MtF 2010: Conditional probability of dependence on cannabis

December 14, 2010

The Monitoring the Future epidemiological survey of drug use in the US has announced some of their updated data today. As always, you can go to their website and look at the figures yourself. The monographs with expanded data tables and summaries can be found here, the current year’s data will be published in June, I think, so until then you’ll have to be satisfied with the selected figures on the website.

One of the many things that I look at in these datasets forms part of the converging evidence on the conditional probability of dependence. These surveys do not assess dependence or abuse or any other DSM categorization of course. They cannot. However since these are broad scale, demographically representative data they can give us a look at both trends and the overall population in a way that is not possible otherwise. I’m not sure how you could do such a representative survey of clinically diagnosed conditions, for example.


The above figure (source pdf) presents the proportion of the sample that says that they have used marijuana on at least 25 of the 30 days prior to survey (I think this is their usual definition of “daily”) for 8th, 10th and 12th grade students. Roughly speaking, over the past half-decade about 5% of high school seniors and 3% of sophomores endorse daily consumption of marijuana.
I encourage you to think about this in the context of your children’s peer groups. They have an easy 100 kids per grade level in many urban / suburban school districts…in elementary school. Forty to fifty kids per class in highschool would not be unusual. One or two kids in each class period is getting high every day. Five kids of your elementary school children’s grade level will be getting high every day when they are seniors.
Every day.
We don’t know that these children are dependent on marijuana, as I mentioned at the top. We do not know that all of them meet criteria for abuse as defined under the Diagnostic and Statistical Manual. We certainly cannot assess the subjective nature of a drug “problem”.
But I think we should be able to think about what it means to be getting high on marijuana on at least 25 days out of each 30.
What about the more casual use of marijuana?


Well, something on the order of 35% of high school seniors and 28% of sophomores have tried marijuana at least once in the past year (pdf source; lifetime stats are in the monographs, btw). Not too informative, of course, since this is a very loose criterion. Still, it is one way of putting a denominator under the conditional-probability calculation.
We’re in the 10-14% ballpark. This is the approximate chance that if a high school student has so much as tried marijuana in the past year, s/he will be smoking daily.
Allowing for the fact that very likely not all the daily smokers will meet criteria for dependence on marijuana, these numbers align pretty well with other similar estimates which point to about 8-10% as the conditional probability of dependence on marijuana.
Related Reading:
Addiction Inbox on the MJ/cigarette trends
All About Addiction (Psychology Today) on MJ use and binge drinking


19 Responses to “MtF 2010: Conditional probability of dependence on cannabis”

  1. AMW Says:

    What happened after 1992?


  2. Lorax Says:

    @1 Clinton didn’t inhale.


  3. ex-hedgehog freak Says:

    Question: having viewed only this year’s figures in the link, and not having delved back into earlier publications, do you know if they plan to sort this data to look for subset trends (economic status, ethnicity, regional variations) – trends in some of these subgroups may or may not be surprising, but may tell us a lot more than the overall picture for educational purposes.


  4. Just Some Guy Says:

    Well, it’s the year Dr. Dre put out The Chronic. I’m assuming it’s his fault.


  5. becca Says:

    “Roughly speaking, over the past half-decade about 5% of high school seniors and 3% of sophomores endorse daily consumption of marijuana.”
    Wow, I knew some weird people. (probably closer to 50% of the folks I knew at sophomore age smoked daily… well, daily is perhaps crude. Most days in any given week is probably more accurate)


  6. DrugMonkey Says:

    #3, yes, the monographs have demographic breakdowns. And some of the tables are reproduced in single-page pdf format and linked on the webpage.
    see and look at “subgroup” links…


  7. herp n. derpington Says:

    it seems like you’re conflating a rise in use as evidence of the addictiveness of pot? perhaps i’m misinterpreting, but if i am right, i just want to ask this:
    among seniors, 35% have used recreationally, 5% regularly (daily). what does the data look like for something that is highly addictive (tobacco) and something not as addictive but still the number one drug in america (alcohol, and don’t say caffeine)?


  8. DrugMonkey Says:

    unfortunately the summarized data are treated a little bit differently. so for example this table gives lifetime and daily for cigarettes (the daily population is about ~25% as big as the lifetime).
    alcohol, 12th graders- any annual use is 65%, “been drunk” is 44%. any use in past 30 days is 40%. I had to dig through the 2009 data /2010 published monograph for fig 5-5b for 30 day prevalence of daily use, ~3-4% going by the graph. So something like 4-5% as a comparable conditional-probability analysis.
    Point being that in both of these cases this particular analysis / inference matches well with other ways of estimating conditional probability from epidemiological data. Again, you cannot get at a firm estimate of meeting dependence criteria. All we are doing here is saying that daily use of alcohol, marijuana or cigarettes in a 12th grader is getting us within viewing distance of the dependent population.
    as with all science, converging inference strengthens our conclusions.


  9. DrugMonkey Says:

    it seems like you’re conflating a rise in use as evidence of the addictiveness of pot?
    I am suggesting that the population that uses marijuana on a daily basis is highly enriched (relative, for example, to the population of ever smoked in the past year, or ever-lifetime) in those that would meet criteria for dependence if they were evaluated clinically. These data do not demonstrate that, however. We’d need to rely on other studies to support that conclusion.


  10. WMDKitty Says:

    You’d also have to, er, “weed out” those who use daily for medical reasons.


  11. drdnc Says:

    If conditional probability means the likelihood of event B given that event A already happened, wouldn’t event A in this case be more accurately characterized as “a high school student has AT LEAST tried marijuana in the past year” rather than “so much as” tried it?


  12. DrugMonkey Says:

    WMDKitty, I would be highly fascinated to learn what conditions are justifying discernible numbers of high school seniors having medical marijuana prescriptions.


  13. WMDKitty Says:

    I’m not saying a large (or even statistically relevant) number of high school seniors have, or would have, a prescription. However, a number of seniors are of legal age, and some of them could, conceivably, have a prescription. Hypothetically, you could have a high school senior who was diagnosed with cancer his sophomore year, and has approval for medical usage.
    I’m simply saying that the number of pediatric medical users, though very, very low, is not “zero”.
    I, for one, was investigating the possibility at that age, as effective medication for spasticity, anxiety, and pain.


  14. Steve Clay Says:

    Someone somewhere is digging up the data to assert that “the conditional probability of dependence on video games” is alarmingly high. Imagine wanting to increase the frequency of a newly found enjoyable activity!
    Kidding aside, I appreciate that you’re trying to say something interesting of this near useless data, and I do find youth cannabis use alarming, too, but it’s very frustrating that this is the best data we can get. And after decades of tinkering we still have no control over availability to kids, and are completely dependent on “sending messages” and “perceptions of risk” to reduce use.
    Care to comment on all the usage numbers that are going down? The drop in binge/regular alcohol use seems like a pretty damn good thing.


  15. islamisohbet Says:

    Bulgaristan’dan göç ettiği öğrenilen Derya A.’yı yakalamak için çalışma başlattı.
    Olay, Acıbadem Faikbey Mescidi Sokak Numara 26′da bulunan binada saat 17.45 sıralarında meydana geldi. İddiaya göre, evde tek başına yaşayan Neriman Şit’e (80) bakıcılık yapan Derya A., henüz belirlenemeyen bir nedenle Şit’in boğazını kesti. Neriman Şit olay yerinde hayatını kaybetti. Zanlı Derya A., apartmanın önünde bulunan bir araca binerek olay yerinden kaçtı. Derya A.’nın çıplak ayakla binadan çıkıp otomobile bindiğini gören Şit’in komşuları, durumu polise bildirdi. Olay yerine gelen polis, kapısı açık bırakılan daireye girdiğinde Şit’in kanlar içerisindeki cesediyle karşılaştı. Yapılan ilk incelemede Şit’in boğazının kesildiği tespit edildi. Sağlık ekipleri de ceset üzerinde inceleme yaptı. Olay yerindeki in


  16. Please make clear that the graphs are on massively different scales. It is very easy not to notice.


  17. DuWayne Says:

    Ahh…I feel like I am back in saneland with you DM, thanks.
    @7 –
    No, DM is correlating the figures for “used in last year” with the figures for “use daily,” making no statement about the slope of either chart. I would note that he is also not making a statement about actual addiction or dependence either.
    @10 & 13 –
    Sorry, but that is a huge, stinking pile of crap. Pediatric medical cannabis use is an oxymoron. What we do know about the effects of cannabis on the developing brain would contraindict the use of cannabis for all but terminal conditions and the figures for pediatric medical use of cannabis for a terminal condition is near enough to zero to make it statistically invisible, much less insignificant.
    That is not to say that there aren’t pharmaceuticals being prescribed to children that are just as problematic, but when prescribed responsibly, they are treating conditions that warrant the use of potentially damaging drugs (disclosure – I have a child on such drugs, who should not be and am more than seriously bloody pissed about it). The problem with using cannabis in the contexts you are talking about, is that evidence would suggest that it is not only not effective, it can be and often is counterintuitive. And the mere fact that you might believe it actually helped is not the least bit compelling.
    There is a great deal of clinical evidence to suggest that using cannabis to deal with anxiety will generally exacerbate anxiety disorders. It tends to create cyclic dependency problems, wherein the anxiety without cannabis is worse than the anxiety was before the initial use. Moreover, cannabis use for anxiety often eventually causes panic attacks – note that this is not when cannabis is not present, the panic attacks will generally happen within about ten to thirty minutes after use, depending on the delivery mechanism, dosage and quality. Another problem is that episodic anxiety will often become more generalized with cannabis use.
    In the context of pain management, the pain relief is not commensurate to the damage cannabis does to a developing brain. I am not, mind you, singling out cannabis for it’s effects on the developing brain – alcohol and tobacco are also particularly bad for the developing brain. It’s just that we aren’t talking about their use in pediatric medicine. While opiates are also bad for the developing brain, it is very likely that the actual pain relief they afford is considerably more commensurate with the level of damage they cause.
    Pain management sucks balls, but while it might be argued that cannabis might be beneficial as a tool for pain management for adults, those arguments just don’t hold for pediatric use. Using cannabis regularly, before age 23-24 is a very bad idea. The development of the medial prefrontal cortex (MPFC) is still very active into the early twenties. Given the nature of how we suspect decision making is mediated in that part of the brain, it would seem likely that cannabis use could negatively effect the mediation of impulse control into adulthood.
    I would also note that that same region of the brain is also likely involved in our susceptibility to emotionally driven appeals. There is evidence that would suggest that the MPFC is critical to processing emotions in teh context of behavior and that an immature MPFC is not only associated with risk taking behaviors, but also with emotion driven decision making. This is probably why young men are particularly susceptible to recruitment into religious extremism, gangs and the military.
    While the longterm effects of cannabis use on these decision making processes is questionable, short term impact (defined by years)is exceedingly likely. Likewise, the impact of long term use that was initiated during MPFC development is a fascinating question and one that I am particularly interested in. Given that the vast majority of long term cannabis use that was initiated at an early age fits clinical criteria for addiction, it is right up my alley. I am particularly interested in the hypothesis that regular cannabis use during MPFC development might create a sort of feedback loop, supporting addiction and sustaining emotional immaturity.
    But all of that has little to do with your idiotic assertion about pediatric medical cannabis use, except as a reason why it is a really bad idea. Essentially the notion of medical cannabis use in children just provides a justification for the establishment of dangerous behavior patterns and a significant likelihood of delayed, if not entirely stunted neural development. As someone who engaged in just those sorts of behavior patterns and had their life rather fucked up because of it, comments like yours particularly piss me off.
    Adolescence is screwed about plenty enough by normal hormonal changes and the tail end of neural development. Throwing drugs that we know at the very least complicate neural development into the mix is irresponsible at best. Imbuing that drug use with a veneer of legitimacy is considerably worse than irresponsible, it is reprehensible. Kids are really rather good at making irresponsible decisions without that sort of “help.”


  18. DrugMonkey Says:

    Care to comment on all the usage numbers that are going down? The drop in binge/regular alcohol use seems like a pretty damn good thing.
    Agreed. I thought Dirk H did a good job in the cigarette data and Adi J discussed the binge drinking so I figured I’d focus on some other aspect.


  19. supratall Says:

    Someone somewhere is digging up the data to assert that “the conditional probability of dependence on video games” is alarmingly high. Imagine wanting to increase the frequency of a newly found enjoyable activity!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: