Cannabis and schizophrenia: Scare stories are not policy arguments.

“Some” isn’t much of an argument. You need “How many?”

The Wall Street Journal has an op-ed by a Yale psychiatry resident warning that the legalization of cannabis will lead to increased disability due to schizophrenia. That’s based on a couple of studies showing that cannabis use, especially early heavy use, is a statistical risk factor for a subsequent schizophrenia diagnosis.

If I had a young friend with a family history of schizophrenia or who had experienced schizophrenic symptoms, I’d advise that person to stay away from cannabis. Why take unnecessary chances? But the evidence of an actual causal link is fairly underwhelming; it’s very hard to tell whether early cannabis use might reflect attempts at self-medication for pre-clinical symptoms rather than being an actual precipitating cause.

At the population level, we have what seems to me like strong negative evidence on the question whether increasing the availability of cannabis will lead to a measurable increase in the number of people with disabling levels of schizophrenia. Those in pre-Boomer and early Boomer birth cohorts in the U.S. – anyone born before about 1952 – had essentially zero experience with cannabis before the age of 18. But that changed rapidly. More than 10% of the high-school seniors of 1979 – roughly speaking, the birth cohort of 1961 – were daily or near-daily pot-smokers. Then the prevalence of heavy adolescent use fell sharply for a little more than a decade, reaching its trough around 1992, and has rebounded since.

Yet the rate of schizophrenia diagnosis shows no corresponding cohort-to-cohort swings. (Nor, for that matter, between high- and low-cannabis-prevalence areas within the U.S. or cross-nationally.) That doesn’t mean there aren’t individual cases in which cannabis precipitates a first psychotic break, or that there aren’t some people with schizophrenia whose disease course is worsened by pot-smoking. (It’s also possible, of course, that cannabis provides valuable symptomatic relief for others, but no doubt our vigilant Institutional Review Boards will protect us from ever learning about that phenomenon, if it exists.)

But those results do help put an upper bound on the number of additional schizophrenia diagnoses we need to fear as the result of the increase in cannabis use that would result from legalization and the resulting changes in availability, price, and attitudes. And that upper bound is fairly low: low enough to keep it off the list of the top ten reasons for or against legalization. Note, for example, that people put in jail or prison are at risk of severe damage, especially if while inside they become victims of physical or sexual assault. Some commit suicide. So the mental-health costs of arresting 650,000 people a year, and holding 30,000 or more prisoners at any one time, for cannabis offenses – costs that would be largely, though not entirely, abolished by legalization – might easily match or exceed the mental-health costs of increased exposure to cannabis.

The author of the WSJ piece solemnly announces, “The claim that marijuana is medically harmless is false.” No sh*t, Sherlock! Nothing is harmless. It’s always a question of counting harms, weighing them against one another, and comparing them to benefits. And we should do that not only when embarking on “social experiments” (i.e., making changes) but also when continuing a high-cost and potentially unsustainable status quo policy.

The costs of cannabis prohibition are large (including $35 billion a year in criminal income), and its capacity to keep consumption in check appears to be breaking down. That’s not a reason to plunge wildly into legalization on the libertarian model, but it is reason enough consider, soberly, the options around legal availability. Mere unquantified viewing-with-alarm (about schizophrenia, or workplace impairment, or intoxicated driving, or increased use by adolescents, or increased substance abuse disorder) no longer counts as a valuable contribution to the debate, any more than mindless sloganeering about “The failure of the War on Drugs.”

Some people will get hurt as a result of legalization; some people are getting hurt now by prohibition. The question before us is, “What policy would minimize total damage, net of the benefits of responsible use?” Continued prohibition in some form – at least the prohibition of commerce – might turn out to be the answer to that question; at least, Jonathan Caulkins and Keith Humphreys both think so, and they’re two of the most thoughtful and knowledgeable people around on this issue.

