, and more previous treatment attempts (Carroll & Rounsaville, 1993). In one study, hyperactive adolescents were seen to be significantly younger than controls when they started, at the point of heaviest use, and when they stopped using cocaine (Hechtman, Weiss, & Perlman, 1984).

Incidences of ADHD within treatment settings have also been observed to be greater for male than female patients (Everett, Schaffer, & Parsons, 1988). Carroll and Rounsaville (1993) reported 78% male vs 23% female attentional disordered treatment seeking cocaine abusers in one treatment study, and a similar ratio, 73% male vs 27% female was reported by Gawin and Kleber (1985a).

In work with chemically dependent adolescents, Ralph and Barr (1989) identified "explosive volatility" as a feature of ADHD behavior not usually included in the clinical description. Individuals with ADHD in substance abuse treatment settings are often seen to be defiant, argumentative, verbally aggressive, and often verging on premature discharge from treatment facilities.

This apparent escalation of negative behaviors in patients with ADHD, is often attributed to their having limited ego skills and resources to cope with life stressors, as well as the additional stress placed on them by a highly structured inpatient treatment setting (Ralph & Barr, 1989).

In studies comparing adolescents treated for ADHD with stimulant medications and adolescents without ADHD, treatment for ADHD was seen to decrease the risk for future adult drug and alcohol use (Beck, Langford, MacKay, & Sum, 1975; Loney, Kramer, & Milich, 1981; Henker, Whalen, Bugental, & Barker, 1981).

Adolescents appropriately treated for ADHD showed similar, and in some cases, less incidences of substance abuse than controls. Fewer studies comparing treated vs untreated individuals with ADHD have been conducted. In one such study however, Kramer, Loney, & Whaley-Klahn (1981) found untreated hyperactive boys tended towards greater drug use than those properly treated for ADHD.

Diagnostic Issues:

Retrospective Diagnoses & Confounding Variables:

The diagnostic criteria for ADHD in adults requires a history of childhood ADHD; therefore, one major problem in determining the incidence of ADHD in adults is retrospective diagnosis. Ward, Wender, and Reimherr (1993) recently constructed the Wender-Utah rating scale (WURS) "in an attempt to surmount this problem of retrospectively establishing the childhood diagnosis of ADHD in adults."

A "cutoff" score of 36 or higher on the 61 item rating scale has been shown to accurately distinguish 96% of individuals with ADHD from controls. The Utah criteria for ADHD includes items of impulsivity, over-excitability, temper outbursts, affective lability, stress intolerance, and disorganization. Wender's diagnostic criteria for adults with ADHD requires: a childhood history of attention deficits and hyperactivity with one of the following: problems in school, over-excitability, and temper outbursts, or an adult history of attention deficits and hyperactivity together with two of the following: affective lability, explosive temper, stress intolerance, disorganization, and impulsivity.

Individuals meeting other diagnoses such as schizophrenia, depressions, and borderline personality disorders, were excluded from test development studies. 

ADHD, Antisocial Personality Disorder, & Conduct Disorder:

The Confounding Triad:

The DSM does not purport closed or fixed categories. Indeed, ADHD itself does not appear to be a mutually exclusive category and has been seen to overlap significantly with oppositional defiant and conduct disorders (Demilio, 1989; Loney, 1988; Milin, et al., 1991; Ralph & Barr, 1989). Ward et al., (1993) found several of the borderline personality disorder symptoms (affective lability, volatile temper, and impulsivity) to overlap with ADHD as well.

It would appear that although the DSM and the Wender-Utah rating scale overlap, they may not necessarily target the same behaviors. For instance, the DSM does not acknowledge emotional lability and volatile temper as components of ADHD, although many research studies have reported these characteristics (Morrison, et al., 1973). As well, the Utah criteria does not recognize ADHD without hyperactivity in its diagnosis. 

There is considerable disagreement as to what constitutes the different diagnostic categories. Some researchers suggest hyperactivity and aggression are separate independent diagnostic categories (Halikas, Meller, Morse, & Lyttle, 1990; Loney, 1988); some feel they are intertwined (Faraone, Biederman, Keenan, & Tsuang, 1991; Printz, Connor, & Wilson, 1981), and others feel they are essentially the same thing (Quay, 1979). 

There has also been considerable disagreement as to which of the disorders is more likely to induce or contribute to later substance abuse problems. In substance abusing samples, ADHD was found in conjunction with conduct disorder and antisocial personality disorder (Carroll & Rounsaville, 1993; Gittleman et al., 1985). Some researchers maintain that aggression or sociopathy, and not ADHD, is related to substance abuse (Halikas et al., 1990).

However, Carroll and Rounsaville (1991) found a high incidence of ADHD in cocaine abusers that was not accounted for by sociopathy. They also found notably more intense and earlier onset of cocaine abuse, irrespective of comorbidity with antisocial personality disorder.

Finally, in a recent unpublished study, Wilens, Biederman, Mick, and Faraone (1995) found ADHD by itself significantly increased the risk for substance use disorders in adults, and even more so when compounded with antisocial, mood, and anxiety disorders.