Amphetamine psychosis: a model for studying the onset and course of psychosis

Neuroscience research within the psychiatric field has tended to focus on detection of “unique” genetic factors and related “endophenotypes” (ie, intermediate phenotypes, assumed to have simpler genetic underpinnings than the disorder itself) in an attempt to identify potential aetiological mechanisms.1 In contrast, limited research has examined exposure to known environmental pathogens as a method of elucidating underlying neurobiological processes. Researchers in the United Kingdom have detailed a “bootstrapping” approach that starts with a known environmental pathogen and links it to specific illness states and possible at-risk genotypes.1 The goal is to achieve convergence, such that the environmental and genetic evidence is directly linked to the same underlying neural substrate.1 They stress that knowing all three factors, rather than the two that dominate most linear models, enhances the likelihood of elucidating the actual underlying pathophysiology.1

Here, we explore the capacity for amphetamine-type stimulants (ATSs) to function as a candidate environmental pathogen linked to schizophrenia and related psychoses. The potential to unravel relevant genotypes or causal neural substrates in cohorts of people exposed to ATSs is evaluated. ATS exposure is particularly attractive to researchers, as it is a common environmental pathogen. Further, the neurochemistry of acute intoxication is well understood. Neurophysiological and neuropsychological markers are available for schizophrenia and related psychoses. Some of these appear to be viable endophenotypes in association studies.2,3 If one accepts the basic proposition that ATS-related psychoses are biological equivalents of schizophrenia and related psychoses, this is an experimental model worthy of closer neurobiological and longitudinal examination (Box 1).

ATS use increases the likelihood of experiencing psychotic symptoms,4,5 and the clinical characteristics of ATS-related psychosis can be indistinguishable from acute paranoid schizophrenia (Box 2).4 In this article, we use the term “ATS-related psychosis” to refer to the presence or history of psychotic symptoms associated with ATS use (ie, symptoms that emerge concurrently with or after ATS use, and were not present before ATS use). The model proposed here is particularly relevant to studies of young people (typically, 15–30 years), as psychotic symptoms are more likely to emerge during this age range.

ATS-related psychosis is thought to be most evident with methamphetamine use, which is currently the most commonly used illicit ATS in Australia. Almost one in 10 Australians (9.1%) have tried methamphetamine, sold on the street under a variety of names, including “speed” and “ice”.6 Early-onset ATS use, typically occurring in the mid-to-late teens, predicts later problems and increases the chance of developing psychosis.4,7 ATS users are at particularly high risk of experiencing precipitated or exacerbated psychotic symptoms, with an estimated prevalence of psychosis that is 11 times higher than the general population.5 Compared with other substances, methamphetamine is highly neurotoxic, and its use can lead to long-term damage, with concomitant deleterious effects on cognition.8 Once ATS-related psychosis has developed, recurrence of psychotic symptoms can develop without ATS use.4 However, current narrow diagnostic definitions in the Diagnostic and statistical manual of mental disorders, fourth edition suggest that some ATS users who continue to experience psychotic symptoms for prolonged periods (beyond 1 month) should be diagnosed with a formal psychotic disorder (eg, schizophreniform disorder or schizophrenia),9 despite limited evidence for the validity of such criteria.10 Diagnostic issues are further compounded by findings that young people presenting with their first episode of psychosis have high rates of concurrent substance-use disorders.11

As with exposure to any other environmental pathogen, not all ATS users develop psychosis. The development of a psychotic episode is assumed to be related to additional vulnerability factors.7,12 Traditionally, ATS-related psychosis is seen as distinct from schizophrenia, with the latter being assumed to be a progressive or chronic deteriorating disorder.12 However, ATS-related psychosis may persist well beyond periods of acute ATS intoxication, and may recur despite abstinence.8 In clinical practice, the dominant model has been a poorly validated hierarchical approach, whereby a diagnosis of schizophrenia or another psychosis is deferred in the presence of prior or ongoing ATS use, unless the patient has been intoxicated with, or withdrawn from, the ATS for over 1 month.13

Attempts at phenotypic differentiation of ATS-related psychosis from schizophrenia and related psychoses have been of limited value. Compared with ATS users who do not develop psychoses, those experiencing psychosis report first using ATSs at a younger age, using greater amounts of ATSs, and have higher rates of premorbid schizotypal traits, major depression, alcohol dependence and antisocial personality disorder.5 These findings suggest similarities with young people who develop schizophrenia and related psychoses without ATS exposure.

