A young man once sat across from me, hollow-eyed, trembling, and confused.
He had tapered off his psychiatric medication and started using cannabis to help him sleep. “It worked,” he told me quietly. “Then it didn’t. Now I can’t tell what’s real anymore.”
I see versions of him almost every week. Some arrive hearing voices. Others are gripped by paranoia or overwhelming anxiety. Most are young. Many are using high-potency cannabis. Stories like these raise a question clinicians increasingly encounter in everyday practice: What happens when a drug that may destabilize vulnerable brains becomes normalized and widely available?
When it comes to normalization, the most effective marketing word in America may not be organic, wellness, or freedom. It may be legal. Under bright lights and tasteful packaging, cannabis has completed one of the more striking public transformations in modern American life. Legalization has changed more than the legal status of cannabis. It has changed assumptions. In public conversation, legal increasingly implies safe, or at least harmless enough to ignore. But legality does not change biology.
And as cannabis becomes a normalized consumer product across much of the U.S., physicians are increasingly left to confront a different question: What does widespread cannabis exposure, especially during adolescence, actually do to the developing brain?
Estimated state-legal cannabis sales rose from roughly $19 billion in 2020 to more than $31 billion in 2025. Adult-use cannabis tax revenue reached $4.42 billion in 2024 and has remained above roughly $4 billion annually since 2021. Public perception has shifted accordingly. Americans are more likely than ever to support the legalization of marijuana in the U.S.
But the emerging psychiatric literature continues to suggest a more complicated reality. A recent cohort study published in JAMA Health Forum offers one of the largest longitudinal examinations yet of adolescent cannabis use and psychiatric outcomes.
Researchers followed 463,396 adolescents aged 13–17 years who were universally screened for past-year cannabis use during routine pediatric care and tracked through age 25 years or the end of 2023. The findings were striking. Adolescents who reported cannabis use in the past year had significantly higher risk of several psychiatric diagnoses by young adulthood, with the strongest associations observed for psychotic and bipolar disorders.
Past-year cannabis use was associated with:
Even after additional adjustment for prior psychiatric conditions, the associations remained significant, with hazard ratios of 1.92 for psychotic disorder and 1.73 for bipolar disorder.
Timing also mattered. Associations with depressive and anxiety disorders were strongest at younger ages and were no longer statistically significant by ages 21–25. In contrast, the associations with psychotic and bipolar disorders remained stronger and more persistent.
The study does not prove that cannabis directly causes psychiatric illness. The relationship may be partly bidirectional. Some adolescents may use cannabis in response to distress or emerging symptoms. Residual confounding is always possible in observational research. But several aspects of the findings are difficult to dismiss.
Among participants who later developed psychiatric diagnoses, the mean time from first reported cannabis use to diagnosis ranged from 1.7 to 2.3 years. Associations remained significant after adjustment for time-varying alcohol and other substance use, after controlling for baseline psychiatric disorders, and in sensitivity analyses excluding adolescents with prior psychiatric histories.
Perhaps most notable, the study detected these associations using a relatively crude exposure measure: simply whether adolescents reported cannabis use within the past year. The signal emerged even without distinguishing daily use, THC dose, or product type.
These findings do not stand alone. A growing body of literature has linked cannabis exposure, particularly frequent or high-potency use, to psychosis-spectrum disorders. In a study published in Lancet Psychiatry, daily cannabis use was associated with over threefold higher odds of psychotic disorder compared with never use, increasing to nearly fivefold among users of high-potency cannabis. Other meta-analyses (2019, 2021) have similarly demonstrated dose-response patterns.
Other population-level signals are emerging as well. A 2023 Danish registry study estimated that nearly 1 in 3 schizophrenia cases among young men may be linked to cannabis use disorder, suggesting a substantial population-level contribution in high-risk groups.
Canada offers another window into the trend. The proportion of new schizophrenia cases linked to cannabis use disorder has nearly tripled over the past 17 years, rising from about 4% before legalization to roughly 10% afterward. Among younger men, nearly 1 in 5 new cases are now associated with cannabis use disorder. Another factor complicating the discussion is potency. According to the National Institute on Drug Abuse, average delta-9 THC concentration in seized cannabis increased from 3.96% in 1995 to 16.14% in 2022, while many dispensary products, particularly concentrates, can exceed 40% THC.
Public health agencies have begun to acknowledge these risks. The CDC warns that regular cannabis use during adolescence and early adulthood can impair memory, learning, attention, and emotional regulation, with potential long-term effects on brain development. A recent JAMA Psychiatry article also cautioned that federal efforts to reschedule cannabis risk should not become politically driven decisions that sideline psychiatric evidence and emerging mental health concerns.
Public debate around cannabis often centers on legalization versus criminalization. That question is important. Criminal penalties carry real social costs. But the real question is whether society can admit that a product may be less deserving of criminal penalties while still carrying real and dose-related psychiatric risk.
For clinicians, this distinction matters. The brain does not respond to political categories. It responds to exposure: timing, dose, potency, and vulnerability. Regulation can increase or decrease the odds that a 17-year-old encounters high-potency products, that a 22-year-old with emerging psychosis has easy access to concentrates, or that a primary care clinic identifies cannabis use disorder before crisis hits.
A more clinically grounded policy discussion would acknowledge these realities. Legal cannabis markets can coexist with stronger protections, clearer THC labeling, tighter regulation of high-potency products, and greater investment in screening and early psychiatric intervention.
Medicine’s role in this landscape is not to relitigate culture wars. It is to restore proportion and informed consent. The point is not to say cannabis risk is all-or-nothing. The point is almost the opposite: the risk is patterned. It rises along lines the evidence keeps identifying: weekly and daily use, higher potency, younger age, and particular vulnerability to psychosis-spectrum outcomes. For patients with personal or family histories of psychosis, that message becomes more urgent, especially given evidence that cannabis-induced psychosis has a substantial transition rate to schizophrenia.
Cannabis may now be legal in much of the U.S. It may be profitable, widely marketed, and increasingly normalized.
But harmless is a different claim entirely.
Has the legal landscape of cannabis changed the way you approach diagnostic screening or informed consent with your younger patients? Share your experience in the comments.
Fawad Taj, MD is an assistant professor of psychiatry at Case Western Reserve University and a psychiatrist at University Hospitals of Cleveland with over a decade of experience in psycho-oncology, emergency psychiatry, and serious mental illness. His leadership lies at the intersection of mental health, policy, and community advocacy. Dr. Taj is a 2025–2026 Doximity Op-Med Fellow.
Collage Jennifer Bogartz / Shutterstock
