I. OUR APPROACH

On the face of it there is a reasonable cause for concern about the public health significance of cannabis use in many developed societies. First there is a high prevalence of cannabis use in such societies, especially among adolescent sand young adults. Second, cannabis is an intoxicant like alcohol that is usually smoked like tobacco so there is a reasonable presumption that it shares at least some of the adverse health effects of these two drugs, both of which have a substantial public health impact.

There are a number of obstacles to providing a confident appraisal of the personal and public health significance of cannabis use. The first set of issues concerns difficulties in making causal inferences about the connections between cannabis use and the adverse health and psychological consequences which have been attributed to it. The second set of issues concerns the quantification of the seriousness of the risks of cannabis use for users and for the broader community.

A third set of issues concerns the difficulties in making comparative appraisals of risk. Since no human activity is ever risk free, the appraisal of the health risks of cannabis presupposes a comparison of its risks with those of other relevant activities. The most obvious comparison is with the risks of using the other major recreational drugs in Western societies namely, alcohol, and tobacco. We have also included some comparisons with opiates which, serves the useful purpose of calibrating the health risks of cannabis against those of a drug which is widely regarded as a major public concern even though it is not widely used.

Our purpose in making these comparisons is not to promote one drug over another but rather to minimise the double standards that have operated in appraising the health effects of cannabis by persons on both sides of the debate about its legal status. On the one hand, there has been a disproportionately greater concern about the use of cannabis than alcohol on the part of many older persons who are opposed to cannabis use. On the other hand, younger cannabis users have denied the existence any adverse health effects of cannabis despite the fact that it intoxicates like alcohol and is smoked like tobacco. A further reason for the comparison is that many of our inferences about the probable adverse health effects of cannabis use depend upon analogies between the health effects of alcohol, cannabis and tobacco.

A factor which complicates the comparative assessment is the way in which the drug is taken. In present-day developed societies, the main route of administration of cannabis and of tobacco is by smoking, that is, inhaling the smoke of a smouldering preparation containing the drug. In pharmaceutical use, opiates are administered orally or by injection, while nonmedical use is primarily by injection, snorting, or smoking. Alcohol is consumed orally, although in combination with a wide variety of other substances. The route of administration of a drug can change over time in a given society: for instance, in North America tobacco was primarily chewed or sniffed in the late 19th century, and there is evidence there of some shifting at present from injecting to smoking heroin. In this analysis, we have focused on predominant routes of administration in developed societies, but it should be recognised that a shift in route of administration can substantially change the size and profile of adverse effects. In particular, the most serious adverse health problems of cannabis and tobacco are associated with smoking as the mode of administration.

The approach we have adopted in addressing each of these issues is as follows. First, we have identified the most probable causal relationships between cannabis use and specific health effects. In doing so we have used standard criteria for assessing the strength of evidence for a causal relationships, although we have had to relax the degree of confidence required so that some provisional conclusions could be drawn.

Second, in so far as it is possible we have attempted to quantify the severity of personal and public health risk for each adverse health effect that can be reasonably attributed to cannabis. We have attempted to estimate the probable relative risk, and the prevalence of the relevant pattern of use.

Third, we have compared these estimates with the best estimates of the mortality and morbidity burden of alcohol, opiates and tobacco. This has been done initially in a qualitative way by indicating whether or not particular adverse health effects that may reasonably be attributed to cannabis have also been attributed to alcohol, nicotine, and opiates. This is followed by a discussion of the probable quantitative risks of cannabis by comparison with those of alcohol and nicotine, and some direct comparative evidence on consequences reported by users of the three drugs.

In making these comparisons we have relied on epidemiological evidence on the health and psychological consequences of cannabis use which is largely based on studies conducted in the English-speaking countries, and most particularly the United States. Unfortunately, countries with a long tradition of heavy cannabis use are not well represented in the research literature. The preponderance of American research on the health effects of cannabis use reflects societal concern about the emergence over the past several decades of widespread cannabis use among adolescents and young adults in the USA; concern that has been translated into funding for epidemiological studies of the health and psychological consequences of cannabis use. The conduct of research on cannabis use in developing countries should be a priority, especially in the countries that have a long history of traditional use, including very heavy use among some subpopulations. These subpopulations are the ones most likely to show any adverse health and psychological effects of chronic heavy use.

Our comparisons of health effects are also largely confined to the effects on the health of users. We have said little about the effects of cannabis use on the health and well being of other persons who do not use cannabis. Such indirect health effects have not been well studied for most drugs, with the limited exceptions of motor vehicle accidents, and violence for alcohol, and passive smoking for tobacco. They perhaps deserve more attention that they have hitherto received but in the absence of the necessary research we are unable to address them in this review.