Associations between cigarette smoking and cannabis dependence: a longitudinal study of young cannabis users in the United Kingdom

Abstract

Aims: To determine the degree to which cigarette smoking predicts levels of cannabis dependence above and beyond cannabis use itself, concurrently and in an exploratory four-year follow-up, and to investigate whether cigarette smoking mediates the relationship between cannabis use and cannabis dependence.

Methods: The study was cross sectional with an exploratory follow-up in the participants' own homes or via telephone interviews in the United Kingdom. Participants were 298 cannabis and tobacco users aged between 16 and 23; follow-up consisted of 65 cannabis and tobacco users. The primary outcome variable was cannabis dependence as measured by the Severity of Dependence Scale (SDS). Cannabis and tobacco smoking were assessed through a self-reported drug history.

Results: Regression analyses at baseline showed cigarette smoking (frequency of cigarette smoking: B=0.029, 95% CI=0.01, 0.05; years of cigarette smoking: B=0.159, 95% CI=0.05, 0.27) accounted for 29% of the variance in cannabis dependence when controlling for frequency of cannabis use. At follow-up, only baseline cannabis dependence predicted follow-up cannabis dependence (B=0.274, 95% CI=0.05, 0.53). At baseline, cigarette smoking mediated the relationship between frequency of cannabis use and dependence (B=0.0168, 95% CI=0.008, 0.288) even when controlling for possible confounding variables (B=0.0153, 95% CI=0.007, 0.027).

Conclusions: Cigarette smoking is related to concurrent cannabis dependence independently of cannabis use frequency. Cigarette smoking also mediates the relationship between cannabis use and cannabis dependence suggesting tobacco is a partial driver of cannabis dependence in young people who use cannabis and tobacco.

Keywords: Addiction; Cannabis; Co-morbidity; Dependence; Longitudinal; Tobacco; United Kingdom.

Fig. 1

Participant flow diagram for opportunistic follow up, 4 years after baseline.

Fig. 2

Multiple mediator model. Paths a, b, c and c′ are OLS regression coefficients in unstandardised form. Path c′ represents the effect of x on y when tobacco variables (m) are included as mediators. Path c represents the effect of x on y when the tobacco variables are not included as mediators. *p < .05, **p < .01, ***p < .001.