Mental state changes secondary to substance use

The use of an illicit substance usually induces acute mental state changes. These are commonly anticipated and desired. Substance use can also lead to adverse effects on the user's mental state (Table 1). Substances that are commonly associated with psychotic experiences in clinical settings include cannabis, synthetic cannabinoids, stimulants (amphetamines, methamphetamines and cocaine) and hallucinogens (Weibell Reference Weibell, ten Velden Hegelstad, Johannessen and Preedy2016). Psychosis can also occur in association with misuse of inhalants (Mustonen Reference Mustonen, Niemelä and McGrath2018), nitrous oxide (Chien Reference Chien, Huang and Chen2020), ketamine (Huang Reference Huang, Lin, Hashimoto, Ide and Ikeda2020) and steroids (Hall Reference Hall, Hall and Chapman2005).

TABLE 1 Effects of commonly misused substances

a Reported but the evidence base is inconsistent.

The relationship between alcohol use and psychosis is more complex, since psychotic symptoms are a recognised feature of alcohol withdrawal, hallucinations can develop in chronic use (alcoholic hallucinosis) and there are rare reports of idiosyncratic reactions to acute use in which psychotic symptoms occur (Brown Reference Brown, Kaneko, Donovan, Donovan and Bird2019). Psychotic symptoms can also be a feature of benzodiazepine withdrawal, and they have been reported in acute intoxication with benzodiazepines (Brown Reference Brown, Kaneko, Donovan, Donovan and Bird2019). Acute use of, or withdrawal from, opiods is not generally associated with psychosis (although exceptions have been described) (Freudenreich Reference Freudenreich2020). To the contrary, the available evidence suggests that opioid agonists can have a positive therapeutic effects on psychosis (Maremmani Reference Maremmani, Rovai and Rugani2014, Maremmani Reference Maremmani, Pallucchini and Rovai2020). The explanations for mental state changes secondary to substance use are varied (Table 2).

TABLE 2 Proposed mechanisms explaining adverse mental state changes secondary to substance use/misuse

Substance use secondary to mental state/functioning characteristics

The direction of causality can also run from mental health factors to substance use and misuse (Table 3). Self-medication is a commonly reported explanation for substance use secondary to mental state disturbance (Torres Reference Torres, Papini and Preedy2016). It includes a pattern of taking the substance with the intention of reducing symptom intensity and/or symptom-related distress. This may be via a direct effect on symptoms (which include withdrawal symptoms) or an indirect effect by making the individual less distressed or concerned about the symptoms. This article focuses on psychotic symptoms co-occurring with substance misuse, but it is important to remember that other mental health problems (e.g. attention-deficit hyperactivity disorder and post-traumatic stress disorder) may increase the propensity to use (or misuse) substances (Head Reference Head, Goodwin and Debell2016; Groenman Reference Groenman, Janssen and Oosterlaan2017; Mergler Reference Mergler, Driessen and Havemann-Reinecke2018).

TABLE 3 Proposed mechanisms explaining substance use/misuse secondary to mental state disturbance

The term self-medication can be extended to include a broader range of reasons for substance use as a consequence of mental state disturbance. For example, substance use may be motivated by a desire to reduce negative affect secondary to non-symptom consequences of the mental disorder such as social and interpersonal problems (Fresán Reference Fresán, Robles-García, Tovilla-Zarate, Compton and Manseau2018). Also, substances may be taken with the intention of ameliorating the side-effects of psychotropic medication. Substance use in response to antipsychotic-induced dysphoria is one suggested mechanism for which there is some empirical support (Awad Reference Awad and Preedy2016).

Mechanisms other than self-medication have been presented to explain the increased propensity for drug use among people with schizophrenia. Impulsivity and its subfacet, negative urgency (a tendency to act rashly in negative affect states), which are risk factors for substance misuse (Smith Reference Smith and Cyders2016; D'Agostino Reference D'Agostino, Peterson and Smith2019), have been found to be increased in schizophrenia (Hoptman Reference Hoptman and Ahmed2016). Recent studies point to the possibility that neurobiological changes in schizophrenia increase proneness to substance misuse (Dyer Reference Dyer, Easey and Heron2019; Petersen Reference Petersen, Toftdahl and Nordentoft2019).

Shared risk for psychiatric disorder and substance misuse

There is empirical evidence for shared genetic risks across different psychiatric diagnostic categories, including alcohol and drug misuse (Carey Reference Carey, Agrawal and Bucholz2016; Pettersson Reference Pettersson, Larsson and Lichtenstein2016). More specifically, substance misuse and schizophrenia have been shown to have a shared genetic liability (Hartz Reference Hartz, Horton and Oehlert2017). If such liability is manifest initially in adolescent substance use, this use may increase the vulnerability to both later substance misuse and psychosis (Khokhar Reference Khokhar, Dwiel and Henricks2018), illustrating the interactive and dynamic nature of the relationship between substance use and psychosis.

Similarly, stress (arising from early adversity and/or more recent life events) is both a shared risk factor for psychosis and substance use (Mizrahi Reference Mizrahi2016; Valentino Reference Valentino2019; Kisely Reference Kisely, Mills and Strathearn2020) and a risk factor that may, in combination with substance use, cumulatively increase the risk of psychosis (Arranz Reference Arranz, Monferrer and Algora2018). A history of multiple acute and chronic adverse experiences over the life course (e.g. trauma, deprivation, homelessness, loss and abandonment) may compromise mental health and personality functioning over and above the increased risk of psychosis and substance misuse (Padgett Reference Padgett, Smith and Henwood2012).

It has been proposed that individuals who are at risk of antipsychotic-induced dysphoria (which is a risk factor for self-medication with substances) are also at increased risk of substance misuse, indicating a very specific type of shared risk underpinning coexisting psychosis and substance misuse (Awad Reference Awad and Preedy2016). Certain personality traits (such as neuroticism) have been found to be associated with both schizophrenia and substance misuse (Ohi Reference Ohi, Shimada and Nitta2016; Rogers Reference Rogers, McKinney and Asberg2018).

Biased causal assumptions

An innate predisposition to understanding cause and effect can be traced to the survival advantages of the consequent ability to make anticipatory evaluations of one's environment and, specifically, to make general predictions about future events (Stuart-Fox Reference Stuart-Fox2015). Although making sense of the complex interrelationship between a large array of events draws on explicit reasoning processes, humans have an innate tendency to implicitly generate causal representations (Muentener Reference Muentener and Schulz2014). Such implicit causal reasoning processes provide an estimate to guide immediate action, but they are prone to produce erroneous conclusions in the face of complex data. Since the causal inferences can occur in the moment without the opportunity for reflective analysis, the underlying processes are more likely to be implicit or at least to rely on pre-existing explanatory paradigms.

Clinical experience suggests that practitioners are particularly vulnerable to making assumptions that favour a formulation involving a causal chain from drug use to mental illness symptoms (often informally described as ‘drug-induced’). The notion of ‘drug-induced psychosis’ is not only overly simplistic, but the way that it is used in practitioners’ explanatory narratives can give the impression that the psychotic symptoms have a lesser status in comparison with psychotic symptoms that are not seen to be causally related to substance use (Wilson Reference Wilson, Shaw and Williams2017). Another common causal inference assumes that the patient's motivation to attend for assessment is linked to an underlying desire to obtain more substances (prescribed substances in this scenario) (so-called ‘drug-seeking’). These types of causal assumption contribute not only to the clinician's interpretation of the patient's responses to questions, but also to the approach to questioning adopted by the clinician. Thus, the clinician is prone to seek out and pay attention to information confirming their biased assumptions. It is not just a matter of the choice of questions and therefore the answers being influenced by a confirmatory bias. Patients’ feedback in clinical settings indicates that they often recognise clinicians’ biases and sometimes they respond in a way that interferes with the assessment. Examples of responses by patients include emphasising a narrative contrary to the one being suggested by the clinician, expressing frustration and/or disengaging from the process. Although understandable, they may be interpreted by the clinician as further evidence in favour of their preconceptions. For instance, if a patient responds oversensitively to a pattern of questioning that appears to be focusing on one causal explanation at the expense of other explanations (such as a drug-induced condition as opposed to an underlying mental illness), the clinician may take this as evidence that the person lacks insight that the ‘real’ problem is their drug use or that they are trying to exaggerate or even feign mental illness symptoms. The patient may, at the same time, interpret the clinician's approach as invalidating, since whatever the causal explanation, to them the symptoms are also a ‘real’ problem. Reaching an informed opinion about the nature of the relationship between substance use and symptoms is likely to be made more difficult by an assessment dynamic in which the clinician is revealing their biases in a way that may provoke a negative response from the patient.

Assessment model

A number of steps can be taken to reduce the influence of pre-existing causal assumption biases on the assessment process and the quality of the therapeutic relationship (Box 1).

First, a model of understanding that relies on the notions of vulnerability to, and the dimensional expression of, psychopathology facilitates a more nuanced explanatory formulation of psychiatric symptoms that co-occur with substance use (Szerman Reference Szerman, Parro-Torres, Didia-Attas, Javed and Fountoulakis2019). In such a model, psychiatric symptoms are seen as a reflection of a vulnerability to developing this type of psychopathology. Approaching the assessment using a framework of the dimensional expression of psychopathology not only is more consistent with the empirical evidence (Ayhan Reference Ayhan, McFarland and Pletnikov2016), but it also encourages attention to symptoms even if it is not possible to be clear about diagnosis. Thus, although active substance misuse may make a definitive diagnosis difficult, it should not be the basis of rejecting the possibility that there is an underlying vulnerability that may be manifest in symptoms even if abstinence were achieved. Removing the pressure to definitively decide whether or not there is a diagnosable mental disorder and focusing on exploring the nature and degree of symptoms allows a more considered approach to exploring the relationship between symptoms and substance use and reduces the risk of patient disengagement. This dimensional approach to the assessment does not preclude a categorical approach to decision-making, such as whether to commence a treatment or refer to a service. Where such a decision needs to be made, this can be done using the best-fit formulation at the time of the decision, while at the same time acknowledging, if necessary, that the formulation is provisional and should remain under review in light of ongoing assessments (Brendel Reference Brendel2003). The dimensional framework also allows problems on different dimensions (such as substance misuse and mental illness symptoms) to be addressed even if it is not possible to reduce the relationship to either a simple drug-induced condition or mental illness.

Second, the clinician should retain an awareness of the potential for the multiple, changing two-way interactions between substance use and symptoms in the same patient (Fig. 1).

Remaining mindful of the complexity of the relationship between substance use and psychiatric symptoms helps with the application of the third element of this model: the clinician should reflect on and resist the influence of their own biased causal representations of the relationship between substance use and mental disorder. These may comprise a tendency to apply oversimplified explanations (such as those relying on the false dichotomy between mental illness and a drug-induced condition) and/or to favour certain causal narratives. As already noted, the narratives that involve substances having a primary causal role appear to be widely prevalent, despite the empirical evidence suggesting a more complex explanation even for those cases where there appears to be a temporal sequence involving substance use first, followed by symptoms.

Fourth, clinicians should represent their open-mindedness overtly in assessment. Adopting a ‘not knowing’ stance in which the clinician remains genuinely curious reduces the constraining influence of predetermined theoretical explanations that the patient may experience as invalidating, with negative consequences for the quality of both the interaction and the information arising from it (Anderson Reference Anderson, Goolishian, McNamee and Gergen1992).

Fifth, if a clear explanatory formulation does not emerge from the assessment (which, in light of the potential for complexity, should not be a surprise), the clinician should be able to tolerate uncertainty rather than imposing unwarranted certainty (Brendel Reference Brendel2003). As already highlighted, if it is necessary to make a decision while there is still uncertainty, the clinician may indicate their preferred formulation on the basis of the available information and advise that the formulation should remain under review.

Management considerations

A comprehensive review of models of treatment for patients with coexisting psychosis and substance misuse is beyond the scope of this article, but in general the approach to management should involve collaborative care planning, attention to physical health and wider social needs, partnership working between health and other support services, adaptation of specialist mental health services for the needs of patients with coexisting psychosis and substance misuse, psychologically informed approaches and the avoidance of excluding these patients from specialist mental health services because of substance misuse (National Institute for Health and Clinical Excellence 2016; Crockford Reference Crockford and Addington2017; Baker Reference Baker, Denham, Pohlman, Badcock and Paulik2020).

With specific reference to pharmacological treatment of psychotic symptoms, it is worth noting that standard treatments used in non-comorbid patient groups may not be appropriate for patients with coexisting psychosis and substance misuse. For instance, maintenance treatments for opioid dependence are sometimes preferable to abstinence-based treatments, which may destabilise individuals with comorbid psychosis and substance misuse. It should be noted, though, that methadone can induce QTc prolongation, as can many psychiatric drugs, so patients with comorbid conditions need extra monitoring for this complication, and medications that are more likely to induce QTc prolongation should be avoided if possible. Individuals who misuse substances are at increased risk of hepatic disease (due to, for example, hepatitis C and the toxic effects of alcohol), resulting in a greater potential for medication toxicity.

There are also specific additional risk considerations in this group. For example, parental mental health problems and substance misuse, together with domestic violence, make up the so-called ‘toxic trio’ for childhood maltreatment (Brandon Reference Brandon, Bailey and Belderson2009). People with coexisting mental problems and substance misuse are at increased risk of being targeted by dealers’ attempts to infiltrate provincial towns (known as ‘county lines’ dealing) through processes such as ‘cuckooing’ (in which a dealer takes over premises for use as a provincial base for drug dealing) (Coomber Reference Coomber and Moyle2018; Williams Reference Williams and Finlay2019).