Daily Cannabis Use Raises Psychosis Risk. It Still Won’t Fix Your Anxiety.

Ask someone who smokes cannabis every day why they started, and “I couldn’t sleep” or “it’s the only thing that touches my anxiety” comes up before “it’s fun” almost every time. Ask a clinician whether that same daily habit is dangerous for someone with a family history of psychosis, and until this month the honest answer was some version of “we think so, but the evidence is messier than either side admits.” The Lancet Psychiatry just ran three papers together in its July 2026 issue that don’t clean up the mess so much as draw its actual shape — and the shape is inconvenient for almost everyone who’s already decided how they feel about weed.

Heavy, daily cannabis use is a real, credible contributor to psychosis risk, and most of what cannabis products are marketed to treat — anxiety, depression, PTSD, sleep — they don’t actually treat.

Both of those sentences are true at once, and neither one cancels the other out. That’s the whole point of publishing them together.

The psychosis finding is narrower than the headlines will make it sound

The lead paper, from Tom Freeman’s group at the University of Bath, isn’t a new study — it’s a synthesis, meaning the authors laid four different kinds of evidence side by side to see if they told the same story: population data tracking who develops psychosis and when, genetic studies isolating inherited risk, controlled dosing experiments in people, and lab research in animals (Freeman et al., 2026). All four pointed the same direction. The authors’ conclusion is that there is credible evidence daily cannabis use is a “contributory cause” of psychosis — plain language, that means daily use measurably raises the odds of developing psychosis over time, not that any single episode of use triggers it. It is a dose-and-frequency finding, not a first-time-user warning.

The population it applies to is specific: people, especially adolescents and young adults in their mid-teens to mid-twenties, who use cannabis daily and go on to develop cannabis use disorder — the clinical term for compulsive use that continues despite the problems it’s causing, not just regular or recreational use. Within that group, the paper also found elevated rates of anxiety, depression, and bipolar disorder, though the authors flag that the bipolar link rests on fewer high-quality genetic and population studies than the psychosis link does, so treat that piece as suggestive rather than settled. For calibration: a 2019 case-control study across 11 European sites found that daily users of high-potency cannabis were roughly five times more likely to be diagnosed with a first episode of psychosis than people who’d never used (Di Forti et al., 2019) — the new synthesis paper is what turns that kind of individual finding into a credible causal claim across the whole evidence base, not a one-off study result.

What cannabis is being sold to treat, it mostly doesn’t treat

The second paper, led by researchers including Jack Wilson at the University of Sydney’s Matilda Centre, asked a blunter question: does the evidence actually support prescribing cannabinoids for mental health and substance use conditions? The team ran a systematic review and meta-analysis — meaning they searched for every qualifying trial rather than citing whichever ones supported a conclusion, screened 5,774 studies, and kept 54 randomized controlled trials covering 2,477 participants going back to 1980 (Wilson et al., 2026). The result: no evidence cannabinoids help anxiety, PTSD, or psychosis, and zero qualifying trials existed for depression at all, despite depression being one of the most common reasons people are prescribed medical cannabis. Wilson put it directly to NPR: “the routine use of medical cannabis products really should be rarely justified for the treatment of mental health disorders” (NPR, 2026).

There were exceptions, and they matter because they’re specific, not a blanket dismissal. Low-quality, low-certainty evidence supported modest benefit for cannabis use disorder itself, insomnia, tics and Tourette’s syndrome, and autism spectrum symptoms. “Low-certainty” here means the true effect could turn out to be substantially different once better trials exist — nearly half the included trials carried a high risk of bias, and the median trial size was just 31 participants. Across the board, cannabinoid treatment also came with real cost: about one additional patient in seven experienced an adverse event compared with placebo. If you’ve been handed a THC or CBD product as the answer to your anxiety or your PTSD, this is the paper that says the prescription got ahead of the evidence — the evidence for a handful of narrower uses is thin but real, and for the big-ticket reasons people actually seek cannabis out, it currently isn’t there.

Low-quality, low-certainty evidence supported modest benefit for cannabis use disorder itself, insomnia, tics and Tourette’s syndrome, and autism spectrum symptoms.

Legal weed isn’t the problem. The business model selling it is.

The third paper, again out of Freeman’s Bath group, is the one that should complicate both sides of the legalization debate rather than confirm what either side already believed. Looking at cannabis policy shifts worldwide from 2000 to 2025, the researchers found that commercialized for-profit legal markets — the U.S. and Canadian model — are associated with increased cannabis use, increased cannabis use disorder among adults, rising product potency, and more hospital admissions for psychosis specifically among people who also have cannabis use disorder (Freeman et al., 2026). Tightly regulated legal models — Uruguay’s pharmacy-and-social-club system with potency caps is the paper’s example — showed none of that pattern. Decriminalization in Europe, Africa, Oceania, and Asia showed little evidence of increased use or psychiatric harm at all. The paper is careful to note there’s no consistent evidence that policy change alone shifts the overall prevalence of psychotic disorders — it’s the hospitalizations among people who already have both conditions that rose. This isn’t an argument for or against legal cannabis as a concept. It’s evidence that how a market is structured — chasing daily heavy users with ever-higher-potency products, the way alcohol and tobacco companies have — determines most of the harm, independent of legality itself.

If you use every day, here’s what actually changes

None of this means casual or occasional use is what these papers are warning about — the psychosis signal is specifically about daily use compounding into cannabis use disorder, concentrated in a population that starts young. If that’s you, or you’re managing a family history of psychosis, the honest read is that frequency and potency are the levers that matter, not whether cannabis is legal where you live. And if you’ve been using cannabis products because someone told you they’d treat your anxiety, your depression, or your PTSD, the evidence right now says you were sold ahead of the science, not that you were imagining the relief. Both papers can be right. The people caught in the middle — using daily, told it’s medicine, worried about what it’s doing to their head — are exactly who this research package is for.

Related: Cannabinoids coverage · Science & Medicine