Cannabis and Psychosis Through the Lens of DSM-5
Nathan T. Pearson* and James H. Berry
Abstract
Evidence for an association between cannabis and psychosis has been documented in literature in many forms including experimental studies, epidemiological data, and case series. The association has implications for psychotic outcomes ranging from mild to severe and occurring over minutes to years. Due to the huge variety of exposures and outcome measures reported, creating a coherent account of all the available information is difficult. A useful way to conceptualize these wide-ranging results is to consider the association between cannabis and psychosis as it occurs within the context of widely used DSM-5 diagnoses. In the present review we examine cannabis/psychosis associations as they pertain to Cannabis Intoxication, Cannabis-Induced Psychotic Disorder, and Schizophrenia. This allows for an understanding of the cannabis and psychosis association along something approaching a continuum. Cannabis intoxication becomes Cannabis-Induced Psychotic Disorder once certain severity and duration criteria are met and Cannabis-Induced Psychotic Disorder is heavily associated with future schizophrenia diagnoses.
1. Introduction
Cannabis use is common and becoming more so. There were an estimated 192.2 million users worldwide between the ages of 15–64 in 2016. This number of worldwide users represents a 16% increase compared to 2006 [1]. Legalization of cannabis for medical use has contributed to this increase [2]. In the United States, states that have passed medical cannabis laws have seen greater increases in illicit cannabis use and in cannabis use disorders compared to states that have not passed medical cannabis laws [3]. As use has increased, population-level perceptions as to the harmfulness of cannabis have decreased [4]. Tetrahydrocannabinol (THC) is of course usually considered the active ingredient but cannabidiol (CBD), several other cannabinoids, and terpenoids play a role in the pharmacology of cannabis [5].
Cannabis use has been associated with psychotic symptoms and disorders including schizophrenia across many populations and in many different study designs [6,7,8,9]. The nature of this association is complex and can be rife with confounders. This is especially so when looking at long-term psychotic outcomes related to cannabis use. There has been debate in the literature as to whether cannabis use is a causative factor for schizophrenia or whether the association between the two rather represents some shared vulnerability to both [8,10]. Another putative reason for the association has been that cannabis use represents an attempt by people with emerging psychosis to self-medicate their symptoms though recently that explanation has been falling out of favor as a primary explanation [7,9]. Cannabis is associated with a range of psychotic symptoms of widely variable severity. Cannabis is also associated with psychotic symptoms of widely variable timeframes. Cannabis-associated psychosis can be seen on the order of minutes, hours, days, or weeks in addition to the months and years timeframe seen in a schizophrenia diagnosis [6,31,32].
A holistic understanding of the link between cannabis and psychosis requires us to look at more than just schizophrenia. For the current review we will describe the association between cannabis and psychosis as it plays out in the context of three Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) diagnoses: Cannabis Intoxication, Cannabis-Induced Psychotic Disorder (CIPD), and Schizophrenia [22]. It is useful to use this lens because the DSM-5 criteria are very widely used and accepted. This gives us firmer footing to describe different “kinds” of cannabis-associated psychotic experiences than we would have otherwise. Delineating the plethora of cannabis/psychosis associations in the literature into these categories is merely meant as a useful way to conceptualize the associations and is not meant to strictly indicate the original works referenced in this review themselves were working with DSM-5 criteria. No single diagnostic framework is used consistently in the cannabis/psychosis literature with DSM-III, DSM-IV, DSM-5, ICD-8, ICD-9, and ICD-10 diagnoses all being used at different times as well as the use of a variety of clinical psychosis rating scales.
2. Cannabis Intoxication
This is a diagnosis made when there is recent cannabis use, significant behavioral or psychological changes that developed during or shortly after cannabis use, and physical stigmata indicating the intoxication such as conjunctival injection or dry mouth [22] (p. 516). With respect to timing cannabis intoxication occurs within minutes for inhalational use but onset can take hours when cannabis is ingested. The symptoms typically last 3–4 hours but depending on dose and tolerance can persist up to 24 hours [33]. This is basically the standard cannabis “high” documented in the DSM-5 as a mental disorder in situations where it causes neuropsychiatric symptoms that are problematic.
Psychotic symptoms are not necessary for a cannabis intoxication diagnosis but can be part of the disorder with the caveat that insight must remain intact and the psychotic symptoms must not be sufficiently severe or persistent enough to warrant clinical attention for their own sake. If the symptoms are severe or persistent enough to warrant clinical attention for their own sake, then that would move us to a CIPD diagnosis. CIPD is discussed in the following section.
Most individuals meeting criteria for Cannabis Intoxication will not present for acute medical care, so looking at psychotic symptoms within this disorder gives us a sense of what psychotic symptoms can be associated with cannabis use in non-clinical populations. We can also note that the vast majority of worldwide cannabis users have at some point met criteria for a Cannabis Intoxication diagnosis (becoming intoxicated is to some degree the goal of any cannabis use) so the psychotic symptoms experienced therein have potential to effect a huge number of persons worldwide. Having described what Cannabis Intoxication is per the DMS-V we can look at the evidence associating cannabis and psychosis as might be seen within the parameters of this disorder.
A 2004 double-blind placebo-controlled experimental study by D’Souza et al that documented psychotic symptoms in healthy subjects after intravenous THC administration provides us with a straightforward and useful example [24]. By administering the Positive and Negative Syndrome Scale (PANSS) at different timepoints before and after intravenous THC administration, the transient or “intoxication” effects of THC with respect to psychotic symptoms were able to be followed. The PANSS is commonly used in research to monitor symptoms of psychosis [34]. The PANSS was administered 60 minutes prior to injection, 10 minutes after, 80 minutes after, and 200 minutes after. It was found that a modest mean increase in positive symptoms occurred and peaked 10 minutes after injection and returned to baseline by 200 minutes after injection. A transient increase in mean negative symptoms also was seen after injection and again symptoms returned to baseline by 200 minutes. Due to the study design using intravenous THC as opposed to inhaled or ingested THC the results seen here show quicker on/off effects than what would be experienced in the population at large where inhalation or ingestion are the common administration routes. The transient increases seen in this study in both positive and negative symptoms measured via PANSS peaked at approximate scores of 10. Putting these results in context the possible PANSS scores for either positive or negative symptom subscales are 7 to 49 and PANSS averages for schizophrenic persons have been reported at 18.2 for positive symptoms and 21.01 for negative symptoms [34]. So, we see that while increases in psychotic symptoms were seen in this study using healthy subjects the magnitude of symptoms was quite small and transient as mentioned above. Also, it is notable that a dose–response relationship was seen in this study with more psychotic symptoms occurring with 5 mg THC injection compared to 2.5 mg THC injection. This finding of an acute transient increase in psychotic symptoms after intravenous THC administration in healthy subjects was replicated by Morrison et al. in 2009 [35]. In human laboratory studies, concerning healthy individuals being administered THC at high doses, it has been approximated that 35–50% will experience psychotic symptoms [16].
The largest pool of evidence describing acute transient psychotic symptoms associated with cannabis use can be found in studies documenting general population cannabis users self-reported psychotic experiences during acute use. This data also gives us some sense of the proportion of cannabis users that experience psychotic effects acutely when using the drug in naturalistic settings. A 2003 review by Green et al. examined 12 studies that surveyed users’ subjective effects when using cannabis [36]. Three of the studies used open-ended questions to elicit subjective effects while nine studies used closed-ended questioning (checklists or questionnaires). All studies used had a sample size over 30. The open-ended studies found 2–14% of subjects reported hallucinations while 6–15% of subjects reported paranoia. The closed-ended questioning studies allowed for results to be combined when the surveys asked similar or identical questions about subjective effects of cannabis. Of subjects in closed-ended questioning studies, 19.8% reported hallucinations/visions (N = 3082), while 51.4% reported paranoia (N = 2708). It is interesting to note that close-ended questioning elicited more psychotic symptoms than open-ended questioning. Cannabis users were seen throughout these studies to endorse mostly beneficial effects when describing effects spontaneously and to endorse proportionally more harmful/bothersome effects when made to consider these via checklists and questionnaires. This is congruent with the cognitive biases typically associated with substance use disorders. It is interesting to see this even in this non-clinical population [37].
