When Does Amphetamine-Induced Psychosis Become Schizophrenia?

This case describes a 36-year-old male with a 20-year history of psychiatric illness. The first decade of illness was characterized by antisocial behavior and heavy polysubstance abuse, primarily with amphetamines. The second decade was characterized by signs and symptoms prototypical for chronic schizophrenia, in the absence of ongoing drug abuse. The nature of the relationship between these two clinically discrete periods of psychiatric illness is discussed. Animal research on behavioral sensitization to stimulants and human studies of methamphetamine epidemics are cited as mounting evidence implicating a direct etiologic role of stimulants in the initiation and maintenance of a chronic schizophrenic disorder. Diagnostic implications are discussed.

The patient was a 36-year-old single white man. He lived with his parents in a small rural town in the same house in which he was raised. He was the youngest of six children, all of whom had passed the expected age of risk for schizophrenia onset and appeared to be leading healthy and productive lives. One would be hard-pressed to predict from childhood family photographs (1) that the patient would be the one who would become ill as an adult: he had a bright smile, was often pictured in the center of friends, and was often engaged in athletic activities (in which he was said to have excelled). His parents stated and medical records confirmed that there was nothing unusual about either the pregnancy or delivery, and there were no perinatal complications. The patient appeared to achieve all motor and verbal developmental milestones at the same pace as his siblings and peers. Early academic performance was in the average range, which was consistent with the performance of his siblings. No problems were noted in either his social adjustment or school performance until his sophomore year in high school.

At that point, he began to exhibit conduct problems: school truancy, excessive fighting with peers, and disobedience with house rules such as curfews. He became uncharacteristically irritable and unreliable. His grades fell progressively, and he finally dropped out in his junior year of high school after he had failed several courses (much to the dismay of his parents, since all of the other siblings had graduated from high school, and two had gone on to college).

His parents were convinced that the patient had “fallen in with a bad crowd,” and they suspected that he had become involved with drugs. Their fears proved to be correct. Although the patient had denied it at the time, he later admitted that beginning early in high school and for many years after he had used a variety of drugs, some in large amounts. His drugs of choice were amphetamines and marijuana, which were often used to counterbalance one another. He estimated that he had taken up to 100 “white crosses” (amphetamines) per week, often for several months running, for almost 10 years. He averaged one to three joints of marijuana per day throughout this time. He also with some regularity had used hallucinogens, both mescaline and LSD, and he estimated that he had “tripped” once or twice a month for at least a few years. Finally, he drank beer, sometimes up to a six-pack per day. This type of behavior persisted throughout his teenage years. By his early twenties he had also begun to use cocaine, although he still preferred amphetamines. He had never used heroin or other narcotics and had never intravenously administered drugs.

The manner by which he had supported himself, a live-in girlfriend, and his drug habit during these years remains questionable, and it is suspected that he had been involved in some small-time drug dealing. However, he had avoided problems with the law except for a few arrests for fighting, public intoxication, and driving while intoxicated. His work history was sporadic; he had worked part-time as a truck driver (which was alarming considering the subject’s substance abuse history). Finally, at the age of 26, after he had lost his driver’s license because of multiple charges of driving while intoxicated, he consented to enter a 1-month inpatient substance abuse treatment program at the state psychiatric hospital. During a month of abstinence, he did not manifest any clinical signs of withdrawal from any substance, which substantiated his claims that he had never experienced any withdrawal symptoms in his life. He was discharged on a regimen of disulfiram and was advised to participate in daily Alcoholics Anonymous meetings.

He complied with neither, but his drug and alcohol intake did decrease substantially. He separated from his girlfriend of several years and begrudgingly moved back in with his parents. Over the following month or two, he was noted to be irritable and withdrawn. His parents attributed this to the fact that along with giving up drugs, he was giving up a lifestyle and social circle. He would need a period of readjustment, they reasoned.

It was at that point (about 2 months after the patient had discontinued his drug and alcohol intake) that he began to express paranoid ideation for the first time. Most of the content related to his girlfriend, beginning with plausible claims that she was being unfaithful to him. The paranoid ideation soon evolved to include totally unfounded accusations of her having a sexual relationship with one of his brothers. Within a few weeks, the allegations had progressed to include more bizarre claims, such as the belief that his girlfriend was having sex with dogs and farm animals and that she had been setting fires all over town. His parents reported that he was increasingly preoccupied by these ideas and was ultimately unable to talk about anything else. He became reclusive and spent almost all of his time in his room.

