When Alex Berenson’s wife, a psychiatrist in New York responsible for evaluating criminals with mental illness, told him “they all smoke” marijuana, he wasn’t convinced. As a former journalist who spent years reporting on the prescription drug industry for The New York Times, he says he was skeptical that cannabis could be linked to mental illness.
“If all of these people are becoming temporarily psychotic after using THC, doesn’t that suggest that THC can do bad things to the brain?”
But after investigating, Berenson grew increasingly certain about this connection and wrote a book that he hopes will be a “bullhorn” to warn of the risks of marijuana use. In “Tell Your Children: The Truth About Marijuana, Mental Illness, and Violence,” he argues that cannabis use can cause both temporary psychosis and chronic psychotic disorders like schizophrenia. According to Berenson, the former is the kind of break that sends someone to the hospital for a few hours or days, while “an unlucky minority of users” will go on to suffer the permanent impairment of the latter. And because it can cause psychosis, Berenson contends that marijuana use can also be linked to violence — something he claims is underreported and “people are embarrassed to talk about.”
Berenson, who is also a bestselling novelist, deliberately chose not to focus on less serious mental illnesses like social anxiety disorder: “There is some evidence that cannabis is linked to those, but because they are so widespread — and measuring and counting them so difficult — causality is very hard to determine.” But he maintains that today’s highly powerful marijuana can have deadly serious consequences for some people. And those under the age of 25, Berenson says, are particularly at risk since THC, the main psychoactive compound in marijuana, can be “toxic to the developing brain.”
For this installment of the Undark Five, I spoke with Berenson about the science behind marijuana use, how the drug has become so much more potent than it was 50 years ago, and what research is still needed. I also asked him about a growing chorus of critics — many with some academic pedigree — who say the inferences Berenson draws from hard-to-interpret data on marijuana use, mental illness, and violence are misleading.
Our conversation has been edited for length and clarity.
Undark: In the book, you offer examples of how marijuana can be linked to mental illness. But what is the evidence that it’s a causal relationship?
Alex Berenson: Science is about building evidence. It’s about blocks of evidence. No one disputes that occasional use of marijuana by people over 25 is generally safe. But when you challenge and dechallenge [or administer, then withdraw] people with THC, you can provoke temporary psychotic episodes, even in healthy people. These episodes are not rare — hundreds of times a day in the United States, heavy marijuana users are brought to emergency rooms with psychotic symptoms. It’s a known phenomenon.
So then the question becomes: is there a relationship between temporary psychotic episodes and permanent psychosis? You would reasonably expect that there is, but you need studies. We now have those. So that’s the second layer of evidence. People ask which way the causation runs. We now have studies that show that people who don’t have pre-existing symptoms of psychosis also have higher rates of schizophrenia and other kinds of psychosis when they use. Even in healthy adults.
If you ask me what is the single strongest piece of evidence that marijuana can cause psychosis, it’s that healthy people who use marijuana — frequently enough that it’s notable — become paranoid or psychotic after smoking. And if all of these people are becoming temporarily psychotic after using THC, doesn’t that suggest that THC can do bad things to the brain?
UD: A 2017 report from the National Academies of Sciences, Engineering, and Medicine (NASEM) states that “the relationship between cannabis use and cannabis use disorder, and psychoses, may be multidirectional and complex.” What if the link works in the opposite way — that those with psychosis are drawn to marijuana? And since psychosis affects a small percentage of the population, why did you choose to focus on it?
“This is not a rare phenomenon — it just goes under the radar. People don’t want to talk about it.”
AB: Cannabis psychosis is a real phenomenon — it’s a diagnosis included in the World Health Organization’s 10th revision of the International Statistical Classification of Diseases and Related Health Problems. People smoke and become [temporarily] psychotic, often so that they wind up in the emergency room. And psychosis is not uncommon. Clinically meaningful, diagnosable psychosis — including bipolar disorder with psychosis, schizophrenia, depressive psychosis — probably affects one person in 25 over the course of their lifetimes. Many more people will have subclinical symptoms — episodes of unpleasant thoughts and paranoia, etc.
So this is not a rare phenomenon — it just goes under the radar. People don’t want to talk about it. There’s shame around this.