There is also evidence from a study by Sami et al. that former cannabis users were more likely to report having had psychotic experiences with cannabis than current cannabis users who were more likely to report pleasurable experiences [38]. Current users who indicated a future intention to quit were more likely to have had psychotic experiences with cannabis than current users who indicated no desire to quit. These findings (along with the differences Green et al reported with open vs closed questions) suggest a potential “in” for insight-driven interventions for helping people quit cannabis such as motivational interviewing.
As it is clear many users do not report psychotic effects from acute cannabis use it becomes important to ask what kind of person is at risk for these bothersome acute effects. Mason et al. looked at acute psychotic symptoms associated with cannabis use and stratified their cannabis users based on high or low pre-intoxication scores on the Schizotypal Personality Questionnaire (SPQ) [39]. SPQ was used as a proxy for baseline psychotic symptoms and can be taken to indicate risk or susceptibility to psychosis [40]. This study found greater acute transient effects on psychotic symptoms in individuals with higher SPQ scores at baseline. Acute effects were taken as the difference between Psychotomimetic States Inventory (PSI) scores 10–15 minutes after use and PSI scores 3–5 days later after at least 24 hours of cannabis abstinence. This result provides evidence that certain individuals, especially those experiencing some mild psychotic symptoms at baseline, are more prone to acute transient psychotic symptoms associated with cannabis use than others.
Having described some of the evidence for an acute association between cannabis and psychosis, as could be seen in a Cannabis Intoxication diagnosis, we will move on to describe Cannabis Intoxication’s more severe and persistent progeny, CIPD.
3. Cannabis-Induced Psychotic Disorder (CIPD)
Substance-Induced psychotic disorders are recognized by the DSM-5 and are placed in the category of Schizophrenia Spectrum and Other Psychotic Disorders. Substance-Induced psychotic disorders related to practically all substances of abuse can be described using this diagnosis [22] (pp 110–115).
A diagnosis of Cannabis-Induced Psychotic Disorder is given when one or both of hallucinations and delusions are present, the hallucinations and/or delusions developed during or soon after cannabis intoxication, the disturbance does not occur exclusively during the course of a delirium, and the disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning. Other criteria for the disorder are that cannabis should be thought to be capable of producing the disturbance seen and that the disturbance should not be able to be better explained by an independent psychotic disorder that is not cannabis-induced (such as pre-existing schizophrenia). The DSM-5 suggests that if symptoms last longer than one month a diagnosis other than CIPD should be considered [22] (p. 110).
Substance-induced psychotic disorders generally can occur in the context of recent intoxication or withdrawal from a substance (for example with alcohol) but in the case of cannabis only psychotic symptoms occurring in the context of recent intoxication are thought to appropriately lead to a CIPD diagnosis [22] (p. 114).
Several things differentiate CIPD from Cannabis Intoxication. First and foremost is that in CIPD the hallucinations and/or delusions are the focus of the clinical presentation and are severe enough to warrant clinical attention/treatment as opposed to the psychotic symptoms that can be seen in Cannabis Intoxication which are more mild and self-limited and are not even required to make that diagnosis. A further distinction is that the hallucinations in CIPD are experienced without insight whereas in Cannabis Intoxication the hallucinations when present are experienced with insight intact and the DSM-5 linguistically downgrades these in places from frank hallucinations to “perceptual disturbances.” In addition to greater intensity/severity of symptoms CIPD can also have a much longer duration than Cannabis Intoxication. Cannabis Intoxication will necessarily resolve within 24 hours whereas CIPD can last for days and even weeks after cannabis exposure [6]. However, criteria for CIPD could also be met in a presentation only lasting on the order of hours if the symptoms are severe.
The concept of a cannabis psychosis apart from simple intoxication has been recognized for literally hundreds of years—take the following example from 1779 describing a preparation of cannabis known as “Bangue”.
“Bangue is an intoxicating herb; in the use of which it is hard to say what pleasure can be found, it being very disagreeable to the taste and violent in its operation which produces a temporary madness, that in some, when designedly taken for that purpose, ends in running, what they call a muck, furiously killing every one they meet without distinction till themselves are knocked on the head like mad dogs [41] (p. 21).”
