Dr. Dost Öngür, chief of the Psychotic Disorders Division at Harvard-affiliated McLean Hospital, recently presented a sweeping slide-talk on complex current issues in treating psychosis, but it was one simple line of his that most stayed with me:
"I tell my students: Smoke all the pot you want — after you're 30."
That very evening, I parroted it to my adolescent children, explaining that Öngür, as an expert on psychosis at a leading psychiatric hospital, was referring specifically to the risk of developing schizophrenia. (Subtext: "When you were little, infections scared me most. Now it's mental illness.")
Virtually all American schoolchildren are told that starting alcohol or drugs early could be bad for their brains. But Öngür's warning to his medical students stems specifically from a body of research that has been accumulating since the 1980s, suggesting that heavy marijuana use early on -- mainly in the teen years, but also into the 20s — is linked to a higher risk of psychosis.
A review paper in the peer-reviewed journal Biological Psychiatry this April summed up 10 long-term studies to date, most with sample sizes in the thousands, and concluded: "Overall, evidence from epidemiologic studies provides strong enough evidence to warrant a public health message that cannabis use can increase the risk of psychotic disorders."
A few past studies have failed to find a causal link, and the April paper's authors say that more research is needed. In particular, uncertainty remains on the pivotal question of causation: Can marijuana use actually cause psychosis? Or could the use be an effect, a symptom, of brewing mental illness?
That question remains controversial. And the vast majority of young people who use marijuana will not develop psychosis at all.
Still, this statistic was enough for me to caution my kids: Öngür cites estimates that for a teen who smokes pot daily, the risk of schizophrenia rises from about 1 percent to about 3 percent.
Öngür is concerned about legalization, but says his central aim is not to influence the impending votes in Massachusetts and elsewhere. Rather, it's to get out a public health message at a time when societal attitudes toward marijuana are shifting rapidly.
He sees many young patients with major mental illness who smoked a lot of pot beforehand, he says, "but that's not necessarily an argument to vote for or against any given policy proposal."
"It’s not just about having picked up pot but about how much pot you’re smoking and at what age."Dr. Dost Öngür
Our conversation, edited:
So when you tell your students to use all the pot they want after age 30, what is your thinking?
The debate here is not about whether cannabis use is risky for everyone at all ages, but specifically with psychotic disorders — like schizophrenia and some kinds of bipolar disorder — the concern is about young people. In particular, it’s about early teenagers, between the ages of 12 and 15 — that’s the highest risk period — and 15 to 18 is the next-highest risk period. The concern is that people who are smoking daily at those ages are at higher risk of developing these major mental disorders later in life.
So not at that time but later?
Correct. There has been a lot of debate about this, and clearly more research is needed, but the best evidence is that adolescents who are smoking daily in those early to mid-teenage years are at risk of developing schizophrenia or a related condition later in life, typically between the ages of 18 and 25.
Some of the studies that have been done involve face-to-face interviews with adolescents, and as best as anybody can tell, these are typically developing adolescents, they’re just smoking a lot of pot. You come back 10 years later and see who has developed something like schizophrenia, and what you find is that the kids who were smoking cannabis daily are at threefold higher risk for developing these disorders.
And they were definitely typically developing teens before their use started? They didn’t have precursors of schizophrenia?
This is what the debate is all about. Of course, nobody can really tell, since we don’t have a special microscope that looks into people’s minds. But as best we can tell in face-to-face interviews, these are kids that seem to be developing typically.
The definitive study to prove or disprove this causative relationship would be to take a large number of adolescents, force half of them to smoke pot daily, and prevent the other half from smoking any pot at all, and do that for about 10 years. Once they’re reaching their early- to mid-20s you would know who’s developed schizophrenia and who hasn’t, and then you would be able to tell. That’s of course never going to be done, and short of that experiment, we’re stuck with this evidence: that kids who are not saying they’re shy or anxious, they’re not suspicious or paranoid, they just enjoy smoking pot a lot — that then, later, they develop a psychotic illness.
What if it’s not daily use?
In the studies, there is in fact a dose-response relationship. Kids who are smoking daily are at the highest risk. Kids smoking on weekends but not every day are at somewhat lower risk. And kids who are not smoking at all are at the lowest risk of all. That’s additional evidence that there is this relationship. It’s not just about having picked up pot but about how much pot you’re smoking and at what age.