The consensus of family, twin and adoption studies is that there is a large genetic contribution to the vulnerability to schizophrenia and related psychoses.14 The high heritability demonstrated in such research led to a wave of linkage studies, and more recently genome-wide association studies, that have generally been inconclusive.2,14 Such limitations have precipitated a focus on biological indices, which lie somewhere between genes and behaviour. These endophenotypes are considered to be measurable components reflecting biological modulation.15 To be valid, endophenotypes must demonstrate an association with candidate genotypes (and/or genetic loci), heritability and functional links to symptoms.15 Candidate endophenotypes in schizophrenia include neurophysiological, neuroanatomical and neuropsychological variables.

As an exemplar candidate gene for schizophrenia and other psychoses, the gene encoding the enzyme catechol-O-methyltransferase (COMT) has clear functional links, due to its role in the catabolic clearance of dopamine (a neurotransmitter centrally involved in the expression of psychosis).16 The COMT gene is one of the most widely studied in psychiatric research, probably because of its function, location, and the apparent simplicity of its genotypic variation.16 Functional polymorphisms of COMT have been shown to be associated with neurophysiological and neuropsychological indices.15 Despite this, it remains difficult to draw strong conclusions about COMT’s role in schizophrenia, as studies have shown small effect sizes and poor reproducibility.16 However, the assessment of COMT in conjunction with the profound dopamine release associated with ATS use may shed light on the underlying neural substrate of psychosis.

Guided by its relevance to the dopamine system, val158met, a COMT functional polymorphism, has been implicated in ATS use, both with respect to its rewarding properties and the potential to develop psychosis.17 Furthermore, the interaction between COMT polymorphism, working memory performance and ATS response provides evidence that individual differences in baseline dopamine functioning plays a key role.18 One study found that ATS users who showed spontaneous psychosis relapse (without subsequent ATS use) were 1.7-fold more likely to have the COMT met allele (which increases dopaminergic signalling) compared with controls.19 That is, the presence of this allele appeared to increase the risk of an adverse (psychotic) response to ATS use.19 Another study of substance use reported that the val/val genotype of COMT in combination with adolescent cannabis use was strongly associated with the later development of psychosis.20 Overall, these findings suggest that the interaction between genes and environmental factors is complex. For example, persistent negative forms of psychosis may be associated with the met allele, whereas the val allele may be associated with episodic reactive forms of psychosis.21

Our major assumption for working with the ATS-related psychosis model is that ATS-related psychosis and schizophrenia share common features and a common neural substrate (Box 2). That is, ATS-related psychosis relates to fundamental deficits in pre-attentional auditory processes and subsequent controlled processing, as indexed by neurophysiological (MMN) and neuropsychological (verbal learning and memory) markers, respectively. These markers are, in turn, mediated by gene–environment interactions, as particular genotypes (eg, COMT) differentially affect psychiatric outcomes associated with ATS use (ie, psychosis). Finally, identifying such markers in young people who use ATSs may assist with the early identification of individuals at risk of developing psychosis.

Future research on ATS-related psychosis must also consider the role of other neurotransmitters, such as glutamate. The hyper-dopaminergic state thought to be at the core of schizophrenia has been proposed to be a consequence of NMDA receptor hypofunction.32 Notably, healthy participants given the NMDA receptor antagonist ketamine, demonstrate reduced MMN, with concomitant increases in “psychotic” symptoms.33 Thus, the relationship between MMN and dopamine activity may be mediated by NMDA–glutamate function. Indeed, genetically elevated dopamine tone, coupled with heavy ATS use, may result in profound dopamine release, potentially resulting in glutamate-mediated neurotoxicity. If so, this may produce impairments in MMN and cognition that are common to both schizophrenia and ATS-related psychosis.

Psychotic disorders in young people are associated with high rates of disability and premature death. Both the clinical symptoms of psychosis and the associated cognitive impairment are indicative of significant neurobiological changes. Young people using ATSs are obviously at risk. Insufficient research has been invested in understanding the neurobiological link between ATS use and major psychiatric disorder. Studies are urgently needed to delineate the early neurobiological markers of subtle brain dysfunction (ie, the ERP marker of schizophrenia, MMN) in addition to other commonly used cognitive measures. These same neurobiological markers may be useful in those with psychoses without ATS exposure. In either group, such measures could be used to develop better early intervention and neuroprotective approaches. This is particularly relevant when onset of illness occurs during critical neurodevelopmental periods, such as adolescence and young adulthood, resulting in significant psychosocial impairment and poor developmental outcomes.

1 A model integrating gene–environment interaction research with psychiatric disorders1

---

Here, amphetamine use and genotypic susceptibility are causally linked to psychosis. This model can be tested by neurobiological markers (eg, mismatch negativity [MMN] and verbal learning or memory), which index disrupted pre-conscious attention and a dysfunctional auditory system.