It’ll be busier.
Those words cut to the heart of the last defense put up by doubters of the link between cannabis and mental illness. As other evidence has piled up, they’ve offered it more and more.
If cannabis causes psychosis, why haven’t psychosis rates risen in countries along with cannabis use?
At first glance, the rebuttal seems persuasive.
It isn’t.
Scientists and health agencies track disease rates in two ways. They count the number of people who have received a new diagnosis in a given period—usually a year—as well as the overall number who have the illness. The first figure is called the incidence of the disease. The second is its prevalence.
For mental illness in general, and schizophrenia in particular, the United States can’t count either number. The federal government doesn’t track incidence by requiring that doctors report new diagnoses of serious mental illness to a central database. It doesn’t track prevalence through a national registry counting patients. In fact, in November 2017, the National Institute of Mental Health suddenly cut its estimates for the prevalence of schizophrenia in the United States from 1.1 percent of adults to 0.3 percent.
The new figure implied that instead of having almost 3 million adults with schizophrenia, the United States instead had fewer than 1 million. The agency quietly put the new, lower estimate on its website without asking for comment—until Dr. E. Fuller Torrey, a longtime schizophrenia researcher, wrote a scathing opinion piece about the change in Psychiatric Times. “Although he has been the director of the National Institute of Mental Health (NIMH) for less than two years, Dr. Joshua Gordon has made 2 million individuals with schizophrenia disappear,” Torrey wrote. “NIMH has not said where they went but officially, they no longer exist. This is a remarkable accomplishment.”
Torrey pointed out that the new estimate was based on a survey more than 15 years old that specifically excluded people in hospitals, prisons, or on the street—all places where many people with schizophrenia live. Northern European nations make serious efforts to count cases, Torrey wrote. “Compared with these countries, the data available in the US on the prevalence of schizophrenia are equivalent to the data available in some developing countries.”
In his response, the NIMH director admitted that the United States didn’t know how many Americans had schizophrenia. The institute had “settled on” the 0.3 percent estimate because it believed the old figure was too high. But it wasn’t sure the lower figure “necessarily reflects the full picture” either, he wrote. In fact, the institute has now once again updated its website. It now uses a range—from 0.25 percent to 0.64 percent.
Even if the United States did a better job counting cases, psychiatrists agree schizophrenia is a problematic diagnosis. It covers symptoms, not causes. It is a little like diagnosing a patient as “crawling” without knowing if he’s on the ground because of a broken leg or a stroke. It overlaps with bipolar disorder, with psychosis, depressive psychosis, schizoaffective disorder, and schizophreniform syndrome. All are treated with antipsychotic drugs.
The common thread in these illnesses is psychosis. But counting psychosis cases is even harder than tracking people with schizophrenia, especially without a national registry. A patient might be diagnosed with bipolar disorder, rediagnosed as having schizophrenia, then called schizophreniform a few years later. That’s one patient counted three times.
Making matters even trickier, some big studies show that schizophrenia fell modestly in industrialized countries from 1960 to 1990. No one is sure why, or even if the decrease was real. Some researchers think better prenatal nutrition and obstetric care led to a genuine decline. Others say that as psychiatric hospitals closed, and patients returned to the community for treatment, doctors became more reluctant to stigmatize them as schizophrenic.
“With respect to psychosis incidence overall, there’s no good evidence that rates are changing,” said Dr. James Kirkbride, an epidemiologist at University College London who led a 2012 study that examined schizophrenia rates in England between 1950 and 2009. Schizophrenia diagnoses dropped, but other psychosis diagnoses rose. More recently, rates of drug-induced psychosis have risen, but from a low baseline, Kirkbride said.
But everyone agrees that the growth in marijuana use from the late 1960s through about 1980 couldn’t have done much to psychosis rates. Most cannabis then simply didn’t contain enough THC to matter except to very frequent users. Those people may have been at extra risk. But they accounted for a tiny fraction of the overall population—not enough to affect overall rates of mental illness.
But beginning in the 1990s, cannabis became far more potent. The number of smokers rose too. Since 2000, the United States and Canada have seen further increases, while use has risen in some European countries and been flat or down in others. But potency has continued to increase everywhere.