And psychosis can have terrible consequences — for sufferers, for their families, and for society. It is also closely linked to crime.
It is also important to understand that NASEM’s core finding in its 2017 report was this: “Cannabis use is likely to increase the risk of developing schizophrenia and other psychoses; the higher the use, the greater the risk.”
UD: How much more potent is the marijuana today than what was being smoked in the 60s?
AB: Most cannabis that grows in the wild is .5 to 1 percent THC — you have to cultivate this stuff to grow cannabis that has enough THC that you can get high easily from. In the U.S., in the 60s, it was maybe .5 percent to 2 percent THC. Yes, it can get you high, but you have to smoke a lot of it. It’s like a light beer. And that had more CBD, the primary non-psychoactive chemical. There’s some evidence that CBD actually mitigates the effect of THC. So people were smoking stuff that wasn’t as dangerous. And most people weren’t smoking as often, or smoking very much.
The drug today, sold in legal states, is extremely high in THC. Flower cannabis, herbal cannabis, the bud you see when you’re walking through the dispensary — those might be 20 to 25 percent THC, with very little CBD. Essentially, they’re genetically modified products. And some people are smoking extract, which is essentially pure THC. So people today are smoking something between 10 and 50 times as strong as the stuff they were smoking 40 years ago.
When potency rises, psychiatric side effects rise. And legalization has seemed to create a class of people who are very heavy smokers. A lot of people who smoked a little in college in 1985, they really don’t have any idea what this stuff is now.
UD: A New York magazine piece argued that “Berenson’s fishiest move is his choice of 2014 as a baseline year” for changes in homicide rates in states that legalized marijuana. Elsewhere, you’ve been charged with wrongfully peddling in “mental health stigma by suggesting that people with psychotic disorders are violent” while some studies show they aren’t. How do you respond?
“I didn’t cherry pick. I counted every aggravated assault and homicide in the four first states to legalize in 2013, and again in 2017, the most recent year for which data is available.”
AB: The charges of cherry-picking in the New York magazine article come from Mark Kleiman, a drug expert and a professor of public policy at New York University.
But here’s what Kleiman himself said in the introduction to the 2016 edition of a book he co-authored, “Marijuana Legalization: What Everyone Needs to Know:”
“Four states have replaced their prohibition policies with commercial legalization, giving us for the first time actual evidence about some of the early consequences of letting companies supply marijuana to every adult.”
That’s why 2014 was so important. That’s why using 2013 as the base year is logical and fair. And that’s why I used it. I didn’t cherry pick. I counted every aggravated assault and homicide in the four first states to legalize in 2013, and again in 2017, the most recent year for which data is available. That is the most definitive dataset possible — the whole bushel.
The definitive meta-analysis for schizophrenia and violence appeared in 2009. It examined 20 earlier studies over a 40-year-period and showed that people with schizophrenia have a five times greater risk of violence and a 20 times greater risk of homicide.
Because people with schizophrenia make up less than 1 percent of the population, they commit less than 10 percent of homicides, but on an individual basis they are much more likely to be violent than healthy people.
Excess risk occurs when they are having their first break, when they are not being treated with antipsychotics, and when they are using cannabis and other illegal drugs. Pretending otherwise distorts reality — and may even endanger their family members, who are often their victims and may be unaware of the severity of these risks.
UD: If you were addressing a group of scientists and philanthropists, what areas of research would you tell them need the most attention?
AB: Ethically, you can’t give 10,000 people cannabis and say “let’s see what the psychiatric outcomes are.” But we can take a sample of, say, 1,000 murders and 10,000 aggravated assaults, and see how many times people used cannabis, or how many times they used cannabis immediately prior to the crime. And really try to see if the violence connection is as strong as it seems to be.
We also need research into the emerging pattern of adults who seemed healthy who are getting really bad psychosis after using cannabis. We need to know how often that’s happening. That’s another hard one — you have to just look at cases. But the idea that there are people who may not realize that they are at risk for this, and wind up with terrible psychotic disorders that are untreatable — that’s scary, and we should find that out.
Hope Reese is a writer and editor in Louisville, Kentucky. Her writing has appeared in Undark, The Atlantic, The Boston Globe, The Chicago Tribune, Playboy, Vox, and other publications.