Associations between tobacco and cannabis use in remote indigenous populations in Northern Australia

Aims To assess whether cannabis use, recently taken up by many indigenous Australians in remote communities, has reinforced tobacco use. Design Cross-sectional study. Setting Three eastern Arnhem Land communities (Northern Territory, NT); total population = 3384, in 2001. Participants From 1247 people aged 17-36 years, 190 (120 males, 70 females) were opportunistically recruited. Measurements Self-reported life-time and current tobacco, cannabis and other substance use were confirmed by local health workers and using clinic records. Participants reported level of substance use, frequency and duration (years used). Associations with tobacco use were calculated (odds ratios: OR) using logistic regression with age, sex, alcohol use and a history of petrol sniffing as confounders. Findings In univariate analyses current tobacco users were more likely than non-users to be using cannabis (OR = 3.1, 1.5-6.2, P = 0.002) and this association remained in multivariate analyses (OR = 3.0, 1.4-6.8, P = 0.006). Tobacco use was associated with the number of years of cannabis use (P = 0.035). The likelihood that tobacco users were also cannabis users increased as quantity of cannabis used increased (P = 0.008). Current tobacco use was no more likely in those who initiated cannabis from 1998 onwards than in those who initiated cannabis before 1998 (OR = 1.1, 0.4-3.2, P = 0.881). One-third of life-time users of both tobacco and cannabis initiated their use at or near the same time, and very few of these (12%) had discontinued either cannabis or tobacco. Conclusions Cannabis appears to have influenced the continued use of tobacco in these populations with possible additional burdens for cardiovascular and respiratory diseases and challenges for interventions.