Another historical example of the recognition of CIPD consistent cannabis/psychosis association comes from French psychiatrist Dr Jacques-Joseph Moreau in 1845, describing the effects of hashish:
Abstract
Evidence for an association between cannabis and psychosis has been documented in literature in many forms including experimental studies, epidemiological data, and case series. The association has implications for psychotic outcomes ranging from mild to severe and occurring over minutes to years. Due to the huge variety of exposures and outcome measures reported, creating a coherent account of all the available information is difficult. A useful way to conceptualize these wide-ranging results is to consider the association between cannabis and psychosis as it occurs within the context of widely used DSM-5 diagnoses. In the present review we examine cannabis/psychosis associations as they pertain to Cannabis Intoxication, Cannabis-Induced Psychotic Disorder, and Schizophrenia. This allows for an understanding of the cannabis and psychosis association along something approaching a continuum. Cannabis intoxication becomes Cannabis-Induced Psychotic Disorder once certain severity and duration criteria are met and Cannabis-Induced Psychotic Disorder is heavily associated with future schizophrenia diagnoses.
1. Introduction
Cannabis use is common and becoming more so. There were an estimated 192.2 million users worldwide between the ages of 15–64 in 2016. This number of worldwide users represents a 16% increase compared to 2006 [1]. Legalization of cannabis for medical use has contributed to this increase [2]. In the United States, states that have passed medical cannabis laws have seen greater increases in illicit cannabis use and in cannabis use disorders compared to states that have not passed medical cannabis laws [3]. As use has increased, population-level perceptions as to the harmfulness of cannabis have decreased [4]. Tetrahydrocannabinol (THC) is of course usually considered the active ingredient but cannabidiol (CBD), several other cannabinoids, and terpenoids play a role in the pharmacology of cannabis [5].
Cannabis use has been associated with psychotic symptoms and disorders including schizophrenia across many populations and in many different study designs [6,7,8,9]. The nature of this association is complex and can be rife with confounders. This is especially so when looking at long-term psychotic outcomes related to cannabis use. There has been debate in the literature as to whether cannabis use is a causative factor for schizophrenia or whether the association between the two rather represents some shared vulnerability to both [8,10]. Another putative reason for the association has been that cannabis use represents an attempt by people with emerging psychosis to self-medicate their symptoms though recently that explanation has been falling out of favor as a primary explanation [7,9]. Cannabis is associated with a range of psychotic symptoms of widely variable severity. Cannabis is also associated with psychotic symptoms of widely variable timeframes. Cannabis-associated psychosis can be seen on the order of minutes, hours, days, or weeks in addition to the months and years timeframe seen in a schizophrenia diagnosis [6,31,32].
A holistic understanding of the link between cannabis and psychosis requires us to look at more than just schizophrenia. For the current review we will describe the association between cannabis and psychosis as it plays out in the context of three Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) diagnoses: Cannabis Intoxication, Cannabis-Induced Psychotic Disorder (CIPD), and Schizophrenia [22]. It is useful to use this lens because the DSM-5 criteria are very widely used and accepted. This gives us firmer footing to describe different “kinds” of cannabis-associated psychotic experiences than we would have otherwise. Delineating the plethora of cannabis/psychosis associations in the literature into these categories is merely meant as a useful way to conceptualize the associations and is not meant to strictly indicate the original works referenced in this review themselves were working with DSM-5 criteria. No single diagnostic framework is used consistently in the cannabis/psychosis literature with DSM-III, DSM-IV, DSM-5, ICD-8, ICD-9, and ICD-10 diagnoses all being used at different times as well as the use of a variety of clinical psychosis rating scales.
2. Cannabis Intoxication
This is a diagnosis made when there is recent cannabis use, significant behavioral or psychological changes that developed during or shortly after cannabis use, and physical stigmata indicating the intoxication such as conjunctival injection or dry mouth [22] (p. 516). With respect to timing cannabis intoxication occurs within minutes for inhalational use but onset can take hours when cannabis is ingested. The symptoms typically last 3–4 hours but depending on dose and tolerance can persist up to 24 hours [33]. This is basically the standard cannabis “high” documented in the DSM-5 as a mental disorder in situations where it causes neuropsychiatric symptoms that are problematic.
Psychotic symptoms are not necessary for a cannabis intoxication diagnosis but can be part of the disorder with the caveat that insight must remain intact and the psychotic symptoms must not be sufficiently severe or persistent enough to warrant clinical attention for their own sake. If the symptoms are severe or persistent enough to warrant clinical attention for their own sake, then that would move us to a CIPD diagnosis. CIPD is discussed in the following section.
Most individuals meeting criteria for Cannabis Intoxication will not present for acute medical care, so looking at psychotic symptoms within this disorder gives us a sense of what psychotic symptoms can be associated with cannabis use in non-clinical populations. We can also note that the vast majority of worldwide cannabis users have at some point met criteria for a Cannabis Intoxication diagnosis (becoming intoxicated is to some degree the goal of any cannabis use) so the psychotic symptoms experienced therein have potential to effect a huge number of persons worldwide. Having described what Cannabis Intoxication is per the DMS-V we can look at the evidence associating cannabis and psychosis as might be seen within the parameters of this disorder.
A 2004 double-blind placebo-controlled experimental study by D’Souza et al that documented psychotic symptoms in healthy subjects after intravenous THC administration provides us with a straightforward and useful example [24]. By administering the Positive and Negative Syndrome Scale (PANSS) at different timepoints before and after intravenous THC administration, the transient or “intoxication” effects of THC with respect to psychotic symptoms were able to be followed. The PANSS is commonly used in research to monitor symptoms of psychosis [34]. The PANSS was administered 60 minutes prior to injection, 10 minutes after, 80 minutes after, and 200 minutes after. It was found that a modest mean increase in positive symptoms occurred and peaked 10 minutes after injection and returned to baseline by 200 minutes after injection. A transient increase in mean negative symptoms also was seen after injection and again symptoms returned to baseline by 200 minutes. Due to the study design using intravenous THC as opposed to inhaled or ingested THC the results seen here show quicker on/off effects than what would be experienced in the population at large where inhalation or ingestion are the common administration routes. The transient increases seen in this study in both positive and negative symptoms measured via PANSS peaked at approximate scores of 10. Putting these results in context the possible PANSS scores for either positive or negative symptom subscales are 7 to 49 and PANSS averages for schizophrenic persons have been reported at 18.2 for positive symptoms and 21.01 for negative symptoms [34]. So, we see that while increases in psychotic symptoms were seen in this study using healthy subjects the magnitude of symptoms was quite small and transient as mentioned above. Also, it is notable that a dose–response relationship was seen in this study with more psychotic symptoms occurring with 5 mg THC injection compared to 2.5 mg THC injection. This finding of an acute transient increase in psychotic symptoms after intravenous THC administration in healthy subjects was replicated by Morrison et al. in 2009 [35]. In human laboratory studies, concerning healthy individuals being administered THC at high doses, it has been approximated that 35–50% will experience psychotic symptoms [16].
The largest pool of evidence describing acute transient psychotic symptoms associated with cannabis use can be found in studies documenting general population cannabis users self-reported psychotic experiences during acute use. This data also gives us some sense of the proportion of cannabis users that experience psychotic effects acutely when using the drug in naturalistic settings. A 2003 review by Green et al. examined 12 studies that surveyed users’ subjective effects when using cannabis [36]. Three of the studies used open-ended questions to elicit subjective effects while nine studies used closed-ended questioning (checklists or questionnaires). All studies used had a sample size over 30. The open-ended studies found 2–14% of subjects reported hallucinations while 6–15% of subjects reported paranoia. The closed-ended questioning studies allowed for results to be combined when the surveys asked similar or identical questions about subjective effects of cannabis. Of subjects in closed-ended questioning studies, 19.8% reported hallucinations/visions (N = 3082), while 51.4% reported paranoia (N = 2708). It is interesting to note that close-ended questioning elicited more psychotic symptoms than open-ended questioning. Cannabis users were seen throughout these studies to endorse mostly beneficial effects when describing effects spontaneously and to endorse proportionally more harmful/bothersome effects when made to consider these via checklists and questionnaires. This is congruent with the cognitive biases typically associated with substance use disorders. It is interesting to see this even in this non-clinical population [37].
There is also evidence from a study by Sami et al. that former cannabis users were more likely to report having had psychotic experiences with cannabis than current cannabis users who were more likely to report pleasurable experiences [38]. Current users who indicated a future intention to quit were more likely to have had psychotic experiences with cannabis than current users who indicated no desire to quit. These findings (along with the differences Green et al reported with open vs closed questions) suggest a potential “in” for insight-driven interventions for helping people quit cannabis such as motivational interviewing.
As it is clear many users do not report psychotic effects from acute cannabis use it becomes important to ask what kind of person is at risk for these bothersome acute effects. Mason et al. looked at acute psychotic symptoms associated with cannabis use and stratified their cannabis users based on high or low pre-intoxication scores on the Schizotypal Personality Questionnaire (SPQ) [39]. SPQ was used as a proxy for baseline psychotic symptoms and can be taken to indicate risk or susceptibility to psychosis [40]. This study found greater acute transient effects on psychotic symptoms in individuals with higher SPQ scores at baseline. Acute effects were taken as the difference between Psychotomimetic States Inventory (PSI) scores 10–15 minutes after use and PSI scores 3–5 days later after at least 24 hours of cannabis abstinence. This result provides evidence that certain individuals, especially those experiencing some mild psychotic symptoms at baseline, are more prone to acute transient psychotic symptoms associated with cannabis use than others.
Having described some of the evidence for an acute association between cannabis and psychosis, as could be seen in a Cannabis Intoxication diagnosis, we will move on to describe Cannabis Intoxication’s more severe and persistent progeny, CIPD.
3. Cannabis-Induced Psychotic Disorder (CIPD)
Substance-Induced psychotic disorders are recognized by the DSM-5 and are placed in the category of Schizophrenia Spectrum and Other Psychotic Disorders. Substance-Induced psychotic disorders related to practically all substances of abuse can be described using this diagnosis [22] (pp 110–115).
A diagnosis of Cannabis-Induced Psychotic Disorder is given when one or both of hallucinations and delusions are present, the hallucinations and/or delusions developed during or soon after cannabis intoxication, the disturbance does not occur exclusively during the course of a delirium, and the disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning. Other criteria for the disorder are that cannabis should be thought to be capable of producing the disturbance seen and that the disturbance should not be able to be better explained by an independent psychotic disorder that is not cannabis-induced (such as pre-existing schizophrenia). The DSM-5 suggests that if symptoms last longer than one month a diagnosis other than CIPD should be considered [22] (p. 110).
Substance-induced psychotic disorders generally can occur in the context of recent intoxication or withdrawal from a substance (for example with alcohol) but in the case of cannabis only psychotic symptoms occurring in the context of recent intoxication are thought to appropriately lead to a CIPD diagnosis [22] (p. 114).
Several things differentiate CIPD from Cannabis Intoxication. First and foremost is that in CIPD the hallucinations and/or delusions are the focus of the clinical presentation and are severe enough to warrant clinical attention/treatment as opposed to the psychotic symptoms that can be seen in Cannabis Intoxication which are more mild and self-limited and are not even required to make that diagnosis. A further distinction is that the hallucinations in CIPD are experienced without insight whereas in Cannabis Intoxication the hallucinations when present are experienced with insight intact and the DSM-5 linguistically downgrades these in places from frank hallucinations to “perceptual disturbances.” In addition to greater intensity/severity of symptoms CIPD can also have a much longer duration than Cannabis Intoxication. Cannabis Intoxication will necessarily resolve within 24 hours whereas CIPD can last for days and even weeks after cannabis exposure [6]. However, criteria for CIPD could also be met in a presentation only lasting on the order of hours if the symptoms are severe.
The concept of a cannabis psychosis apart from simple intoxication has been recognized for literally hundreds of years—take the following example from 1779 describing a preparation of cannabis known as “Bangue”.
“Bangue is an intoxicating herb; in the use of which it is hard to say what pleasure can be found, it being very disagreeable to the taste and violent in its operation which produces a temporary madness, that in some, when designedly taken for that purpose, ends in running, what they call a muck, furiously killing every one they meet without distinction till themselves are knocked on the head like mad dogs [41] (p. 21).”
Another historical example of the recognition of CIPD consistent cannabis/psychosis association comes from French psychiatrist Dr Jacques-Joseph Moreau in 1845, describing the effects of hashish: