From LSD to cannabis: an old suspicion about substances and madness
The idea that certain psychoactive substances can mimic—or even trigger—psychosis is not new. In the mid-20th century, LSD was the star of the so-called psychotomimetic model: due to its chemical similarity to serotonin and the changes it produced in perception, cognition, and emotion, it was considered an experimental window into schizophrenia. Some researchers ingested it to “experience psychosis” firsthand.
Other candidates followed. DMT—present in ayahuasca, but also endogenous in our bodies—was studied for similar reasons, though measurements in schizophrenic patients were not conclusive. Psilocybin from mushrooms completes that group of serotonergic hallucinogens. Today, it is assumed that these compounds primarily reproduce acute manifestations, not the chronic clinical picture of the disease, and that their effects tend to be more visual than auditory, unlike the characteristic auditory hallucinations of schizophrenia.
The picture became more complex with other models: amphetamine and paranoia opened the dopaminergic model, useful for “positive” symptoms but blind to “negative” ones (apathy, attention deficits, cognitive problems). Substances like PCP and ketamine, which act on glutamate and produce dissociation, later expanded the repertoire. The underlying message is important: there is no single model of psychosis, but rather several that illuminate different facets.
Cannabis under the microscope: an association documented for centuries
The suspicion regarding cannabis is ancient. In traditional Chinese medicine, there are millennial references stating that its excessive use could cause visions or communication with spirits. It is advisable to read these accounts with caution: seeing or hearing entities is strongly determined by culture and does not, in itself, equate to a disorder. However, it is also true that, in certain people and contexts, those experiences can be part of a clinical picture.
As early as the 19th century, the French psychiatrist Moreau de Tours described similarities between the effects of cannabis and some psychotic symptoms—a century before LSD gave its name to the psychotomimetic model.
Acute effects: temporary, but real
In both consumer reports and controlled studies, there is evidence that cannabis can produce effects lasting minutes or hours that resemble aspects of psychosis: paranoia, illusions, hallucinations, delusional ideas, and mental confusion. The risk seems higher in people with little experience, at high doses, or when there is some predisposition. These phenomena have been reproduced in the laboratory with pure THC and are listed as potential adverse effects in the package inserts of medical-use derivatives.
Two key nuances. First: these acute effects are temporary and do not constitute psychosis in themselves. Second: in a minority—novices, very high doses, or prior predisposition—the symptoms can last for weeks and require professional attention. In those cases, speaking of a psychotic break triggered by cannabis is a reasonable explanation.
The picture also has a paradoxical edge. Cannabis derivative use is frequent among people with schizophrenia, and administering pure THC can temporarily exacerbate their symptoms. Why do they consume it, then? One hypothesis is that, although it may worsen some symptoms and complicate treatment, it could alleviate others. In fact, there are lines of research into the potential of CBD—and even THC in certain contexts—against psychotic symptoms. This is open research, not an established indication: much remains to be studied before any approved therapeutic application.
The long term: risk factor or cause?
The difficult question is whether cannabis can cause psychosis or schizophrenia in a healthy person with no history. Here, the evidence is controversial, although several indications point to a relationship.
The most cited data come from epidemiology: those who have used cannabis tend to receive more schizophrenia diagnoses in subsequent years than those who have not. When controlling for other variables that could also explain the result, the effect is reduced, but it does not disappear. A dose-dependent pattern and the weight of age are also described: consumption in adolescence and in larger quantities is associated with higher risk, and some reviews suggest that the onset of the disorder could be earlier in consumers. One influential review estimated an approximately 40% increase in the relative risk of a subsequent diagnosis.
Conversely, other researchers consider the risk to be only moderate and point to an uncomfortable counterargument: despite the expansion of cannabis use in the West over decades, schizophrenia rates have not skyrocketed proportionally, and have even stabilized. If the relationship were strongly causal, one would expect otherwise.
What we can (and cannot) affirm
In summary: cannabis can induce, in a transient manner, states similar to psychotic symptoms; people with psychosis or a predisposition tend to experience more intense and lasting effects, and higher rates of relapse. Long-term, the association with the development of psychosis is consistent, but causality remains under debate.
The reason is methodological: the relationship meets some criteria for causality, but not all. The vast majority of those who use cannabis do not develop psychosis, and the majority of people with schizophrenia have never used it. The most prudent conclusion is that exposure to cannabis is neither necessary nor sufficient: it likely acts as one more component factor that, together with others (genetic, environmental, or still unknown), can contribute to the condition.
Critical reading and harm reduction
Some keys to interpreting this topic without falling into alarmism or minimization:
- Correlation is not causation. Just because two things appear together in data does not prove that one causes the other; there may be shared variables (prior vulnerability, use of other substances, social factors).
- Reverse causality is plausible. Early prodromal symptoms could push someone toward consumption, and not the other way around.
- Risk is not distributed equally. Family history of psychosis, onset in adolescence, high frequency, and very potent products concentrate most of the described risk.
- Context matters. Mood, environment, and expectations modulate the probability of an unpleasant acute experience.
If symptoms appear that persist beyond the acute effect—stable paranoid ideas, voices, disorganized thinking, intense anxiety that does not remit—it is advisable to seek professional help without delay. This piece is educational and does not replace a clinical assessment.
This review adapts and reworks an educational text by Rafael Guimarães dos Santos for Psiconáutica.org. Academic sources that have addressed this issue include works by Ben Amar and Potvin, Moore et al. (review in The Lancet), D’Souza et al., Large et al., and the research line of Zuardi, Crippa, and Hallak on cannabinoids in psychiatry.
