Twin study fails to prove 'gateway' hypothesis
by Dr Andrew Byrne

Original paper

Dear Colleagues,

Twin studies can be informative in causation theories. These authors state that in addition to having close or identical genetic make-up: ".. twin pairs, having been reared in the same household, would be expected to be highly concordant for environmental experiences." Thus most twins are exposed to alcohol, tobacco and other drugs at much the same age.

In their lead item in JAMA, Lynskey et al. find that for the exceptional minority of twin pairs (~300 out of 4000) in whom 'cannabis use before age 17' was discordant, that subsequent reported drug abuse/dependency was 2 to 5 times more prevalent in the early cannabis users. The authors find that this association lends weight to causation while admitting it is 'not possible to draw strong causal conclusions' of the 'gateway' theory. It is intriguing that they would address causation when this is a retrospective, cross-sectional study, a design which is not able to determine causation.

Since twins who used cannabis in the same year were eliminated, conclusions based on their similarities of upbringing must be guarded. These twins demonstrated at least one major difference in their environment and/or decision making on at least one occasion during adolescence. Whatever caused this may also explain the higher reported rates of other drug use, quite independent of any theoretical 'chemical priming' or 'gateway' effect.

These results are all derived single follow-up telephone interviews with an unknown party over matters relating to illegal drug use, child sex abuse and other personal issues up to 15 years earlier. Some may have chosen to (falsely) deny childhood cannabis use and then to also deny adult abuse or dependency. Others may have had faulty recollection for such distant events, making the findings less secure.

A certain minority of young people use hard drugs prior to using cannabis (around 1 - 2% from household surveys). Such subjects should be of considerable interest to those addressing the so-called 'gateway' theory. Lynskey et al. however, having found that up to 17 of their subjects used hard drugs before being exposed to cannabis, chose to exclude them from their study.

Another problem with this study is that the drug abuse/dependence findings are so high that they may indicate an atypical sample. For 'any illicit drug abuse/dependence' the prevalence was 33-48%; for 'alcohol dependence' it was 30-43% ['non cannabis by age 17' group first percentage followed by 'early users group' %]. Alcoholism is only thought to affect around 5% of adult males in the general population.

The authors state that their findings "..were consistent with early cannabis use having a causal role as a risk factor for other drug use and for any drug abuse or dependence." But further, they state: "While the findings of this study indicate that early cannabis use is associated with increased risks of progression to other illicit drug use and drug abuse/dependence, it is not possible to draw strong causal conclusions solely on the basis of the associations shown in this study." A 'causal' link between early cannabis use and later hard drug use remains unlikely on balance (National Academy of Science review) - and it is hard to imagine that a study of this nature could clarify the issue, no matter how well it was performed and analysed.

An associated editorial by Kandel teases some of the matters out while still seeming to assume that all cannabis use is problematic and needs to be discouraged by any effective means. Her own work from 19 years ago showed the strong association between alcohol/tobacco and illicit drug use.

Kandel indicates three factors necessary to prove the gateway hypothesis: (1) sequencing, (2) association and (3) causation. As she points out, the third is the hardest to prove.

In discussing the modern calls for prescribed cannabis, Kandel states that there is no empirical knowledge on whether medical cannabis will lead to problems. But cannabis products were widely prescribed in the first half of the 20th century without apparent problems arising (eg. tincture of cannabis). She says that it is a 'curious phenomenon' that morphine used for medical purposes 'does not lead to addiction'. Could it be equally 'curious' that medical cannabis does not do so either?

And unlike Lynskey et al., Kandel does not address the known non- chemical associations between cannabis and hard drug use. Having found from personal experience that drug education about cannabis was unreliable, young people may then reason that information about heroin and cocaine being dangerous is also unreliable. Similarly, having broken the law on cannabis, they may then have less compunction about breaking laws relating to other drugs. There is at least anecdotal evidence that some heroin addicts first used heroin when their dealer could not supply cannabis, amphetamine or other drugs of current choice. To his credit, Lynskey writes: ".. access to cannabis use may provide individuals with access to other drugs as they come into contact with drug dealers. ... [Dutch decriminalization of cannabis] may have been partially successful as rates of cocaine use among those who have used cannabis are lower in the Netherlands than in the United States." [One wonders what criteria the authors would need for 'complete' success of Dutch cannabis policy!]

Citations: Lynskey MT, Heath AC, Bucholz KK, Slutske WS, Madden PAF, Nelson EC, Statham DJ, Martin NG. Escalation of Drug Use in Early-Onset Cannabis Users vs Co-twin Controls. JAMA (2003) 289:427-433

Kandel DB. Does Marijuana Use Cause the Use of Other Drugs? JAMA (2003) 289; 4: Editorial