And in fact, saying ‘Smoke all you want when you’re 30 — you’re not going to have this kind of problem’ is exactly about this: There’s a risk window, and how much you smoke in that risk window seems to determine how high your risk is going to be.
What’s the extent of the evidence?
This is another area where there’s been debate. Many of the earlier studies were simply cross-sectional, that people who develop schizophrenia said they smoked a lot of pot when they were younger. And then the question always had been: Well, what if those people were somehow unusual to begin with, they had abnormal experiences, and they were drawn to pot because it did something for them, because they were a little different?
The more recent studies have been prospective [subjects are followed for a period of time after signing up] and they’ve been accumulating, and at this point we’re talking about a solid evidence base. So some of the leading medical journals, including Lancet, for example, have actually written editorials taking the position that there's a relationship between cannabis use in the teenage years and subsequent risk of psychotic illness.
And the biggest counterargument is the chicken-and-egg idea that there could be some other cause for both marijuana use and psychotic disorders?
Right. The kids who happen to like smoking pot more than the average teenager, maybe there’s something about them. It’s not necessarily pre-psychotic in the sense we think of it these days, that they were at risk, more paranoid. But maybe there’s something else that predisposes you both to smoking a lot of pot in your teenage years and to psychotic disorders.
It’s possible; it’s very difficult to disprove that hypothesis, but the fact remains that those are the kids who are at risk. So if you’re starting from a harm-reduction point of view, you want to intervene with the kids who’ve already declared themselves to be at risk. And if you could actually remove that risk factor, it stands to reason that the outcomes would be better.
Though there’s also the argument that you’d have to prevent several thousand kids from smoking marijuana to prevent one case of psychosis, given the current numbers.
That’s probably one of the reasons why we don’t have a major public health campaign already happening out there. Because psychotic disorders are not all that common. We’re talking about, depending on how you count, something like 1 to 3 percent of the population gets diagnosed with a psychotic disorder like schizophrenia and related conditions or the psychotic forms of bipolar disorder. And these kids who smoke daily in their early- to mid-teenage years are at a roughly threefold higher risk. So we’re talking about the risk going from 1 percent of the population to 3 percent. That means 97 percent of the kids who are smoking every day when they’re in the 12-15 age range are actually not going to develop a major mental illness.
So the math is such that even if you prevented every teenager and adolescent from smoking marijuana, you’re not going to make as huge a public health impact as some other kinds of prevention can, such as smoking cessation and lung cancer.
But with cannabis and the psychotic disorders, the thing to consider is that these major mental illnesses are not afterthoughts, they’re not minor problems, they’re a big burden for people afflicted by them and their families. They tend to be relapsing and remitting but typically lifelong. They’re not something you have once and then feel better and you get on with your life. Many of our patients are not able to live independently, hold jobs, have meaningful relationships. So the cost to the individual is incalculable.
Also, many of the people who have these disorders are supported by disability income, public health insurance, state mental health systems provide housing for them, and other opportunities. When you add it all up, some have calculated numbers as high as $200 billion a year in the U.S. that goes toward supporting and caring for people with psychotic disorders. So a meaningful reduction in those numbers by prevention of cannabis use in teenagers would still actually have a significant impact on society.
One paper I read noted that it might be the biggest 'modifiable risk' if you want to prevent psychotic disorders…
That’s right. We talk about multiple risk factors in psychotic disorders. There’s no doubt that these disorders tend to run in families. So having a family history, and presumably a genetic predisposition that goes with that, puts you at risk. But that’s not a modifiable risk factor. There’s not a whole lot we can do about that. But this [marijuana use] is one thing where early intervention could make an impact on the trajectory of these individuals.
Since you mentioned genetics — is the genetic testing ready for prime-time? If you want to be able to tell your kid, ‘Look at your genes — you better not smoke marijuana’?
There are studies out there linking certain genetic markers to the risk of developing psychosis if you smoked a lot of marijuana in your teenage years. So some links have been identified, but to answer your question, it’s not ready for prime-time. Those studies need to be replicated, and in much larger samples. At the moment, they really are just intriguing leads.
What’s the best guess at this point about the mechanism — how might marijuana increase risk of psychosis?
It seems to be about the excitatory and inhibitory signals in the brain. The way nerve cells communicate with one another, there’s usually two kinds of signals: one where one nerve cell makes it more likely for the next nerve cell to be active, that’s an excitatory signal, it’s like the gas pedal in your car. The other kind is an inhibitory signal, where the next nerve cell is less likely to be active, like the brake pedal in your car. And cannabis, in the brain, impacts both systems, and it changes the balance of excitation to inhibition. That’s probably what’s causing the abnormalities in information processing and how we think and how we behave.
And there would be a ‘window of risk’ because the brain is still developing in adolescence?
Right — the way the brain develops, there are many systems that are plastic, are changeable, and it depends on environmental inputs how the brain’s final organization is going to work. If you stimulate what’s called the endocannabinoid system in an unnatural way with cannabis long enough, it seems to alter the balance permanently, and cause these problems that eventually manifest as schizophrenia or related conditions.
The Biological Psychiatry review of past studies in April says the findings are consistent enough for a public health message that says teen marijuana use can raise your risk of psychotic disorders, but there have been some studies that did not find a link.
Yes, but the preponderance of evidence is in the direction that there is a relationship here. There have been occasional studies that don’t find it, but in general, high-quality, large, well-done studies have supported this idea. When you take all the evidence together, the meta-analyses say that the relationship is there.
What do you see in your division, your practice?
People like me who work with patients with psychotic disorders were not surprised to see this scientific literature develop. This is not something that’s novel to us. We know that cannabis and psychosis go hand in hand.
The average patient, in our experience, is somebody who did in fact start smoking pot as a teenager, usually not teenager as in 18 or 19, but more like 13, 14, 15. They smoked for several years, things were OK at first, but as the teen years went on, they became more withdrawn, their grades started dropping. And then they started developing unusual preoccupations, concerns about conspiracies or paranoid feelings about being in harm’s way somehow, until those thoughts and feelings actually culminated in what we call a psychotic episode, and the person comes to psychiatric care, and the diagnosis of a psychotic disorder is actually made.
Once the diagnosis has been made, there’s also very strong evidence that continued cannabis use worsens existing symptoms and is associated with future episodes. So there’s a relationship in people who already have schizophrenia or related conditions: If they are smoking a lot of pot, that is actually putting them at risk for more trouble. You can’t say it’s causative, but in the group of people who already have a diagnosis, cannabis use is associated with more symptoms and more episodes.
There’s this notion of a psychotic ‘break’ with reality. I was imagining a young teen smoking a lot of pot and then having a ‘break’, but it sounds like what you’re describing is that the break comes much later…
It can. It can go either way, which is more evidence that what cannabis is doing is not intoxication in this case. It’s not that you smoke and then you get upset or paranoid. It’s actually triggering something that culminates in psychosis eventually.
We see it both ways: In some kids, the pot-smoking reaches a crescendo — they’re smoking multiple times a day every day for months, and then they become psychotic — in that case it would be easier to argue that the pot’s messing with your head and you’re getting psychotic because you’re smoking so much pot. But there are other youngsters we see who were smoking a lot and then they knew something was wrong and they stopped it, and then some weeks later the psychosis starts to emerge more fully.
So this is a process that unfolds over time, and pot-smoking has a clear role in it, but it’s not a direct relationship. The parallel would be: If you drink too much alcohol and you have a blackout, it’s clearly related to the effects of the alcohol, but nobody would have a blackout a week or two after being drunk.
And acute psychosis is just different?
Right, that’s another thing I make clear to our medical students and residents, that we’re not talking about what is sometimes called 'secondary' psychosis — secondary to another factor, in this case substance use, that induces a psychotic illness. Rather, we’re talking about pot smoking as a risk factor for a psychotic illness.
So it’s a risk factor, but it’s not the same thing as ‘you smoke it and you get psychotic.’ Many, many people smoke a ton and they never get psychotic. So it’s simply a risk factor, it’s not a direct secondary relationship.
So what do you actually say to your medical students and psychiatry residents?
To make the point to them that we’re not talking about a generic negative effect of pot smoking but rather a very specific risk window, I say, 'Once you’ve passed that risk window, you can smoke all the pot you want and you’re not going to get psychotic from it, you’re not going to get schizophrenia from it.'
Most of them are in their mid to late 20s or even 30, and I say, 'For you guys, this is now safe, you can smoke all the pot you want from the point of view of schizophrenia.' It’s the teenagers we worry about, and that’s where we should be reducing exposure to cannabis.