But being “against legalization” does not by itself name a policy position. No one I know has a serious proposal to put the genie back in the bottle, reversing the trend toward more cannabis use, and heavier use, that started around 2003 and seems to be accelerating. So it’s time to try some innovation. Who knows? We might be able to construct a licit market, and norms of responsile use, that would stop the progression toward more potent and less CBD-buffered, and thus probably riskier, cannabis. And then we should evaluate the results of those innovations with as much cool detachment as we can summon up: not to “prove” that one team of culture warriors or the other was right, but to consider what to try next. That’s the way grown-ups make policy.

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact:

39 thoughts on “Cannabis and schizophrenia: Scare stories are not policy arguments.”

  1. Well, speaking of health consequences of the use of marijuana, the current American Journal of Medicine has an epidemiologic study of current marijuana and insulin/glucose metabolism, finding that current users of marijuana had lower insulin resistance than non-users, and also had smaller waist circumferences. The authors note that animal studies have suggested that cannabis may reduce insulin resistance and make mice less prone to obesity.

    This does not mean that everyone with Type 2 diabetes should smoke marijuana, but it does mean that there is a crying need for more basic and clinical research into the health effects of an increasingly available drug.
    Op-ed pieces by psychiatry residents in the Wall Street Journal will not do.

  2. Just a reminder that some of us don’t care about the harms or the benefits of legalization. Imprisoning people for pot possession or sale isn’t wrong because the mental-health costs of doing so exceed the mental-health costs of increased exposure to cannabis. It is wrong because it is IMMORAL.

    1. Gee, I’m sure Mark needed a reminder that some of his commentators feel that way. I’m sure it slipped his mind.

    2. Most people would agree that adultery is immoral and therefore wrong, yet we don’t jail adulterers. The law should have no say in deciding which immoral acts should be punished, especially when no one else is harmed (as is the case with cannabis use).

    3. Henry – you are spot on. Religions threaten the very existence of humanity. That doesn’t justify religious prohibition! When you make inevitable, consensual human activities illegal (like religion or drug use), you criminalize being human. That, by definition, is a holocaust.

      1. People aren’t imprisoned for pot possession. Get real.

        And stop ignoring all of the evidence that is mounting against marijuana. It may not be en vogue to do so, however.

  3. Just a reminder of why I dropped my thirty odd year long sub to the Wall Street Journal a few months ago.

    Too much dishonesty on the editorial page now creeping into the news sections.

    The New York Times has its issues but as a replacement sub it is superior in reader satisfaction (at least mine).

  4. First, I’d like to say, Hi, it’s me again. Second, there are now two reasons for my returning and actually posting. Having read the entry above, I agree with your take on incidence of schizophrenia and cannabis use, but for one observation I have from sixteen years as a clinical psychologist. People with a serious thought disorder are experiencing symptoms LONG before those symptoms become evident. Even to their family and closest friends. Many schizophrenics self medicate with cannabis. So it is likely that their cannabis use to self medicate will pre date the appearance of obvious symptoms. Schizophrenia is a complex disease, with many different manifestations and a complex and difficult to sort out set of individual symptoms. It is difficult to accurately diagnose, and there are many different form it can take. So simply observing that the use of cannabis occurred before the onset of a clinical diagnosis is not only insufficient for a cause effect assignation, it is impossible to say definitively whether the cannabis use occurred before or after the patient began having distress.
    I understand you feel there are harms associated with cannabis use. “The author of the WSJ piece solemnly announces, “The claim that marijuana is medically harmless is false.” No sh*t, Sherlock! Nothing is harmless.” I have heard that in the past. Just seems a little unsupported. Usually, I will hedge, and take the position that perhaps cannabis use is not for everybody. But I would love it if somebody could actually point out some “harms”. Seventy years of intensive and (for the last forty) government supported research have failed to conclusively find any negative medical effects. The major “harms” associated with cannabis use all seem to be caused by it’s prohibition.
    Second, and the primary reason for this post is to get your reaction to this.
    Perhaps you’ve seen it, perhaps you have not. But I’m interested in your take. After all, economics are an important consideration in the crafting of policy.

    1. Economics? That?? Thanks, I need a laugh today.
      As to the harms of cannabis, we have a chapter on them in Marijuana Legalization.

