DrugFacts 2010 Repost: Recreational Drug Use in the Year After Initiation

November 12, 2010

This is Drug Facts Week, an effort of NIDA to promote understanding of the effects of recreational drugs. Although I’m slightly busy with other matters, I wanted to participate, partially, with a series of re-posts. This post originally appeared April 6, 2008; I was reminded of it by a tweet from A3Addiction.


This week’s fax from the Center for Substance Abuse Research at the University of Maryland touches on an issue of continual interest, namely the determination of “how addictive” different drugs of abuse may be. As I have mentioned a time or two (or three) before, I believe this is an area where the scientific research tends to skirt a key issue. From my perspective, this is one of the hardest questions to answer on the basis of the available human data and the animal models tend to drive right over the essential concepts.
This week’s CESAR fax (sign up here) reports rates of discontinuation, continued use without dependence and dependence for most major drugs of abuse. These data can help us to answer the question of “how addictive” are various recreational drugs.


The first measure which might answer a question of “how addictive” is the rate of discontinuation or non-use in the year after first trying a substance. On this scale, crack cocaine is “least risky” with 76% of individuals who tried it no longer taking it a year later. Inhalants (73%) and heroin (69%) also result in high quit rates. On this measure, alcohol (26% discontinuing- wow, this high?) is the “most risky” by some margin over the nearest competition (marijuana, 42%; pain relievers, 57%). This is a decent start. From a population point of view if a drug has properties (inherent, environmental or circumstantial) that result in high percentages of those who try them failing to continue use, it is “less addictive”.
I’m not going to get into the question of why people might try specific substances once or a few times and then not continue using because these data can’t illuminate the question. I’m sure the commenters will have fun proposing the likely testable hypotheses. And of course, “use” is not that congruent with “addiction” although, of course, discontinuation of a substance is a fantastic way to prevent addiction. So it is somewhat relevant to the human condition to consider this factor. And taking the canonical self-administration laboratory model of drug use as an example, it should be obvious that the major governing hypothesis is “with enough exposure, in the right pattern, all individuals will become dependent/addicted”. Charles “Bob” Schuster, who is one of the major figures behind the long history of self-administration, drugs-as-reinforcers laboratory animal models reiterated this concept in his Dews award lecture I recently attended. That pole of the drug-abuse research community is not backing down one bit, apparently.

CESAR041408-UseStatusOneYear.jpg

Notes: Data are from persons aged 12 or older responding to the 2004 to 2006 National Survey on Drug Use and Health who reportedinitiation of a substance 13 to 24 months prior to the interview. Pain reliever, sedative, tranquilizer, and stimulant use refers to nonmedical use of prescription-type drugs not prescribed for the respondent or used only for the experience or feeling they caused. Percentages may not sum to 100 due to rounding.

Source: Substance Abuse and Mental health Services Administration, “Substance Use and Dependence Following Initiation of Alcohol or Illicit Drug Use,”The NSDUH Report, March 27, 2008. Available online at http://www.oas.samhsa.gov/2k8/newUseDepend/newUseDepend.cfm.

The second measure which might answer the “how addictive” question is the dependence rate. For many people, the idea of drug dependence is essential to the question of whether or not we should be concerned about recreational drug use. On this scale, the “most risky” substances are heroin (13% dependent) and crack cocaine (9%) with a second grouping of marijuana (6%), stimulants (5%) and powder cocaine (4%).
Now we move to the final measure which helps us to arrive at the crux of the matter. The percentage of the sample that continues to use a substance and yet does not meet criteria for dependence. Of course, a failure to meet dependence criteria does not mean that one is not at risk from continued substance use. Not at all. There are a host of other health risks that can result from continued drug use other than dependence per se. However, this non-dependent use measure helps us to fill in the gaps between those who discontinue use and those who become substance dependent. Here we find that 71% of those who initiated alcohol use in the past year continue to use alcohol without dependence. The next-closest substance is marijuana (52% still using without dependence) with almost all of the rest hitting the 34-40% range. Interesting, but we can go a bit further along the path.
The most interesting comparison to my eye is to take the population of drug-initiators who are still using a year later and compare the relative rates of dependent and non-dependent use. Heroin (13% dependent/17% nondependent) and crack cocaine (9%/15%) are the “most risky” by this measure. One way to express this might be to say that given continued use of the substance, 43% will become dependent on heroin, 38% on crack cocaine and 10% on marijuana. On the other end, only 4% of continued users will become dependent on alcohol.
The CESAR figure fits well with a favorite paper of mine to which I refer in discussions of relative “addictiveness”. It is a bit old, but presents some interesting data. The study population in the National Comorbidity Survey was:

“based on a stratified, multistage area probability sample of persons 15 to 54 years old in the noninstitutionalized civilian population in the 48 coterminous United States, including a representative sample of students living in campus group housing. …between September 14, 1990 and February 6, 1992.”

I’ve graphed some data presented in tabular form (Table 2 from the publication) and added one additional analysis. The data from the paper include the proportion of the sample that reported extramedical use of the various substances in his/her lifetime as well as the proportion that reported sufficient symptoms to meet criteria for dependence (including at any time, not just when surveyed). The estimate of life-time prevalence of use is quite familiar if you browsed any further into the Monitoring the Future data after reading this post which included one of the annual prevalence graphs. The dependence rate given extramedical use from Anthony, et al (1994) is quite similar to the data from the CESAR fax, above. Strikingly similar given that it is derived from a different sample and gated on the one-year interval after initiation instead of considering the whole lifetime.
To my eye, the most striking outcome of these types of data is the realization that it is only a minority of individuals who sample a drug that will become dependent. From the Anthony et al (1994) data, about 23% of the population who has ever sample heroin becomes dependent. As outlined above, only 43% of the population of heroin initiators become dependent in the first year.

Anthony94-Use-DependenceRate.jpg

Anthony JC, Warner LA, Kessler RC.
“Comparative Epidemiology of Dependence on Tobacco, Alcohol, Controlled Substances, and Inhalants: Basic Findings from the National Comorbidity Survey”.
Experimental and Clinical Psychopharmacology. 1994;2:244-68.

My interpretation of these analyses is that we are doing something wrong with our “usual” drug abuse models in the laboratory. Those models are usually predicated on 100% of the animals developing the “phenotype” of interest by, e.g., self-administering the drug. This leads to a theoretical approach that drug abuse is simply a matter of “sufficient exposure” to a given drug. And in some ways this may very well be true. From a certain perspective.
It overlooks the fact that the human population of interest consumes drugs in highly variable patterns, influenced by any number of factors. If the development of addiction is a result of the “correct” repeated exposure to drug, this just begs the question of why some individuals end up exposing themselves according to the addiction protocol and some do not! It is my belief that a more interesting way to move forward in understanding drug abuse would be to concentrate our models on population outcomes that mimic the human data more closely. With appropriate subsets of the sample discontinuing use after a few samples, using repeatedly without dependence and developing drug dependence.
The final point results from the analysis I added to the graph above to reflect the population prevalence of dependence on the respective substances. The point is to illustrate the respective scope of public health concern related to various recreational drugs. For example the estimate suggests that the chances of becoming dependent on heroin (23% of lifetime users) are greater than the chances for alcohol (15%) or cannabis (9%). However, the much greater population prevalence of alcohol (91.5%) or cannabis (46.3%) in comparison with heroin (1.5%) means that the fraction of the entire population that is dependent on heroin (0.3%) is miniscule in comparison with the fraction dependent on cannabis (4%) or alcohol (14%). It is true that the personal and societal impact of heroin dependence may be much greater than that of alcohol or cannabis. Nevertheless it is clear that the number of individuals affected by alcohol or cannabis dependence dwarfs the number affected by heroin addiction.
MonkeyEgg.jpg
return to hunt
Those who think that cannabis is perfectly benign from a public health standpoint should grapple with the data suggesting that population dependence rates for cannabis (4%) exceed those for cocaine (2.7%), stimulants (1.7%) and anxiolytics (1.2%) as well as that for heroin (0.3%).

Advertisements

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: