You are on page 1of 2

EDITORIAL Editorials represent the opinions

of the authors and THE JOURNAL and not those of
the American Medical Association.

Does Marijuana Use Cause
the Use of Other Drugs?
Denise B. Kandel, PhD on initiation of use of other illicit drugs would also explic-
itly be taken into account. The argument would be stron-

A
DEVELOPMENTAL SEQUENCE OF INVOLVEMENT IN ger if the 2 models had been tested and the common-factor
drugs is one of the best replicated findings in the model was found to provide as good a fit as the causal model.
epidemiology of drug use. Regular sequences and While the simulation model of Morral et al does not dis-
stages of progression in which the use of alcohol prove the existence of a causal relationship between mari-
and cigarettes precedes the use of marijuana (cannabis), and, juana and other illicit drugs, it provides an alternative in-
in turn, the use of marijuana precedes the use of other illicit terpretation.
drugs, has been observed in the United States as well as in A second strategy is to evaluate the impact of prevention
other western societies.1 Very few individuals who have tried or intervention programs implemented among youths to pre-
cocaine and heroin have not already used marijuana; the vent or reduce drug use. Such programs provide an imper-
majority have previously used alcohol or tobacco. Such behav- fect substitution for an unrealizable social experiment in which
ioral regularities are subsumed under the “gateway hypoth- adolescents would be randomly assigned to initiate the use
esis.” The gateway hypothesis implies 3 interrelated propo- of different drugs. However, programs designed to prevent
sitions about sequencing, association of initiation, and or stop use of lower-stage drugs also seem to stop or reduce
causation.1 Sequencing implies that there is a fixed relation- use of higher-stage drugs, as reported in Kandel.1
ship between 2 substances, such that one substance is regu- Still a third strategy is the use of genetically informative
larly initiated before the other. Association implies that ini- samples. This approach has not previously been imple-
tiation of one substance increases the likelihood of initiation mented to test the gateway hypothesis and represents a
of the second substance. Causation implies that use of the unique contribution of the work of Lynskey et al.2 In a pre-
first substance actually causes use of the second substance. vious report6 based on data from a large sample of Austra-
Causation, a controversial proposition, is the one most widely lian twins born in 1964 to 1971, Lynskey et al concluded
invoked in policy debates and is the proposition addressed that genetic risk factors are important determinants of the
in the article by Lynskey et al in this issue of THE JOURNAL.2 risk of marijuana dependence, at least among men. In the
Several strategies are available for determining the causal current report, based on a subsample of 311 same-sex twin
role of a lower-stage drug in the sequence on initiation of a pairs from the Australian cohort who were discordant for
higher-stage drug. In one strategy, epidemiologists attempt early marijuana use by age 17 years, Lynskey et al find that
to specify the role of prior drug use on the subsequent use of early marijuana use by itself, even after control for other co-
other drug classes by controlling for theoretically relevant co- variates, increases significantly the use of other illicit drugs.2
variates and other confounding factors. Analyses based on this As the authors emphasize, the strength of the twin design
approach find that marijuana retains a significant associa- is that twins are assumed to share the same environment
tion with the subsequent use of other illicit drugs even after and family experiences, and monozygotic pairs share the same
control for risk and protective factors.3,4 genetic risk. If these factors explained the association be-
However, in naturalistic population studies, it is not pos- tween early marijuana use and the use of other illicit drugs,
sible to control for all potential relevant factors. Morral et the risk of using these drugs would be the same for early
al5 have attempted to overcome this limitation by develop- marijuana-using twins and their discordant co-twins. But
ing a simulation model, which assumes that the use of mari- it is not. Thus, Lynskey at al conclude that “The re-
juana and other illicit drugs is explained solely by a general sults . . . were consistent with early cannabis use having a
propensity to use drugs and the ages at first opportunity to causal role as a risk factor for other drug use and for any
use and at first actual use of marijuana and other illicit drugs. drug abuse or dependence.”2 But is the assumption of com-
This common-factor model replicates the empirical asso- mon environmental influences among twins too strong? An
ciation between marijuana and other illicit drugs observed argument can be made that even identical twins do not share
in a national sample without positing an effect of the use of the same environment during adolescence and that envi-
marijuana itself.5 However, the fit of this model was not com-
pared with one in which the causal effect of marijuana use Author Affiliations: Department of Sociomedical Sciences, School of Public Health,
and Department of Psychiatry, College of Physicians and Surgeons, Columbia Uni-
versity, and New York State Psychiatric Institute, New York, NY.
See also p 427. Corresponding Author and Reprints: Denise B. Kandel, PhD, Department of Psy-
chiatry, Columbia University, 1051 Riverside Dr, Unit 20, New York, NY 10032.

482 JAMA, January 22/29, 2003—Vol 289, No. 4 (Reprinted) ©2003 American Medical Association. All rights reserved.

Horwood LJ. Stages and Pathways of Drug Involvement: Examining the Gate- of marijuana and other drugs. identify some of the critical biological processes involved inferences can be made from appropriate medical use of mor- in drug use behavior. Bucholz KK. DA00081 from the National Institute on Drug Abuse. It is important. cultural norms. ior in animal models because each has much to contribute to Lynskey et al cite 2 recent studies. shaping its development. un- ing the use of various substances. no empirical data to guide policy. tence of a gateway effect. which does not lead to addiction. Solinas M. et al.74:673-681. Nelson EC. consistent with the exis. Brenner S. 3. prevention efforts will presumably affect the underlying risk 7. epidemiolo- pecially with marijuana. 6. progressing to other substances? It appears so. Presented at: The Nobel Assembly at Karolinska Institutet. Inter- iors. norms. January 22/29. the use of illicit drugs. (Reprinted) JAMA. to explain the association between the use of different drug mental question that underlies the gateway hypothesis. Addiction. personal traits. one explore the many other social. This is the funda. termined by availability. England: Cambridge University Press. Lecture. In: Kandel DB. 2003—Vol 289. whether they are common to mari- One of the best strategies to test the gateway hypothesis juana and the use of other illicit drugs or whether they are may be the use of animal models. 4 483 . the central question remains: does marijuana use cause behavior in human populations. Cardoni C. Nature’s gift to science. Reverse sensitization suggests that Funding/Support: Work on this editorial was supported by Research Scientist Award the particular sequence observed in any society may be de. raises 2 issues depending on whether the population of in. Pisanu A. animal mod. Psy- marijuana. 2002. For youths who have already used exposure to ⌬9-tetrahydrocannabinol and cross-sensitization with morphine. Stages and Pathways of Drug Involvement: Ex- and protective factors related to the onset of marijuana use. With cross. the association between the 2 way Hypothesis. increase the risk of using other illicit drugs? There are. or moral definitions regard. chopharmacology. Morral AR. III: predictors of progression. Nobel Prize for Physiology or Medi- grams be developed to target this group at very high risk for cine. 2003. amining the Gateway Hypothesis. Escalation of drug use in early-onset preciate that the progression is not inevitable. marijuana for medicinal purposes? Will medical use dependent of any social. however. but also reverse sation and provide insights into the neurobiological mecha- sensitization between morphine and cannabinoids. absence of direct experimental manipulation may be elusive. 1984. to ap. by Sydney Brenner in his recent Nobel Prize lecture9—can sensitization. the gateway hypothesis 4. els can examine whether prior use of a drug per se increases Nevertheless. juana use that also support intervention. The effect attributed to early marijuana use could in ventions should focus on those factors that are associated with part be due to nonshared environmental influences. Animal models provide a test of progres. 2002.7 ior and the role of psychological and social conditions in While animal models cannot mimic the variety of cul. et al. Sensitization as a process underlying the progression of drug use via gateway drugs. the use of other illicit drugs? The search for causes in the personal meanings. whether or not these factors are shared with the onset of the 2002:318-336. Lynskey MT. epidemiologists can et al8 is of particular interest because it has documented not pose and specify the overall questions that need to be addressed. Yamaguchi K. 2002. Such models may also help fortunately. 1997. Schenk S. psychological. which in turn would lead to differences in behav. illicit drugs needs to be pursued. Cambridge.32:195-207. Paddock SM.289:427-433. Cambridge. December 8. McCaffrey DF. 2002.2 there are health risks associated with chronic mari- use paradigms. A marijuana Stockholm. Lynskey MT. In particular. has been well established. and measures to brain. 5. es. what inferences can be drawn about the use of well-defined prior experiences with specific drugs and in.92:279-296. Prior exposure to one class of drug phine. Early onset cannabis use and psychosocial adjust- ment in young adults. 8. sion in which drug seeking can be observed in relation to Finally. Psychol prevent the use of other illicit drugs? Hopefully it will. the progression from marijuana to other very few such animal experiments have been conducted. Fergusson DM. Heath AC. or other factors. peer influences. These nisms underlying progression in drug use. for Med. 9. 2001. Whether or not a true causal link exists between the use REFERENCES 1. and so on. exposure to one class of drug increases con. All rights reserved. the issue is: will preventing the use of marijuana butions to cannabis dependence in a national young adult twin sample. ed.158:259-266. Addiction. the issue is: can and should intervention pro. only cross-sensitization between repeated exposure to whereas animal researchers can use their models to test cau- ⌬9 -tetrahydrocannabinol and opiates. tural. In this collaboration. Kandel DB. Not all those cannabis users vs co-twin controls. However. This is a curious may sensitize the animal to the use of other drug classes and phenomenon that points out the complexity of drug behav- increase the consumption of these drugs. textual factors that are also important in drug use behavior. England: Cambridge University Press. Sweden.97:1493-1503. Behavioural sensitization after repeated use of other illicit drugs. ©2003 American Medical Association. For policy makers.”as poetically defined behavior and for the gateway hypothesis. social. JAMA. in specific to illicit drugs other than marijuana. But only in a experiments have important implications for human drug human—“a model organism for the gods. Am J Public Health. and con- sumption of other drug classes. such as rats or mice. No. The work by Cardoni the other’s quest. But classes. Patterns of drug use from adolescence to young adult- who try marijuana will subsequently use cocaine or become hood. prevent subsequent use of these drugs are warranted. Heath AC. and most have focused on priming gists must collaborate with scientists who study drug behav- by one class of drug on the subsequent use of the same drug. For a population of effect. As noted by Lyn- which animals can be assigned to a number of different drug skey et al. and psychological factors that determine drug Thus. et al. legal. Reassessing the marijuana gateway terest has or has not yet used marijuana. heroin addicts. Kandel DB. In this search. EDITORIAL ronmental differences can lead to learned differences in the user is at risk for using other illicit drugs. ed. 2. eg. Genetic and environmental contri- nonusers. the search for mechanisms is necessary if only the risk of the use of another drug class.