      1. I should have guessed you would avoid making any substantive comment. Eventually you’ll be asked about that one by someone who you will have to answer, in a forum that may be public. Sure hope your response is more polished then. BTW, your chapter?, I read that…It also failed.

        1. Along the same lines, when is Mark going to prove to my satisfaction that aliens didn’t land at Roswell? When? Someday he’ll have to tell the truth…some day…until then I will keep the flame of truth alive…only I can see it but it’s real…so real…

      2. The question before us is, “What policy would minimize total damage, net of the benefits of responsible use?” Continued prohibition in some form – at least the prohibition of commerce – might turn out to be the answer to that question; at least, Jonathan Caulkins and Keith Humphreys both think so, and they’re two of the most thoughtful and knowledgeable people around on this issue.

        To quote a renowned deep-thinker’s robustly detailed and devastatingly convincing rebuttal: “Thanks, I need a laugh today.”

  5. If one or more genetic components are responsible for reactions to marijuana use by schizophrenics in those whose total symptomology has not yet emerged, such as the gene involved in dopamine signaling that codes for a protein called RAC-alpha serine/threonine-protein kinase (Akt1), then one resolution of the problem will lie in early detection of an Akt1 mutation, rather than brute force legal methods. The good news is that new technologies are making such testing simpler and cheaper to perform.

  6. Who will get hurt as a result of legalization?

    The same people who were being hurt from cannabis use before legalization? Or will there be a new cohort? Who will these people be? Law abiding people who hitherto abstained because of the arbitrary legal status but whom will suddenly become victims of cannabis legalization?

    1. Are you sure you meant to use question marks there? You sound like you didn’t.

  7. I am hoping you can speak more on what you feel are the dangers of less CBD-buffered cannabis in the marketplace. How does it modulate THC in a way that minimizes “risk”? Thank you from a first-time commenter! Always dig your blog.

    1. We know much less about cannabinoid pharmacology than we should, or than we would if the federal government weren’t hostile to the relevant research. It appears that CBD buffers the effects of THC in two distinct ways. It competes for CB1 receptor sites, thus acting as a partial antagonist. And it also (in ways not fully understood) seems to relieve anxiety and protect against panic attacks and dysphoria.

    2. I suffer from schizoaffective disorder, have a blog called Schizophrenic Confidential, and my latest post was JUST ON Cannabidiol. It’s really interesting stuff. (The CBD of cannabis, not referring to my blog, although i’d also like to think so).

      Personal story, it actually was pot that brought about my mental illness hearing voices. However, marijuana did not cause my schizophrenic spectrum disorder. People who revert into psychosis after smoking pot (and some time after the high when they’re sober) have to have a highly uncommon genetic predisposition to undergo psychoses. I.E. we produce a surplus of Dopamine transmission in our brains, the THC in cannabis increases dopamine transmission EVEN FURTHER and there you have it. You’re gone into No Man’s Land. Marijuana acts more like a catalyst, not a match.

      I am still pro-legalization. It won’t increase the rate of people developing schizophrenic spectrum disorders. Just because myself, my fellow schizophrenics, and addicts can’t enjoy marijuana doesn’t mean the large majority of the population shouldn’t.

  8. The author of the WSJ piece solemnly announces, “The claim that marijuana is medically harmless is false.” No sh*t, Sherlock! Nothing is harmless.

    Including, say, large servings of sugary beverages which I’m sure the WSJ has solemnly argued for the prohibition of as well.

  9. Until comparable studies for schizophrenia and the cases show one preferred substance for self medication over another, I’ve no idea what the point of the Yale psychiatry resident actually is. Is he saying that a more accessible Mj will trigger more schizophrenia? What is the use of Alcohol and approved Pharms in relation to “increased disability due to schizophrenia”? Is he saying increased instability doesn’t happen on its own or in conjunction with other inebriates… or for that matter, general behaviors?
    Is he saying that any entry to use by anything new is a instability progenitor? Then why are new drugs on the market eventually, almost all, taken back off the market due to side effects… are those not instability creators?
    It would seem that ONDCP funded studies like this one go as far as needed to cause doubt, and no more. When the Yale resident gets more comparable and has some comparisons, he might as well be discussing Mothers milk, aspirin or the notably toxic anti-psychotics on the market today.
    Psychotic breaks, or some people with schizophrenia will always happen regardless of anything consumed or activity so this piece leaves almost everything to the imagination. Sorry Yale, early heavy use of anything can cause these problems, go blame drinking coffee to early and too much also, as legalization is not about individuals whom will find something at hand for dealing with, or starting a psychotic break/schizophrenia.

  10. Jonathan Caulkins and Keith Humphreys both think so, and they’re two of the most thoughtful and knowledgeable people around on this issue

    At least based on his posts on this blog, I’ve seen no evidence that Keith Humphreys is thoughtful on drug policies. He basically seems to be have a pretty hard core moral objection to people taking this particular risk with their life (but not other risks that aren’t associated with the DFH’s), and that’s basically the exact opposite of thoughtful and the entire reason we have this irrational drug policy.

    Basically, the admission ticket to being “thoughtful” about drug policy is respect for people’s right to do things with their bodies that you may not like and which may be risky. If you don’t accept that, you aren’t thoughtful.

    1. So “thoughtfulness” means “agreeing with Dilan Esper”?
      Live and learn.

    2. You don’t have to accept libertarianism to be thoughtful, but you do have to be willing to give some weight to the benefits of cannabis use. Caulkins and Humphreys will not do so. They seem unaware of the fact that people are different, and even if they so no use for cannabis in their lives, there are many reasonable people who do.

  11. At the risk of being repetitious, we have nowhere near enough information about the pharmacology of marijuana to make sound policy regarding its use. There is at least a decent amount of information about its psychopharmacology but its effects on other organ systems are barely researched.

    1. We haven’t had scientifically sound cannabis policy since the 1930’s.

      Are you implying that incomplete scientific knowledge confers upon us a freedom to ignore the knowledge that we already have, or to postpone the action that it appears to demand?

      1. Does incomplete scientific knowledge confer upon us a freedom to ignore the knowledge that we already have, or to postpone the action that it appears to demand? As it happens, you can make and often must make policy on the basis of inadequate information, but there are systems for discussing the level of certainty you have in your policy recommendations.

        Developers of evidence-based medical treatment guidelines have a system for ranking the strength of evidence concerning various recommendations. The GRADE (Grading of Recommendations Assessment, Development, and Evaluation) process begins with a well-focused and explicit practical question which considers all relevant outcomes. Some outcomes are critical, others important, others useful to consider but not critically so. is the group’s website with tons of other information

        The level of evidence takes into consideration issues of study design, risk of bias inherent in the methodology of the studies, their consistency across settings and populations, the directness of the outcome measures to the outcomes which matter most in real life, and the precision (narrow versus wide confidence intervals) of the measured treatment effect sizes.

        Recommendations can be made and are made on the basis of less than high quality evidence, but the recommendations are open to revision as new evidence becomes available. Evidence-based policy making should be no different, being transparent about the level of evidence for or against measures to be taken with respect to marijuana.

        The critical thing about evidence-based guidelines is that they be honest about the level of evidence supporting their recommendations. It is very unlikely that future research is going to show that tobacco smoking is good for us after all. There is high quality evidence that the harms outweigh the benefits (a few of which do exist). The harms have been measured in studies which are well designed, have low risk of bias, are consistent across settings, with outcomes (like death) which are directly relevant to things that matter to people, and measured with considerable precision when all studies are considered. Tobacco policy-makers can have a high degree of confidence in the scientific basis of their recommendations. Marijuana policy-makers have a less massive evidence base on which to build their edifice. They need to be transparent about which outcomes are known with precision, how directly pertinent these known outcomes are to issues which matter to society, and which outcomes are known with great uncertainty.

        The outcome in the current American Journal of Medicine cited above (effects on lipid metabolism, abdominal obesity, and insulin resistance) is, in my opinion, important but probably not crucial in considering the totality of its health effects. There is considerable uncertainty about these effects, which future research may show to be large or may show to be non-existent. If marijuana improves glucose metabolism and decreases the risk of abdominal obesity, that would be a relevant fact to weigh in the scales when policy is made. We have insufficient information about its physiologic effects in this area.

