A Parent’s Guide to Cannabis and Teen Anxiety: What the Research Supports, What It Warns About, and What Actually Treats Adolescent Anxiety

The largest randomized trial of adolescent anxiety treatment, published in the New England Journal of Medicine, found that 80.7 percent of children improved on combined cognitive behavioral therapy and sertraline, compared with 23.7 percent on placebo. Effective treatment for this condition exists and is well characterized.

The closest thing to cannabinoid evidence in young people with anxiety is a 31-participant open-label trial of pharmaceutical cannabidiol in treatment-resistant cases, with no control group and adverse events in 80.6 percent of participants. That is a signal worth studying, not a basis for a parent to act on.

Anxiety disorders are among the most common psychiatric conditions in adolescence, and a teenager who has found something that reduces the feeling in the moment will reasonably conclude it is working. Distinguishing acute relief from treatment is the central task here, and it is not a semantic distinction.

This is a page about what a parent should do next, and the answer is the same regardless of how the conversation started: an evaluation by a clinician who treats adolescent anxiety, not a product decision.

The Child/Adolescent Anxiety Multimodal Study remains the reference trial. Published in the New England Journal of Medicine by John Walkup and colleagues, it randomized 488 children aged 7 to 17 with a primary diagnosis of separation anxiety disorder, generalized anxiety disorder, or social phobia to 14 sessions of cognitive behavioral therapy, sertraline at up to 200 mg per day, the combination, or placebo for 12 weeks.

The proportion rated very much or much improved was 80.7 percent for the combination, 59.7 percent for cognitive behavioral therapy alone, and 54.9 percent for sertraline alone, against 23.7 percent for placebo. All three active treatments beat placebo, and the combination beat both monotherapies. Adverse events including suicidal and homicidal ideation were no more frequent with sertraline than with placebo, and no child in the trial attempted suicide.

Those numbers are the reason this page opens here rather than with cannabis. A parent weighing options is not choosing between an unproven option and nothing. They are choosing between an unproven option and a treatment with a four-in-five response rate in the best available trial.

There is one study worth taking seriously, and its limitations are as important as its result. Maximus Berger and colleagues at Orygen in Melbourne enrolled 31 young people aged 12 to 25 with a diagnosed anxiety disorder who had not improved despite cognitive behavioral therapy, antidepressant medication, or both. All participants received add-on pharmaceutical cannabidiol on a fixed-flexible schedule titrated up to 800 mg per day for 12 weeks.

Mean scores on the Overall Anxiety Severity and Impairment Scale fell from 10.8 to 6.3, a 42.6 percent reduction. Depressive symptoms, clinician-rated severity, and functioning improved as well. Adverse events were reported in 25 of 31 participants, or 80.6 percent, including fatigue, low mood, and hot flushes or chills. There were no serious or unexpected adverse events.

This was an open-label trial with no control group. Everyone knew they were receiving cannabidiol, everyone was already engaged in specialist care, and anxiety scales improve substantially under those conditions even without an active drug. The authors said so themselves, concluding that randomized controlled trials are needed to confirm efficacy and longer-term safety. That conclusion is the honest one, and it is not a green light.

It is also worth noting what this trial was not. It was not cannabis. It was a pharmaceutical-grade cannabidiol preparation, dosed and monitored inside a youth mental health research service, in patients who had already failed standard treatment. None of those conditions describe a teenager buying a product.

Selective reporting is the main way this topic goes wrong, so the failures deserve equal space. A double-blind, randomized, placebo-controlled trial published in Psychopharmacology tested single oral doses of cannabidiol at 150 mg, 300 mg, or 600 mg against placebo in 61 healthy young adults, using a well-established carbon dioxide inhalation challenge that reliably provokes fear and panic symptoms.

There was no effect of condition on self-reported fear, on panic symptoms, or on heart rate. The authors described it as the first study to document that cannabidiol does not reduce fear reactivity in humans.

One negative challenge study does not settle whether cannabidiol helps clinical anxiety, and provoked fear in healthy volunteers is not the same construct as an anxiety disorder. It does complicate the simple story in which cannabidiol is a reliable anxiolytic, and it is the kind of result that rarely appears in consumer-facing articles about CBD and anxiety.

A systematic review and meta-analysis in JAMA Psychiatry by Gabriella Gobbi and colleagues pooled 11 longitudinal studies comprising 23,317 individuals, examining cannabis use before age 18 and outcomes in young adulthood between ages 18 and 32, with estimates adjusted for baseline symptoms.

The odds ratio for developing depression was 1.37. The odds ratio for suicidal ideation was 1.50, and for suicide attempt it was 3.46. The pooled odds ratio for anxiety was 1.18 and was not statistically significant, with a confidence interval from 0.84 to 1.67. That last result should be reported as it stands: this meta-analysis did not find a significant link between adolescent cannabis use and later anxiety, even though it did find one for depression and suicidality.

The psychosis literature is separate and stronger. A meta-analysis in Schizophrenia Bulletin covering 66,816 individuals found an odds ratio of 3.90 for schizophrenia and related outcomes among the heaviest cannabis users compared with non-users, with a consistent dose-response relationship. A multicentre case-control study across 11 sites published in The Lancet Psychiatry found daily cannabis use associated with a 3.2-fold odds of psychotic disorder, rising to 4.8-fold for daily use of high-potency products.

The authors of the depression meta-analysis framed the individual-level risk as moderate to low while emphasizing that the number of adolescents using cannabis makes the population burden significant. Both halves of that sentence are worth carrying.

A teenager who says cannabis is the only thing that calms them down is usually describing something accurate. Acute anxiolysis at some doses is a genuine pharmacologic effect, and dismissing the report damages the conversation without changing the behavior.

What complicates it is that regular use creates its own anxiety. A meta-analysis in JAMA Network Open covering 47 studies and 23,518 participants estimated pooled cannabis withdrawal syndrome prevalence at 47 percent among people with regular or dependent use, with daily use associated with higher prevalence. Irritability, restlessness, sleep disruption, and anxiety are core features, and they resolve with the next use.

That loop is indistinguishable from treatment while it is happening. Relief arrives, the relief is real, and the interval before the next dose becomes progressively less comfortable. A teenager experiencing that pattern will report with complete honesty that cannabis is treating their anxiety, and the pattern will still be getting worse.

This is why the clinical answer is not to argue about whether it helps. It is to get an accurate picture of how much, how often, and for how long, alongside a proper anxiety evaluation.

Book an evaluation with a clinician who treats adolescent anxiety, usually starting with the pediatrician or family physician and moving to child and adolescent psychiatry or psychology. Anxiety disorders in this age group are diagnosable and treatable, and the delay between recognizing the problem and starting treatment is the variable most within a family’s control.

Bring the cannabis information to that visit rather than treating it as a separate disciplinary matter. Frequency, route, product type, and how long it has been going on all change the clinical picture, and a teenager is more likely to be accurate if the framing is medical rather than punitive.

Do not source cannabis or cannabidiol products for a minor on the strength of the research above. The one supportive study used a pharmaceutical preparation, in a specialist research service, in patients who had already failed standard treatment, with no control group. Retail products differ in content and purity, cannabidiol has documented interactions with medications including some antiepileptics, and no product sold to consumers has been tested for adolescent anxiety.

If a clinician does consider a cannabinoid in a specific case, that decision belongs to a specialist who is managing the whole picture and can monitor for interactions and adverse effects. The route to that conversation runs through an evaluation, not around one.

The treatment evidence for adolescent anxiety is strong. A 488-participant randomized controlled trial with a placebo arm and clear separation between conditions is about as good as psychiatric treatment evidence gets in this age group.

The cannabinoid evidence in young people with anxiety is weak by design rather than by accident. One open-label trial with 31 participants and no control group cannot distinguish drug effect from expectation, regression to the mean, or the effect of being in a specialist service for 12 weeks.

Longitudinal studies of adolescent cannabis use cannot fully separate cannabis from what precedes it. Teenagers who use cannabis differ from those who do not in family history, adverse experiences, and existing symptoms, and adjustment for baseline symptoms reduces that problem without eliminating it.

The anxiety result in the depression meta-analysis was null and should be treated that way. Reporting the significant depression and suicidality findings while implying the anxiety finding went the same direction would be exactly the selective reading this page is arguing against.

None of this shows that cannabis causes anxiety disorders in adolescents, and the largest relevant meta-analysis specifically did not find a significant association with later anxiety. What it shows is elevated odds of depression and suicidality, and a separate and stronger dose-dependent association with psychotic outcomes.

The cannabidiol research also does not show that pharmaceutical cannabidiol is ineffective in youth anxiety. An uncontrolled trial with a large apparent improvement is a reason to run a randomized trial, and that trial has not been reported.

Adolescent anxiety and adolescent cannabis use overlap heavily in the same population, at the same developmental stage, during a period when brain systems that cannabinoids act on are still maturing. Untangling which came first is one of the hardest problems in developmental psychiatry and will not be solved by better cross-sectional data.

The more actionable point is that the treatment side of this question is not in doubt. Whatever the causal story turns out to be, adolescent anxiety disorders respond to treatments that have been tested against placebo in children, and access to those treatments is the rate-limiting step for most families.

I take the teenager’s report seriously. When a sixteen-year-old tells me cannabis is the only thing that stops the feeling, they are not lying and they are usually not exaggerating. Arguing with that observation is the fastest way to end the conversation.

What I try to get across is that the relief is real and the arrangement is unstable. Something that works within minutes and wears off within hours will, over months, reorganize a young person’s day around the interval between doses. That is not a moral claim. It is how short-acting relief behaves in a developing nervous system.

And I will say the thing I say to every parent who asks: I do not recommend cannabis for an adolescent with anxiety, and the evidence gives me no reason to. The treatments that work for this condition have been tested against placebo in children and they perform well. That is where the effort should go.

No randomized trial supports cannabis as a treatment for adolescent anxiety. The one supportive cannabinoid study in young people used pharmaceutical cannabidiol, in a specialist research service, without a control group, in patients who had already failed standard treatment. Meanwhile, cognitive behavioral therapy and sertraline produced response rates of 59.7 and 54.9 percent against 23.7 percent for placebo in a 488-child randomized trial, and 80.7 percent in combination. A worried parent’s next step is an evaluation, not a product.

Notice that the anxiety finding in the adolescent risk meta-analysis was null while the depression, suicidal ideation, and suicide attempt findings were not. Any source that presents adolescent cannabis use as a proven cause of later anxiety is overstating that literature, and any source that uses the null anxiety result to imply adolescent cannabis use is benign is ignoring the rest of the same table.