About half a million adolescents followed for almost a decade. It is the data base on which one of the most impressive studies ever conducted on the link between cannabis use in youth and the development of psychiatric disorders is based. The results, published in February 2026 on JAMA Health Forumreignite a scientific and public health debate that has never really died out.
The research involved 463,396 U.S. adolescents between the ages of 13 and 17, all served by Kaiser Permanente Northern California, one of the largest integrated health systems in the country. Unlike many previous studies, the children were not selected on the basis of problematic use or a pre-existing diagnosis but were simply subjected, during normal routine medical visits, to a universal and confidential screening for substance use. An approach that significantly reduced the selection biases typical of this type of research.
Participants were followed until age 25 (or until the end of 2023), with an average observation time of approximately 3-4 years for each outcome. The analysis cross-referenced the responses to the questionnaires with the diagnoses documented in the electronic medical records, trying to answer a precise question: are those who declare having used cannabis in the last year more likely to subsequently receive a psychiatric diagnosis?
The results
The response, adjusted for age, sex, ethnicity, socioeconomic level and use of other substances, was affirmative and quite clear for some disorders. Adolescents who reported cannabis use in the past year showed a more than doubled risk of psychotic episodes compared to abstinent peers (adjusted risk ratio: 2.19). The figure for bipolar disorder is almost identical (2.01).
Significant, but smaller, increases were also observed for depressive disorders (1.34) and anxiety disorders (1.24).
Of particular note was the fact that, on average, cannabis use preceded psychiatric diagnosis by 1.7-2.3 years. This data reinforces, at least on a temporal level, the hypothesis that exposure to cannabis is not simply a consequence of an already existing discomfort, but can actively contribute to its development.
The associations remained significant even after excluding children with a previous history of psychiatric disorders from the analysis — an important methodological step, which reduces (without completely eliminating) the risk that the results simply reflect the tendency of those who are already ill to look to cannabis as a way to feel better.
However, it would be misleading to read these data as definitive proof that cannabis causes mental illness. The authors themselves are explicit on this point: the relationship between cannabis and mental health is complex and probably bidirectional.
Those who already have prodromal symptoms – those early, often barely recognizable signs that precede a full-blown disorder – may be more likely to use cannabis to ease discomfort, even before receiving any diagnosis. In this case, cannabis use would be both the cause and effect of a pre-existing vulnerability. There are also shared risk factors – genetic, environmental, social – that can predispose both to substance use and to the development of psychiatric disorders, without one necessarily causing the other.
What the study solidly documents is a robust and persistent statistical association, which cannot be canceled out by adjustments for the main confounding variables. An important difference, but one that should not be underestimated.
Cannabis has changed
What makes the picture more worrying is the evolution of the product in circulation. Today’s cannabis is not the same as it was twenty years ago. In legalized markets, average levels of THC – the main psychoactive component – in California cannabis flowers exceed 20%. Concentrates can reach 95%. Values that were unthinkable until a few decades ago, and which make it difficult to directly apply the conclusions of studies conducted on populations exposed to much less potent products to today’s young people.
THC acts on the brain’s cannabinoid receptors, which are particularly abundant and active during adolescence, a phase in which the central nervous system is still fully developing. The most accredited neurobiological hypothesis is that early and repeated exposure can interfere with the maturation of brain areas linked to emotional regulation, motivation and reality processing – precisely those most involved in psychotic and mood disorders.
Who is more exposed
The study also found that cannabis use was more frequent among adolescents enrolled in Medicaid (the American public health program for the most vulnerable groups) and among those living in socioeconomically disadvantaged neighborhoods. A fact that raises important questions about inequalities: if the expansion of legal cannabis markets ends up concentrating risks on the already most fragile segments of the population, the problem is not only individual but structural.
However, the study does not ask to prohibit or demonize but to take a documented risk seriously, especially in a phase – that of the expansion of legalization – in which the perception of danger tends to decrease just as the availability of the product increases.
The authors’ indications go in concrete directions: clearer and more visible health warnings, stricter restrictions on marketing and packaging, rigorous control of sales to minors, targeted prevention and early intervention programs. And, on the clinical front, the systematic integration of screening for cannabis use in routine medical visits, exactly as already happens in the Kaiser Permanente system which made this study possible.
It’s about ensuring that parents, adolescents and doctors have access to up-to-date, clear and scientifically based information to make informed choices. In a context in which the product has changed, the markets have changed and knowledge about risks has deepened, it is reasonable to expect that public communication will also update accordingly.
