GENETIC EPIDEMIOLOGY OF CANNABIS USE, ABUSE AND DEPENDENCE: A COMMENT ON AGRAWAL & LYNSKEY (2006)

Abstract

Agrawal & Lynskey [1] have written a comprehensive and thoughtful review of the current state of knowledge in this area. First, they review studies demonstrating a wide range of estimates for the relative contribution of genetic factors, i.e. heritability, to the interindividual variation in cannabis use, abuse and dependence, with a heavy emphasis on twin studies. Through twin studies they also explore the nature of the relationship of cannabis use with licit (tobacco and alcohol) and illicit drug use, and show that there are probably underlying genetic mechanisms and dispositions in common. They conclude by reviewing avenues for future research that have not yet been explored to any great degree. These include broadening the phenotype to include other aspects of cannabis use/abuse, and its potential endophenotypes [2]. In addition to those mentioned by the review, neurophysiological, neuroimaging and neuropsychological-based endophenotypes should be explored in genetically informative designs. These have proved very useful in other substance use research and in schizophrenia research [3, 4]. As cannabis use is a risk factor for schizophrenia and psychosis [5, 6], understanding the aetiology and pathology of cannabis addiction might also help us understand psychoses and schizophrenia better. A limitation of current studies on the genetic epidemiology of cannabis addiction is their very restricted provenance, as acknowledged in the review. Virtually all have been conducted in Anglo-Saxon societies, despite which a very wide range of heritability estimates has been derived, and virtually all studies indicate a major role for shared environmental effects in addition to genetic effects. Studies from other cultures with different experiences of cannabis use are needed. We have assessed cannabis use as part of our ongoing longitudinal studies of adolescent twins in Finland. In the FinnTwin 12 study of Finnish adolescent twins born in 1983–87, more than a third of subjects reported alcohol use when surveyed by questionnaire immediately after their 14th birthday [7]. Among 1854 of the same twins interviewed at age 14 years, only 1.1% had experimented (at least once) with cannabis and no genetic modelling was possible. The third wave of the same study was carried out in 2000–05, when the twins were aged 17 years [8]. Among 1905 monozygotic (MZ) and dizygotic (DZ) pairs, > 13% reported experimentation and 1.6% had used 20 or more times in a mailed questionnaire survey. Pairwise tetrachoric correlations for any use were high in male (0.89) and female (0.87) MZ pairs, while correlations were somewhat lower in male DZ pairs (0.71), female DZ pairs (0.70) and male–female DZ pairs (0.52). In the FinnTwin16 study of Finnish twins born from 1975 to 1979, cannabis and other illicit drug use was asked of them as young adults [8] at an average age of 24 years during 2000–02. More than 22% reported experimentation and 3.7% had used 20 or more times. Among the surveyed twins (n = 2005 MZ and DZ pairs), tetrachoric correlations for any use were high in male and female MZ pairs (both r = 0.81), while correlations were lower in male DZ pairs (0.64), female DZ pairs (0.49) and male–female DZ pairs (0.33). Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA)-based [9] interviews of a subset of subjects (n = 602 twins) indicated a high reliability of the history of any use of cannabis (kappa = 0.85) from the questionnaire. The pattern of correlations in the two Finnish studies indicates the presence of substantial genetic effects and shared environmental effects even without carrying out formal genetic modelling in this non-Anglo-Saxon population in which cannabis use is less frequent than in the reviewed studies. In many aspects, the genetic epidemiology of cannabis resembles that of alcohol use, for which substantial shared environmental effects have also been demonstrated and for which gene–environment interaction effects are recognized [10]. The heritability of adolescent alcohol use ranges widely in different social contexts [11-13] even within Finland, which is a relatively homogeneous country genetically and culturally. It is likely that gene–environment interaction accounts for a substantial proportion of the apparent heritability of cannabis use and abuse. This would suggest that the relative importance of direct versus indirect genetic influences on cannabis use should be explored carefully before embarking on expensive molecular genetic studies.