ENGLISH :
Aim To examine the associations between the frequency of cannabis use and the use of other illicit drugs. Design A 25-year longitudinal study of the health, development and adjustment of a birth cohort of 1265 New Zealand children. Measurements Annual assessments of the frequency of cannabis use were obtained for the period 1425 years, together with measures of the use of other illicit drugs from the same time period. Findings The frequency of cannabis use was associated significantly with the use of other illicit drugs, other illicit drug abuse/dependence and the use of a diversity of other drugs. This association was found to be particularly strong during adolescence but declined rapidly as age increased. Statistical control for confounding by both fixed and time dynamic factors using random- and fixed-effects regression models reduced the strength of association between frequency of cannabis use and other illicit drug use, but a strong association between frequency of cannabis use and other illicit drug use remained even after control for non-observed and time-dynamic sources of confounding. Conclusions Regular or heavy cannabis use was associated with an increased risk of using other illicit drugs, abusing or becoming dependent upon other illicit drugs, and using a wider variety of other illicit drugs. The risks of use, abuse/dependence, and use of a diversity of other drugs declined with increasing age. The findings may support a general causal model such as the cannabis gateway hypothesis, but the actual causal mechanisms underlying such a gateway, and the extent to which these causal mechanisms are direct or indirect, remain unclear. (Author' s abstract)