Marijuana users generally start smoking between 14 and 19; first-time psychotic breaks most often occur from 19 to 24 for men, 21 to 27 for women. In other words, almost no one develops a permanent psychotic illness the first time he uses marijuana—or even after a few months. The gap between when people start smoking and when they break averages six years, according to a 2016 paper in the Australian & New Zealand Journal of Psychiatry that examined previous research. The Finnish paper showing that almost half of cannabis psychosis diagnoses convert to schizophrenia within eight years is more evidence of the time lag. A problem that seems temporary becomes permanent.
The time lag is crucial. It implies that the 1990s increase in cannabis use—and the increase in potency that began then and continues today—wouldn’t have immediately affected psychosis rates. Instead, if marijuana slowly drives some people into permanent psychosis, rates of schizophrenia and other psychotic disorders might have trended higher in the 2000s, with the increase visible after 2010.
That trend is exactly what some research has found.
Not surprisingly, the strongest evidence comes from northern Europe, where countries can track mental illness more precisely.
The first red flag came from one of the world’s most remote areas. In 2016, Finnish researchers reported that rates of psychotic disorders had nearly doubled between 1993 and 2013 in people born in Oulu and Lapland provinces, the country’s two northernmost regions.
The study began more than fifty years ago. First, scientists tracked everyone born in the area in 1966, more than twelve thousand babies. Twenty years later, scientists began another study, tracking every infant born in the area in 1986.
When they compared the generations, the researchers found that the people born in 1966 had a 1 percent chance of being diagnosed with schizophrenia or other psychoses by age 27. The people born in 1986 had a 1.9 percent chance of a psychotic disorder by age 27—a near-doubling of the risk.
The researchers said they did not know what had driven the increase, though they noted it might be due to improvements in Finland’s mental health care system. They published their findings in the journal Epidemiology and Psychiatric Sciences in March 2016. The paper received little attention.
By American standards, Finland is relatively drug-free. But the rise in psychosis followed a sharp increase in cannabis use there. Surveys show that use doubled among Finnish teens and young adults between 1992 and 2002 and remained at the higher levels through 2010.
A separate 2009 study of schizophrenia across Finland found a similar trend. New admissions rose from the early 1990s through 2006. The fraction of admissions that included a diagnosis of drug abuse rose from 1 percent to almost 10 percent after 2000. (Cannabis use was not broken out separately from other drugs, but the use of drugs like cocaine is very rare in Finland.)
In other words, marijuana use, psychosis diagnoses generally, and psychosis diagnoses that were specifically drug-related all rose side by side in Finland.
On the other side of the Baltic Sea, a similarly troubling trend took place. Researchers in Denmark recently examined that country’s mental illness registry to see if the incidence of schizophrenia had changed between 2000 and 2012.
They found a striking increase. Schizophrenia diagnoses rose about 30 percent. The Danish researchers published their findings in the journal Schizophrenia Research in October 2016. Like their Finnish counterparts, the Danish researchers couldn’t pin down a reason for the increase in schizophrenia.
But like Finland, Denmark saw a big increase in cannabis use during the 1990s. A 1994 government survey showed that fewer than 4 percent of people aged 16 to 24 had used the drug in the last month. By 2000, the rate doubled to 8 percent. It stayed there through 2010.
In other words, the Danish experience exactly paralleled what happened in Finland.
So, what about the United States?
Again, the short answer is no one knows. Considering that the agency that oversees mental health research here can’t even decide whether 1 million or 3 million Americans have schizophrenia, marijuana would have to cause a massive increase in psychosis before anyone would notice.
And unlike Europe, the United States has had two separate waves of growth in marijuana use—first in the 1990s, then more recently, since 2006. The more recent period is particularly striking because adults, not teens, have had the biggest increases in cannabis consumption. Among adolescents 12 to 17, use has risen only in the last couple of years. In fact, dangerous behavior generally has fallen in teens since 2000, although the trend may now be starting to reverse. (Researchers are not sure why. One theory is that tobacco is the ultimate gateway drug, and the sharp decline in teen cigarette smoking has decreased all kinds of risky behavior—though the more recent rise in vaping may be undoing that progress.)
Put all those facts together, along with the time lag, and even if adolescent cannabis use sharply raises the risk of psychosis, proof of a population-wide increase across the United States won’t exist for years.
But some evidence has already surfaced. Researchers for Kaiser Permanente and another health insurer recently looked at rates of newly diagnosed psychosis among patients from 2007 through 2012. The insurers cover almost 5 million people aged 15 to 59, the age group the researchers studied. And they are in areas with heavy cannabis use—Colorado, the Pacific Northwest, and California.
The researchers examined electronic medical records across hospitals, mental health clinics, and doctors’ offices. They then double-checked a sample of the diagnoses to make sure they were accurate. They specifically excluded cases of dementia-related psychosis, which becomes more common after 50.
Using this more comprehensive approach, they found psychosis rates higher than many previous reports. In their study, which was published in May 2017 in the journal Psychiatric Services, they reported that 86 out of 100,000 people aged 15 to 29 received a new diagnosis of schizophrenia, bipolar disorder with psychosis, or other psychosis each year. Based on that figure, the average person has about a 1.3 percent risk of being diagnosed with a psychotic disorder before age 30.
An even bigger surprise came when they looked at adults from age 30 to 59. Psychiatrists expect schizophrenia and bipolar disorder to become obvious by the late twenties. But the researchers found that 46 out of 100,000 people in the over-30 group received a new psychosis diagnosis every year. “The high proportion of true cases in this sample presenting after age 30 contrasts with conventional wisdom that first onset of psychosis typically occurs at younger ages,” Dr. Gregory Simon, the study’s lead author, wrote.
Translated nationally, the researchers’ findings suggested that 115,000 Americans aged 15 to 59 will develop a psychotic disorder every year, Simon wrote. Almost 3 percent of people will receive a psychotic diagnosis before age 60—1 person in 35. And that figure probably underestimates the true prevalence, because adults who are uninsured, covered by Medicaid, or in prison are much more likely to have psychosis than those functioning well enough to have private insurance like Kaiser.
(Kirkbride, the British epidemiologist, says the most comprehensive studies show that 4 percent is a reasonable figure for the prevalence of adult psychosis, though only a minority of those people will be diagnosed with schizophrenia.)
Of course, that estimate doesn’t mean all those adults are actively psychotic at any time. Some people recover completely. Others get by with antipsychotics. Still others die relatively young—an ugly but real reason that the overall prevalence of people with psychosis can’t be determined just by adding up new cases every year.
Still, the Kaiser study suggests that psychosis is a quiet epidemic. If its figures are accurate, Americans under 50 are nearly half as likely to be diagnosed with psychotic disorders as cancer.
The paper did not directly examine whether marijuana had led to any psychotic diagnoses. But it hints at a cannabis-psychosis link in a couple of ways.
First, the fact that new cases are high in marijuana-friendly states may only be coincidence, or a result of the comprehensive counting techniques the researchers used. But it might also be evidence that cannabis was causing more mental illness in Colorado and other high-use states as early as 2007. Another hint might be if the number of cases trended higher over the six-year period. (Unfortunately, Dr. Simon turned down an interview request, and the paper doesn’t break out data by year.)
The second effect is more speculative. Even cannabis advocates acknowledge the drug can speed the development of schizophrenia. But they argue those people would have been psychotic anyway. The cases are accelerated, not caused, they say.
But if marijuana causes psychosis in otherwise healthy people, its neurotoxic effects might not be limited to teenagers and young adults. In other words, marijuana may cause schizophrenia to develop more quickly in young people who are already on the precipice—but it may also slowly cause psychosis in adults who are outside the usual window for the disorder.
That result is precisely what the Psychiatric Services paper found. Many adults over 30 are becoming newly psychotic every year.
What is causing all those breaks?
Dr. Erik Messamore believes marijuana is a big part of the answer, and he has a theory as to exactly why prolonged exposure to THC might be neurotoxic.
Messamore, a psychiatrist in Ohio, runs a clinic specializing in schizophrenia treatment. He focuses on helping patients having their first break, or “first-episode psychosis.” Evidence shows that the more quickly people are treated, the more quickly they recover. “If you can get people into remission from psychosis within a year, they have much better outcomes,” he said.
Messamore began his career with a PhD in pharmacology, intending to research drug development. But he decided he wanted to help people more directly and went to medical school. He is relatively optimistic about his profession. Antipsychotics don’t work for everyone, and their side effects are real, but they help many patients, he says. “Half or more of people with schizophrenia by the modern definition will have meaningful recovery,” he said—though the disease may still impair their lives.
After completing his residency in psychiatry at the Oregon Health & Science University in Portland, Messamore worked at an outpatient clinic there, treating patients who had common mental health issues like depression or insomnia.
Along the way he noticed an odd trend.
Messamore’s patients were often in their thirties and forties, with no history of psychosis. Many used cannabis. “Oregon is an exceptionally green state,” he said. A patient would fail to show for an appointment, then another—and Messamore would learn he had been hospitalized with a psychotic break.
“There’d be no hint whatsoever of any abnormalities,” he said. Psychiatrists learn to look for subtle signs of psychosis that other people don’t notice: jumbled speech patterns, emotional flatness. Messamore’s outpatients didn’t have those, much less more obvious problems such as delusions or hallucinations.
Until, suddenly, they did.
The experience wasn’t common, but it happened enough to make an impression. Messamore was also working at a state psychiatric hospital, where he saw devastating cases of schizophrenia, patients whom even the strongest antipsychotics couldn’t help. They too had something in common. “I had this hard, solid core of patients for whom almost nothing worked,” he said. “And among this group they’d almost all been heavy [marijuana] users from their early teens.”
In 2013, Messamore moved to Ohio to practice at the Sibcy House, a high-end private psychiatric medical center associated with the University of Cincinnati. The Sibcy patients generally had private insurance. They weren’t poor. They had families, jobs, and sometimes advanced degrees. They also had severe psychosis.
A surprising number of them seemed to have used only cannabis and no other drugs before their breaks. The disease they’d developed looked like schizophrenia, but it had developed later—and their prognosis seemed to be worse. Their delusions and paranoia hardly responded to antipsychotics.
How often had he seen these cases, I asked?
“Once a month.”
After about two years, Messamore left Sibcy to become an associate professor at Northeast Ohio Medical University and run his clinic. But he is convinced that long-term marijuana use can cause mental illness. He has developed a theory as to how cannabis exposure might harm the brain.
Anandamide, the natural cannabinoid that triggers the CB1 receptor, is known to play a role in the brain’s inflammatory processes. Inflammation is a normal response when the body is repairing damage or fighting infection. But prolonged inflammation damages everything from blood vessels to nerve cells. In the brain, chronic inflammation is connected with dementia and other degeneration.
Messamore points to studies showing that long-term marijuana use may be associated with an increase in the enzymes that break down anandamide. That part of the theory is biologically plausible. THC turns on the CB1 receptor, overwhelming the cannabinoid system. The brain has no way to know that THC and not anandamide has caused the extra activity. It might try to bring itself back to normal by reducing its anandamide levels.
But ultimately, that response could cause a shortage of anandamide—thus leaving the brain open to long-term inflammation and increasing the risk of psychosis.
Messamore’s theory is unproven and could be wrong. But other psychiatrists I asked said it was worthy of further study. Messamore notes that his theory suggests that patients whose schizophrenia is caused by cannabis may have more inflammation in their brains than those with other types of schizophrenia. Inflammation leaves specific markers, so neuroscientists may be able to test that part of the theory.
Even if his theory is wrong, Messamore believes researchers will eventually discover a biological mechanism that links cannabis and psychosis. He is disappointed that other psychiatrists have not spoken more aggressively about marijuana’s dangers.
“They’re very, very slow to act in the public interest,” he said. “This is something that is an extremely important public health problem, and it is going to be of disastrous consequence to a number of people.”
It is going to be, and it already is.
• • •
It’ll be busier, Scott Simpson had told me about his Denver Health psychiatric ER.
He was only half-right. It—along with its cousins all over the country—is busier already. And marijuana is part of the reason.
The number of people arriving at emergency rooms with marijuana-related problems has soared in the last decade. In 2014, the most recent year for which full data is available, emergency rooms saw more than 1.1 million cases that included a diagnosis of marijuana abuse or dependence—up from fewer than 400,000 in 2006.
Cases involving marijuana rose far faster than those involving cocaine—and even faster than those involving opiates. In 2006, cannabis cases were less common than the other two drugs. By 2014, they were more common than opiates and twice as common as cocaine. Only alcohol, which is far more widely used, contributed to more emergency visits.
Those figures come from a giant government database called HCUP, the Healthcare Cost and Utilization Project. The federal Agency for Healthcare Research and Quality manages HCUP, which uses billing records to track emergency room visits and inpatient hospital stays. Emergency rooms now receive about 140 million patients a year—though that figure doesn’t mean 140 million different Americans are going every year, since some people wind up in the emergency room over and over.
HCUP created the emergency room database in 2006. It uses billing records to track more than one-fifth of ER trips nationally. In 2014, it covered 31 million visits to 950 emergency rooms in 34 states—enough to paint a nationally representative picture. The database contains information about everything from the procedures patients undergo to whether they are admitted, discharged, or die. It includes all the diagnoses that patients are given, not just the main one.
HCUP makes limited data, like case counts for different diagnoses, available for free on its website at www.hcup-us.ahrq.gov/. When I found the database, the trends in the free data immediately struck me.
Besides the huge increase in marijuana use disorder, the database showed a big rise in psychosis-related cases. In 2006, emergency rooms saw 553,000 people with a primary diagnosis of schizophrenia, bipolar disorder with psychosis, or other psychosis. By 2014, that number had risen almost 50 percent, to 810,000. Including cases where psychosis was either a primary or secondary diagnosis, the increase was even faster, from 1.26 million in 2006 to almost 2.1 million in 2014.
I wanted to be sure the increase was real, since the American health care system puts a premium on “upcoding.” That word is a fancy name for not-quite-insurance fraud—making the most severe diagnosis possible to maximize insurance reimbursements. But upcoding long predates 2006. And the Agency for Healthcare Research and Quality, which has its own psychiatric epidemiologist on staff, told me that it doesn’t believe that upcoding or diagnostic changes caused the increase.
For whatever reason, lots more people are showing up at emergency rooms with psychosis—and with marijuana addiction.
But the crucial question was whether the two problems were occurring together. I needed to count cases where a patient had a primary diagnosis of psychosis and a secondary diagnosis of marijuana abuse or dependence. That combination is a proxy for cannabis psychosis, especially in cases where cannabis is the only drug a patient is abusing.
As far as I could tell, no one had conducted that research. I couldn’t find it in any journal. Once again, I turned to my friend Sanford Gordon for help.
Gordon was not thrilled, especially once he realized how much work coding his analysis program to grind through these massive datasets would be. But he said he would.
Fortunately, 2006 is a good starting point for the emergency department data, because adult marijuana use in the United States began rising about then. If cannabis does cause a significant amount of psychosis, the change in the overlapping figures should be obvious.
Through the spring of 2018, we wrangled the databases for an answer. Actually, Gordon wrangled. I tried not to get in the way. Finally, on a sunny Friday afternoon in June, Gordon emailed me the answers.
In 2006, about 30,000 emergency room patients had a primary diagnosis of psychosis and a secondary marijuana use disorder. Eight years later, that number had almost tripled, to nearly 90,000. Put another way, every day in 2014 almost 250 people showed up at emergency rooms all over the United States with psychosis and marijuana dependence. They accounted for more than 10 percent of all the cases of primary psychosis in emergency rooms. Most of those patients had problems only with cannabis, not other drugs, our analysis found.
Those figures understate the problem. Gordon and I used a conservative definition of psychosis, excluding many people with bipolar disorder. If those patients are included, the number is closer to 170,000. And the HCUP data doesn’t provide a way to track occasional users. It includes only patients whose marijuana use was so severe that physicians diagnosed them with abuse or dependence. (The HCUP dataset does include a separate diagnosis for “drug-induced psychosis,” which also rose rapidly. But it’s impossible to know from that diagnostic code whether the drug was marijuana or something else, so I didn’t include it.)
Marijuana disorder was also associated with more severe psychosis—as measured by being hospitalized instead of released following emergency treatment. Psychotic patients with a marijuana sub-diagnosis were about twice as likely to wind up hospitalized as those who didn’t have one.
Finally, the emergency room data showed that marijuana dependency was linked to opiate and cocaine addiction. The number of emergency room patients who had a primary diagnosis of opiate addiction and a secondary diagnosis of marijuana use disorder nearly tripled between 2006 and 2014—more evidence that the theory that marijuana can help people stop using opiates is dangerously wrong.
For cocaine, the link was even stronger. In 2014, almost 15 percent of people who had a primary diagnosis of cocaine dependence or abuse also had a marijuana use problem.
Without reviewing charts or interviewing patients, it is impossible to confirm every case. But the trend is sharp and obvious. Marijuana use and potency rose between 2006 and 2014. Marijuana-related emergency room visits—for psychosis or any other reason—rose even faster.
HCUP’s inpatient data shows similar trends over a longer period, since the group has collected data on inpatients since 1993.
The number of hospital inpatients has remained roughly flat at around 35 million for the last 25 years. But the number of those patients with a primary or secondary diagnosis of cannabis abuse has soared. It rose from 96,000 in 1993 to 300,000 in 2006 to more than 600,000 in 2014. Those numbers don’t include people treated at psychiatric hospitals or Veterans Administration hospitals.
Seeing that trend led me to ask Gordon if he would run the same screen on the inpatient data as he already had on the emergency room visits. The long pause that followed let me know that I had pretty much used up my favors. But he agreed.
Gordon found that the situation for inpatients was similar to that for emergency room visitors. The number of cases where a patient had a primary psychotic diagnosis and a secondary diagnosis of cannabis abuse rose 70 percent between 2006 and 2014. By then, people who had a cannabis abuse diagnosis accounted for more than 15 percent of all the psychosis cases that American hospitals treated—far more than any other drug. Again, in most of those cases, marijuana was the only drug being abused.
The annual federal survey on drug use and mental health offers more evidence that cannabis is now leading to increases in mental illness significant enough to be seen at the national level.
Every year, the Substance Abuse and Mental Health Services Administration sponsors a survey of drug use and mental illness among 70,000 Americans.
The survey does not try to count different types of disorders, lumping them into one umbrella category of “serious mental illness.” So it cannot be used to make estimates of the prevalence of schizophrenia or other psychotic disorders. Still, without a federal registry, the study is the best source of national data on mental illness.
SAMHSA released data from the 2017 survey in September. It found a marked rise in serious mental illness in the United States too, especially among adults 18 to 25, the heaviest users of cannabis.
According to the study, about 2.5 million young adults met the criteria for serious mental illness in 2017, a rise of more than 25 percent from the previous year and double the rate in 2008. Suicidal thoughts and suicide attempts—which both depression and psychosis can trigger—soared, too.
And though the survey does not break out psychotic disorders from other forms of serious mental illness, it showed that inpatient treatment among young adults has risen even more rapidly. In 2017, 220,000 young adults received inpatient treatment, compared to 178,000 in 2016 and 97,000 in 2008. People with psychotic disorders are much more likely to be hospitalized than those with depression, so that increase is still more evidence that psychosis is increasing among young American adults.
What’s especially striking is that adolescents 12 to 17 don’t show the same increases in cannabis use, or in severe mental illness.
Further, about 10 percent of all cannabis-using adults over 18 met the criteria for serious mental illness in 2017, and another 25 percent met the criteria for less severe conditions. In contrast, among nonusing adults, fewer than 4 percent met the criteria for serious mental illness, and 13 percent for other mental illness.
I know I’ve just given you an all-you-can-eat buffet of numbers. They can all blend together, and I understand if you skimmed that previous section. But I decided to include them because cannabis advocates so often insist that real-world proof that psychosis is increasing doesn’t exist. They’re wrong. No one goes to an emergency room for a fun Friday night or is admitted to a hospital for a psychotic disorder on a whim.
Considering the relatively short time frame of the data, these are huge increases. More than 1 million times in 2014, Americans were told that their marijuana use was a diagnosable medical problem. Not by loser friends who don’t know how to have a good time. Not by cranky parents who just can’t chill. By emergency room doctors.
Imagine if marijuana were actually dangerous.
The epidemic isn’t coming. It’s here.
Why hasn’t anyone noticed?
In part because cannabis psychosis is the drug equivalent of a car accident, a semiprivate incident that doesn’t make news unless someone is injured. People rarely broadcast that they have been hospitalized for psychotic episodes. When celebrities are involved, news sometimes trickles out. Even then the privacy of mental health information means that the story is often murky.
• • •
During 2017, the Saturday Night Live comedian Pete Davidson disclosed he’d suffered repeated “mental breakdowns” after years of smoking marijuana. Davidson had been a vocal cannabis supporter. In a 2016 interview with Rolling Stone, he had called himself a “pothead.”