His parents were afraid that he must be using drugs again, although he continued to deny it. They persuaded him to return to the state hospital for reevaluation. Instead of being admitted to the drug rehabilitation unit, however, he was admitted for the first time to a general adult psychiatric inpatient unit (he was 26 years old at this time). Because the delusions were the predominant clinical feature, occurred in the absence of delirium or dementia, and were judged to be etiologically related to a specific organic factor (in this case amphetamines), the patient was given an admission DSM-III diagnosis of organic delusional syndrome. (In DSM-IV terms, the diagnosis would be amphetamine-induced psychotic disorder, with delusions.) Results of a urine drug screen on admission revealed no presence of any drug, and the patient persistently denied having used any drugs over the previous few months. The intensity of his delusions did not diminish after he spent 1 week on a locked unit, during which time substances of abuse were presumably absent. Hospital records indicate that in light of these factors, the diagnostic impression of the patient’s illness was changing toward a schizophrenic disorder. The patient began treatment with haloperidol on the 10th day in the hospital. While his delusions remained intact over the 6-week hospitalization, he was less preoccupied and upset by them. He consistently denied any type of hallucinatory phenomena. The primary DSM-III discharge diagnosis was schizophreniform disorder, and he was instructed to continue taking moderate doses of haloperidol.

The parents noted a clear improvement in the patient during the first few weeks after his discharge, but he apparently soon became noncompliant with medications. He settled into an isolative lifestyle and continued to live with his parents. He spent most of his time alone in his room, although he occasionally went into town, where he would sit by himself at a restaurant and drink coffee and smoke cigarettes. His sleep cycle tended to be reversed, i.e., he typically stayed up most of the night and slept during most of the day. Despite prolonged periods of abstinence from drugs and alcohol, paranoid and referential ideation and reclusive behavior persisted, and his condition would intermittently worsen. However, it appeared that he had indeed given up most of his drug use. When offered any congratulations on his success in stopping drug use, he indicated that it was not really very hard and that with perhaps the exception of marijuana (which he was still using intermittently), he had “lost his taste” for most drugs.

Over the next few years, his course waxed and waned. He would have periods of increasing preoccupation with delusions and presumed hallucinations, during which his behavior would appear totally guided by these symptoms. He would seclude himself in his room for days at a time, with multiple locks on his bedroom door, and his parents reported that he would frequently talk aloud in his room. He slept with a gun under his bed and a knife by his side. Such periods would often last several weeks at a time. These would be intermixed with periods of relative calm, during which he seemed distant and empty but not irritable or agitated. During these calmer times, he also tended to stay in his room much of the time but did not lock his door or appear guarded or upset. When urged to get out of his room or the house more, he said that he would soon, but just did not feel up to it yet.

He was first evaluated by the primary author during one such quiescent period, which was some 4 years after his first psychiatric hospitalization. At the urging of the parents, who were seeking clarification regarding diagnosis, prognosis, and treatment recommendations, the patient agreed to enter our inpatient research unit. Throughout his 3-week hospitalization, he was passively cooperative but quite superficial and brief in his interactions with both peers and staff. He spent most of his time either alone in his room or on the smoking porch (he smoked between one and two packs of nonfiltered cigarettes daily). His affect was guarded, constricted, and blunted, and he consistently reported a neutral mood. His thought processes were markedly impoverished, he would rarely initiate a conversation or spontaneously elaborate, and his responses to questions were typically brief and concrete. While he initially downplayed any problems, he gradually revealed that he felt that he had “enemies” who had been “harassing” him. As he described this harassment it was clear that he had elaborate and long-standing delusions and hallucinatory phenomena. He was convinced that his enemies had been coming into his house at night and raping him while he slept. He said that this had been occurring off and on over the previous several months and probably longer and that this was the reason for the elaborate locks on his bedroom door and the weapons under his bed. While in the hospital, he slept more during the day than at night, which he admitted was because of his ongoing concerns about being sexually assaulted while sleeping at night. Although he continued to deny frank hallucinations, he ultimately described having “dog ears” and explained that this enabled him to hear conversations that were going on far in the distance. He also admitted to concerns that his audio equipment had been tampered with in such a way that “the wrong lyrics come out of the music.” Finally, he had entertained the idea that some sort of an electrical device had been placed in his body that allowed his enemies to hear everything he said, but he was not convinced this was true.

Family history of psychiatric illness was remarkable only for drug and alcohol abuse by one brother and one sister during their teenage years and into their twenties. Both have been abstinent for many years and are free of other psychiatric problems. Results of a physical and neurological examination of the patient, as well as routine admission laboratory tests, were all within normal limits. A magnetic resonance imaging scan of his brain revealed no gross structural abnormalities, and there were no suggestions of more subtle abnormalities, such as ventricular enlargement or sulcal prominence.

At the conclusion of this inpatient evaluation, it was strongly recommended that the patient try antipsychotic treatment once again. This time, he voluntarily agreed to take haloperidol. The dose was titrated toward a target haloperidol plasma level of 8 to 18 ng/ml (2), which required an oral dose of 20 mg/day. Within about 1 week at this dose, he began to exhibit clinically significant extrapyramidal signs, primarily parkinsonism and akathisia. These side effects improved with the addition of benztropine, 2 mg b.i.d., and propran-olol, 20 mg b.i.d. In terms of psychotic symptoms, the patient continued to express some vague paranoid ideation, but he seemed much less bothered by it, and his sleep cycle normalized. The DSM-III-R discharge diagnoses were chronic schizophrenia, undifferentiated type, and amphetamine dependence, which was in full remission.

Over the following 2 years, the patient was stable and appeared to be fully compliant with his oral medication regimens. There was no recurrence of prominent positive symptoms. When confronted with his previously held delusional beliefs, he offered no explanation as to why he believed them at the time, but he denied any ongoing concerns. Negative symptoms, however, remained profound. He spent most of his time at home, usually in his room. He would lie on his bed for hours on end but would not sleep, read, watch television, or listen to the radio. Yet, all the while he would report no subjective experience of boredom. He denied any ruminations or frequent thoughts or fantasies. He did, however, describe a sense of his mind being empty and his thoughts being slowed down, but neither was particularly bothersome to him. He was never subjectively depressed but was persis-tently anhedonic. He ate one meal a day, usually around midday, and described a sense of eating for sustenance rather than for pleasure. He was superficially pleasant with his parents and with anyone else with whom he came in contact, but he ran out of things to talk about after a few words were exchanged.

After this 2-year period of only residual symptoms, the haloperidol dose was cut from 20 to 10 mg/day because of concerns that this treatment might have been at least a partial contributor to the severity of the negative symptoms. Within about 2 weeks, the parents noted a clear change in the patient’s demeanor. He again appeared irritable and hypervigilant. He was sleeping and eating less and coming out of his room less frequently. He was brought in for an emergency evaluation and was instructed to resume the 20 mg dose. However, by that time, he was already quite suspicious and expressed a lot of hesitation about continuing with medication at all. Over the next week or two he was probably completely noncompliant with the medication regimen, and his paranoid ideation increased progressively to the point that he was once again locking his room, sleeping with a gun, and muttering angrily to himself throughout his waking hours. He refused to come to the hospital and had to be committed by his parents and brought in by the sheriff. He was grossly delusional on admission. He had incorporated his parents, the sheriff, and the treatment team into his delusional system and adamantly refused medication. He required placement in a seclusion room and was medicated against his will with parenteral haloperidol during the first hospital day, after which he accepted oral medications. Within a few days he began to show signs of improvement, and within 2 weeks he was essentially back to his postmorbid baseline, i.e., passively pleasant and cooperative but markedly alogic and avolitional. For the subsequent 3 years, he has remained in this condition on a regimen of oral haloperidol, 20 mg/day.

This case reflects a common clinical scenario: a presentation and course that are prototypical for schizophrenia, both in terms of positive and negative symptoms, that arise subsequent to severe polysubstance abuse, particularly with amphetamines. Such cases raise important questions regarding the role of stimulants in the etiology of a chronic psychotic syndrome and the impact of these etiological considerations on the diagnosis. Can chronic amphetamine abuse in and of itself cause a syndrome that is indistinguishable from schizophrenia, not only cross-sectionally during the intoxication phase but also over time, even after long periods of abstinence? Or does amphetamine abuse merely precipitate the onset of schizophrenia among vulnerable individuals? In genetic terms, can chronic amphetamine abuse itself result in a phenocopy of schizophrenia, or does it act to increase the penetrance of an underlying vulnerability to schizophrenia? In either case, if an important etiologic role for amphetamines in the development of a schizophrenic syndrome can be established, how should this affect diagnosis?

While the wording of “organic exclusion criteria” for the diagnosis of schizophrenia varies somewhat across most standardized criteria sets (e.g., DSM-III, DSM-III-R, and DSM-IV, as well as ICD-10), the intent has been consistent: schizophrenia is not to be diagnosed in the presence of an identifiable factor that can be etiologically implicated in both the initiation and maintenance of the psychosis.

In terms of initiating psychosis, it has long been recognized that the use of amphetamines or other central dopamine agonists can mimic the active symptoms of schizophrenia during the acute intoxication phase, even among individuals without a genetic diathesis for schizophrenia (3, 4). Indeed, investigations in both animals and humans of the acute behavioral effects of these drugs have been a cornerstone of the dopamine hypothesis of schizophrenia (5). Thus, there is little question that amphetamines can initiate a psychosis that is often clinically indistinguishable cross-sectionally from a schizophrenic disorder (especially the paranoid subtype). The differential diagnosis between substance-induced psychosis and schizophrenia therefore depends upon the degree to which the substance use can be implicated in the maintenance of a chronic disorder. In the clinical setting, this ends up turning upon whether the psychotic disorder persists for a designated period of time in the absence of ongoing substance use (e.g., over 1 month according to DSM-IV). While this may be the only way to operationalize our nosology, it is probably an oversimplification, especially as it applies to amphetamine-like psychostimulants.

There is a large body of preclinical literature (reviewed by Robinson and Becker [6] and Kalivas and Stewart [7]) that has demonstrated that certain behavioral effects of intermittent amphetamine administration in animals may persist long after its discontinuation. Further, these studies have demonstrated that responses to a variety of both pharmacological and nonpharmacological stimuli can also be chronically altered by pre-exposure to amphetamines and related compounds (8). The phenomena of “behavioral sensitization” and “cross-sensitization” are invoked to explain these findings. Sensitization to a drug (or any stimulus) can be thought of as the opposite of habituation or tolerance: the more a drug is used, the lower the dose that is required to produce the same or greater effects (9). This phenomenon is ubiquitous across species and demonstrable after a variety of diverse stimuli, but amphetamines have been the focus of more studies than any other stimuli to date. Animals given low, nontoxic doses of amphetamines that were administered intermittently over a period of time were found to have exaggerated behavioral responses when the same or lower doses were subsequently administered, even after long periods of abstinence. Behavioral sensitization to central nervous system stimulants was recognized in animal studies as early as the 1930s (10) and in humans by the 1960s (11), but only over the past two decades have the cellular mechanisms and neural substrates of this phenomenon been elucidated in animal models. The process appears to involve augmentation of axonal dopamine release that results in (or combines with) enhanced dopamine transmission (6). Research has indicated that the initial phase of the sensitization phenomenon involves a psychostimulant effect on dopamine cell bodies in the ventral tegmental area, while the enduring effects involve enhancement of both pre- and postsynaptic dopamine transmission in the nucleus accumbens or striatum (7, 12). The role of nondopaminergic systems in modulating these effects is also being actively explored (13).

It is of interest that there is abundant evidence for a cross-sensitization phenomenon with amphetamines as well. That is, not only are amphetamine-exposed animals subsequently more sensitive to amphetamine effects, but these animals demonstrate exaggerated behavioral responses to other drugs and nonpharmacological stressors (e.g., food deprivation, foot shock) as well (14). Thus, there are animal models that suggest that the intermittent use of amphetamines can have long-lasting behavioral and physiological consequences that include a vulnerability to recurrences of amphetamine-like effects in the context of a wide variety of stressors.