        The GRADE process is NOT an ivory tower academic exercise! It is designed for the considerations of highly practical decision-making in the real world. By all means make policy on the basis of incomplete scientific knowledge; just give the public a transparent account of the level of evidence on which the policies are made. Mark is as well-qualified as anyone in the universe to make transparent recommendations and to distinguish between fact and opinion.

    2. So, Ed Whitney, what you are saying is there is not enough evidence to demonstrate that cannabis does not belong in schedule 1 of the CSA, the category of drugs with no medical use and that are too dangerous/ addictive?

      Astounding, considering that when cannabis was summarily banned in 1937, the state of science was woefully inadequate to prove that such a drastic measure was necessary, let alone beneficial. Are you saying lawmakers had sufficient evidence to make this drastic move then, but not now, in 2013? The historical record clearly shows that the prohibition on cannabis was never based on any real science, let alone concern for public health or safety. In fact, and this can easily be verified, the ban was based entirely on racial bigotry against people of color, and on maintaining a monopoly on paper production using trees, shutting out the more efficient hemp. Anyone who would have science-based policy must first call for an end to cannabis prohibition, since that law is based on as much science as were the Salem witch trials. Once we clean the slate, then we can decide what laws and regulations regarding cannabis, if any, are needed. America prospered and grew just fine for 160 years while cannabis and hemp were legal and widely used. The world has used cannabis and hemp safely with no need for laws against it for some 5000+ years. Only in the last 70 years have we tried to ban this vital substance, receiving as our reward: organized crime, corrupt militarized police, a prison nation, decreased safety and privacy, diminished public health, obstacles to free science, and it goes on…
      Those who support prohibition uphold a legacy of racism against blacks and Mexicans, a legacy of ignorance of science and basic facts.

      1. There is moderate to high quality evidence that marijuana has some accepted medical uses and high quality evidence that it is safe when used under medical supervision. Therefore, there is high quality evidence that marijuana should be reclassified under the CSA.

        There is very low quality evidence that it reduces insulin resistance and has a possible role in the treatment of Type 2 diabetes.

        There is high quality evidence that Dr. Andrew Weil has access to the best goddamn weed on the face of the earth.

        Policy makers should make strong recommendations that clinical trials of marijuana for a variety of indications be sponsored by the government.

  12. This was published in the Wall Street Journal. The Wall Street Journal is owned by Rupert Murdock. Anything published by Rupert Murdoch is a lie. The Wall Street Journal is a lie from cover to cover. Anyone who cites anything published by Rupert Murdoch is citing lies.

    The study may or may not be a total lie, but if it was published in The Wall Street Journal, you may be sure that it is a lie.

    1. The WSJ is there to tell rich people how to get richer. If it takes too many liberties with reality in its reporting (this does not apply to the editorial page), rich people who make decisions based on its contents will lose money and stop reading it. Is there evidence that this has happened?

  13. WSJ is there to appeal to the world view of those who buy ads in it. That is where the money comes from, not in telling readers how to make money (after all, WSJ has lots of ads from actively managed mutual funds, rarely runs stories about the well-known problems with such funds).

    As Steve Forbes has said many times his father told him you don’t get rich from taking the advice, you get rich from selling it.

    IMHO, the Journal’s vaunted news pages are a whole lot less reliable than they used to be.

  14. To me it seems like nonsense that pot causes schizophrenia. Here is why. Pot has been used by mankind in certain regions for ages. Numerous studies have shown that schizophrenia affect a consistent 1-2 percent of the population throughout the world. If it were true that pot caused schizophrenia wouldn’t schizophrenia spike in regions where pot use is common? It does not. In fact many of the studies also state that it may be the case that schizos seek pot to relive the symptoms of their affliction. I can buy that.

  15. ^ my apologies. I misread. I will agree, however, that there is so little research. I fully support researching. I have no doubt the research will be in favor of legalization.

Comments are closed.