Then, in March 2017, he announced on Instagram that he had entered rehab. He elaborated in a podcast interview in September 2017, explaining that he had gone into treatment after months of quasi-psychotic episodes related to his cannabis use:
Around October, I would say, September, of last year, I started having these mental breakdowns, where I would freak out . . . blind rage . . .
I had no memory of it . . . I wouldn’t know what would happen until I broke something or after I came to, so I was under the assumption that maybe it’s the weed, you know, I never really did any other drugs . . .
He went on to explain that he could not stay sober after finishing treatment:
I get off weed . . . [and] I got out, and then I started smoking weed again . . . Two months go by, and I just snapped, I was smoking weed every day, I just like snapped, and I had a really bad mental breakdown . . .
Davidson didn’t elaborate on his “really bad” episode. But he said he had initially been diagnosed with bipolar disorder and suffered from anxiety and low-grade paranoia:
[I would] think everybody’s mad at me, everybody hates me, I’m gonna lose my friends, I’m gonna lose my girlfriend, my family hates me . . .
Ultimately, Davidson was told he had not cannabis psychosis or bipolar disorder but borderline personality disorder. That diagnosis is marked by manipulative behavior and resistance to treatment, neither of which seemed to be an issue for Davidson. On the podcast, he sounded puzzled and upset by his problems. He was taking medicine and going to therapy, he said. “This whole year has been a fucking nightmare.” (Through his agent, Davidson declined an interview request.)
Yet despite his efforts to quit and awareness of the problems marijuana had caused him, Davidson continued to use. “I smoke a lot of pot—it’s really hard for me to concentrate,” he said at a comedy show in Atlantic City in June 2018. He told the crowd that before coming onstage, “I got very, very high.”
At about the same time, he told another interviewer that marijuana was helping his anxiety and Crohn’s disease—a chronic disorder of the digestive tract. As for his previous concerns? “I found out I had a mental disorder. I thought I had a drug problem. It’s a completely different thing.”
• • •
Similarly, the nexus between Kanye West’s mental illness and his marijuana use briefly grabbed attention, in August 2017, when lawyers for West sued insurers over a claim related to concerts West canceled after a psychiatric hospitalization.
In November 2016, West walked offstage during a show in Sacramento after ranting at the crowd for several minutes. (In 2015, West had given a similarly incoherent speech at a music award ceremony, during which he said he had smoked “a little something” beforehand.)
Two days after quitting the Sacramento concert, West was hospitalized for psychiatric evaluation in Los Angeles. He was held for eight days for a “debilitating medical condition” and released under “full-time care and supervision,” according to the lawsuit he filed.
The incident led West to end his tour, which his representatives had insured through Lloyd’s of London—a group of insurance companies—for $10 million against cancellation. West’s lawyers asked Lloyd’s to pay the policy. But the insurers said they needed to investigate. In August 2017, West’s lawyers sued, alleging the insurers were unfairly blaming his marijuana use for the cancellation.
The insurers countersued. They said they could not comment on the specifics of West’s hospitalization out of respect for his privacy. But they noted the policy did not cover canceled dates due to “the possession or use of illegal drugs.”
In 2018, the two sides resolved the suits without disclosing terms. West went on to blame opiates for his 2016 breakdown. Then, in May 2018, he disclosed he had been recently diagnosed with bipolar disorder, which he called his “superpower.” By then the hospitalization and lawsuits had been written off as just another semicomic episode in West’s chaotic celebrity life. (“I hate being bipolar it’s awesome,” his new album cover proclaimed.) No one seemed to connect his diagnosis to his cannabis use.
• • •
So it goes. Marijuana, THC, and wax are smoked, vaped, ingested, dabbed. Hundreds of times a day in the United States—and that’s a low estimate based on a conservative reading of the HCUP data—psychosis follows.
A hospitalization here, an emergency room freak-out there, an involuntary commitment here. A diagnosis of schizophrenia here, cannabis psychosis there, bipolar disorder with psychosis here—though in that case psychiatrists will likely play down the “with psychosis” aspect to the patient. Bipolar disorder may be a superpower. Psychosis, not so much.
No one notices. Not with the United States now topping 70,000 overdose deaths a year. Not with marijuana’s backers shouting its benefits.
Not even when the blood flows.