Tobacco and cannabis are two of most widely used drugs in the world, with recent estimates demonstrating 22.5% of adults using tobacco products and 3.5% using cannabis. There is substantial overlap in their use, with approximately 90% of cannabis users also identifying as cigarette smokers. Furthermore, cigarette use increases during cannabis abstinence, and decreases following renewal of cannabis use suggesting direct substitution between the drugs.
The causal relationship between cannabis and tobacco use is complex. The ‘gateway’ hypothesis suggests cigarette smoking is a precursor (‘gateway drug’) to cannabis use, whereas the ‘reverse gateway’ hypothesis suggests cannabis use predicts tobacco smoking onset. To complicate matters, both substances are often used concurrently (‘mulling’) and sensitisation may occur, in which smoking tobacco may enhance the subjective effects of cannabis.
In a recent study published in Drug and Alcohol Dependence, the authors investigated the degree to which cigarette smoking predicted cannabis dependence both cross-sectionally but also longitudinally, at four-year follow up. Importantly, they also examined whether the effects of cannabis use on cannabis dependence were mediated by tobacco use.
A subset of 298 recreational cannabis users who also smoked tobacco was recruited from a larger study examining acute cannabis effects. All participants were aged between 16-23 years old, had no history of psychiatric illnesses, or learning impairments. During baseline data collection participants were visited in their homes and had to abstain from the use of all recreational drugs for 24 hours prior to data collection.
Measures taken from participants included demographic information, and drug history. Semi-structured, questionnaire-based interviews were conducted to examine past-use of cannabis and tobacco. This provided detailed measures of time since last use of tobacco and cannabis, how many days per month individuals used each drug and how many years individuals had been using each drug (years of use). Cannabis dependence was measured using the severity of dependence scale (SDS). Finally, measures of depression, anxiety, impulsivity and childhood trauma were taken.
For follow-up data, four years later, the participants were contacted via telephone and took part in a semi-structured interview. This included questions about individual’s drug use and questions were identical to those administered at baseline. Data was collected from 65 individuals at this stage.
Participants were 20.55 years old on average, and the majority (71%) were male. At baseline 134 (46.6%) met the criteria for cannabis dependence and at follow-up 17 (26.2%) met the criteria.
The authors conducted various models predicting cannabis dependence from cannabis use, tobacco use and measures of depression and anxiety. The most efficient model predicted 29.5% of variance in cannabis dependence. In this model, significant predictors were:
- Days of cannabis use per month (B = 0.11; 95% CI 0.08 to 0.13; p < .01)
- Days of tobacco use per month (B = 0.03; 95% CI 0.01 to 0.05; p = .01)
- Years of tobacco use (B = 0.16; 95% CI 0.05 to 0.27; p < .01)
- Depression scores (B = 0.05; 95% CI 0.01 to 0.10; p = .02)
To examine cannabis dependence at 4 year follow up the authors entered the variables from the cross-sectional model, including baseline cannabis dependence. The model predicted 24.8% of the variance in cannabis dependence at follow-up. The only significant predictor was dependence at baseline (B = 0.27; 95% CI 0.05 to 0.530; p = .02).
Multiple mediation analysis demonstrated a significant effect of days of cannabis use per month on cannabis dependence, mediated by days of tobacco use per month and years of tobacco use (B = 0.015, 95% CI = 0.007, 0.027). Importantly, these effects remained when controlling for depression.
In light of the increased medicalisation and legalisation of marijuana around the world, research into understanding the relationship between tobacco and cannabis use is essential. The results from this study suggest that tobacco use enhances the vulnerability to cannabis dependence, over and above cannabis exposure. Importantly, cigarette smoking mediated the relationship between cannabis use and dependence even when controlling for confounding variables such as depression.
The authors speculate that:
these results suggest that cigarette smoking may enhance the dependence forming effects of cannabis.
However, it is important to note that causality cannot be assumed in the cross-sectional analyses.
This study has some limitations:
- First, the follow up sample only had 65 participants which limits the statistical power for the analyses (the cross-sectional associations were reasonably well powered!), so future studies should examine any causal relationships using much larger samples.
- Second, they were unable to control for the simultaneous use of cannabis and tobacco (a popular way of smoking), therefore the analyses are limited to separate effects of each drug.
- Finally, the data was all self-reported which is often biased as it relies on retrospective recall and honesty from individuals (which may not be reliable when recalling illicit drug use).
This naturalistic study demonstrates that cannabis use and tobacco use predicted dependence cross-sectionally, but not longitudinally. The findings tentatively suggest tobacco use enhances vulnerability to cannabis dependence.
Hindocha C, Shaban NDC, Freeman TP, Das RK, Gale G, Schafer G, Falconer CJ, Morgan CJA, Curran HV. (2015) Associations between cigarette smoking and cannabis dependence: a longitudinal study of young cannabis users in the UK. Drug and Alcohol Dependence, 148, 165-7.
Allsop DJ, Dunlop AJ, Saddler C, et al. (2014). Changes in cigarette and alcohol use during cannabis abstinence. Drug and Alcohol Dependence, 138, 54-60.
Gowing LR, Ali RL, Allsop S, et al. (2014). Global statistics on addictive behaviours: 2014 status report. Addiction, 110, 904-19.
Belanger RE, Akre C, Kuntsche E, Gmel G, Suris JC. (2011). Adding tobacco to cannabis – its frequency and likely implications. Nicotine and Tobacco Research, 13, 746-50.