Lieberman et al. (15) and Sato et al. (16) have discussed variations on similar models that reflected potential mechanisms by which amphetamine abuse may lead to the development of a chronic schizophrenic syndrome in humans. They have suggested that the initial aggravating event is in the form of enhanced presynaptic dopamine activity. This enhanced activity may be the result of exogenous dopamine, as in the case of chronic use of drugs with dopamine agonist effects, or from some endogenous factor (e.g., excessive dopamine production or deficient breakdown, presumably determined largely by genetic factors). Behaviorally, this would be manifested as positive psychotic symptoms and would cause further regulatory disruptions at the cellular level for dopaminergic neurons both pre- and postsynaptically. Continued exposure of sufficient amount and duration is then posited to result in presynaptic neuronal degeneration, which ultimately leads to a relatively depleted dopaminergic state. Behaviorally, this may be manifested as negative symptoms. The depleted dopaminergic state should then result in an increase in postsynaptic neuronal sensitivity, which would be manifested behaviorally as a propensity for rapid and extreme reemergence of positive symptoms during conditions of transiently increased dopaminergic availability. Such a model is partly supported by challenge studies with dopamine agonists (particularly methylphenidate), which have demonstrated that relatively low doses of these drugs transiently exacerbate positive symptoms among some stabilized patients with schizophrenia (17), but not among normal control subjects (18) or patients with other psychiatric disorders (19). These studies further show that a psychotogenic response to such a pharmacological challenge may be an important indicator of the need for ongoing maintenance with dopamine antagonist medications (i.e., typical neuroleptics) (20). Thus, a subgroup of patients with chronic psychotic syndromes who have been diagnosed as schizophrenic appear to be highly sensitive to fluctuations in central dopaminergic availability. A history of amphetamine abuse may be one risk factor for such a sensitivity. Probably the best information on the natural history of amphetamine abuse in humans can be gleaned from the methamphetamine epidemics in Japan. Shortly after World War II, huge military stores of methamphetamine were dumped onto the open market in Japan, which led to large-scale methamphetamine abuse, usually in intravenous form. This epidemic was abruptly curtailed by the government in the 1950s, and methamphetamine was scarce until a resurgence of use in the 1970s. Since that time, abuse has been steadily increasing and has once again reached epidemic proportions. Thus, clear “on and off points” are easily identified. Further, other drugs of abuse have been extremely scarce in Japan during both of these periods, making these epidemics quite “clean” (as opposed to stimulant use in this country, which typically occurs in the context of polysubstance use). Studies of these epidemics have shown that in a substantial minority of patients, apparent methamphetamine-induced psychoses do not resolve within the first month; many go on to a chronic schizophrenic-like state. For example, Tatetsu et al. (21) found that almost 15% of their postwar methamphetamine-psychotic patients took 5 years or more to recover in the absence of ongoing use. Similar figures were obtained in approximately a dozen other studies of the Japanese epidemics (reviewed by Sato et al. [16]). The chronic psychoses of the patients were described as clinically indistinguishable from paranoid schizophrenia in terms of cross-sectional pre-sentation as well as longitudinal course, which was characterized by vulnerability to psychotic exacerbation interspersed with negative symptoms.

The patient presented in this clinical case conference exemplifies a number of features that are consistent with a chronic, amphetamine-induced schizophrenic syndrome. First, psychotic symptoms initially developed about 2 months after the cessation of amphetamine use, which is roughly the time interval that one would expect for up- or down-regulation of neurotransmitters to occur. Second, the patient displayed a rapid response to neuroleptic medication and a florid exacerbation of positive symptoms in response to a decreased dose of these dopamine-blocking agents. Third, the patient did not appear to have a diathesis for schizophrenia: family history was clearly negative for psychotic disorders, and premorbid functioning was good, both of which suggest an important etiological role for a later event. Finally, negative symptoms were prominent interepisodically.

Is there enough evidence in this case to implicate amphetamines in both the initiation and maintenance of the chronic disorder? If so, should the diagnosis be schizophrenia? We have attempted to demonstrate many of the reasons that amphet-amines can be so implicated. However, we are faced with the observation that the majority of amphet-amine abusers do not go on to develop a chronic schizophrenic-like syndrome. Until the factors that mediate the vulnerability to stressors such as amphetamines are better understood, it remains appropriate to use the diagnosis of schizophrenia to categorize such patients. However, cases such as this serve as an ongoing reminder of the almost certain etiologic heterogeneity of that diagnosis.