3.1. Route of tobacco use and cannabis-related behaviors
Cannabis use and abuse/dependence were more common in those who reported any tobacco use when compared to those who had never used tobacco () and was most common in those who had used both smoked and smokeless forms of tobacco. However, when examining exclusive routes of administration, both cannabis use and abuse/dependence were more common in those who reported using smoked forms of tobacco than those reporting only use of smokeless tobacco.
Prevalence of cannabis use and abuse/dependence in those using neither smoked nor smokeless tobacco, smoked tobacco only, smokeless tobacco only and those using both smoked and smokeless tobacco.
3.2 Characteristics of smoked and smokeless tobacco users
There were also a number of socio-demographic, psychiatric and substance –related measures that differed between individuals using smoked tobacco products exclusively and those who reported use of smokeless tobacco alone (). Men were more likely than women to use smokeless forms of tobacco as were those who were from older (50+ years of age) cohorts. The smokeless forms of tobacco were significantly more popular in those from the South/Midwest (MOR=1.96) and more common in those who were U.S. born (MOR=12.5). No notable differences were observed between these two groups for ethnicity or living below the poverty line. College education was not associated with using smokeless tobacco only.
Multinomial odds ratios [95% Confidence intervals] showing the association between route of administration of tobacco use and socio-demographic, psychiatric and substance use covariates in the NESARC.
Psychiatric diagnoses varied significantly across the smoked only and smokeless only groups (). Major depression, generalized anxiety disorder and panic disorder were markedly more common in those reporting smoked forms of tobacco use, even higher than in those using both forms of tobacco, with the prevalence in the smokeless group being equal or less than the prevalence in those who used neither form. For instance, those with a history of DSM-IV major depression were 1.64 times more likely to be exclusive users of smoked tobacco – this multinomial odds ratio could be equated to the MOR in those using both smoked and smokeless forms (MOR=1.47) and was not statistically significant in those using smokeless tobacco alone (MOR=0.80).
Conversely, alcohol abuse/dependence was more common in those reporting smokeless tobacco use (MOR=5.40, see ). A modest overrepresentation of a family history of substance-related problems and other illicit drug use was noted in the smoked versus smokeless groups. While we do not show the multinomial odds ratios for nicotine dependence in (primarily, because the reference group is never using any form of tobacco regularly, the MORs for nicotine dependence are rather large), post-hoc tests revealed that those using both forms of tobacco were most likely to be nicotine dependent, followed by those using smokeless forms only with the least risk in those using smoked forms alone.
3.3 Smoked tobacco and cannabis-related behaviors
To test the hypothesis that smoked forms of tobacco use would be more strongly associated with cannabis use and abuse/dependence, than smokeless forms, even after covariate adjustment, we conducted multinomial logistic regression analyses comparing those who exclusively used smoked forms, those who exclusively used smokeless forms and those who used both forms of tobacco. The odds of cannabis use and abuse/dependence were highest in those reporting use of both smoked and smokeless forms (). Those using smoked forms of tobacco (cigarettes, cigars or pipes) were 3.3 and 4.5 times more likely to also report a lifetime history of cannabis use and abuse/dependence respectively, with the risk being considerably lower in those endorsing use of smokeless forms only. Remarkably, even after controlling for socio-demographic, psychiatric and substance-related covariates, including nicotine dependence (which was strongly associated with cannabis use and abuse/dependence), a smoked route of tobacco administration was significantly associated with cannabis use (MOR 1.99) and abuse/dependence (MOR 1.55). In contrast, smokeless forms of tobacco use were not significantly associated with cannabis use and abuse/dependence (MOR 0.96 and 1.04) after covariate adjustment. The odds-ratios in those who used smoked forms alone could be equated to the odds in those who used both smoked and smokeless forms, suggesting that after covariate adjustment, only smoked tobacco use was associated with cannabis use and abuse/dependence.
Unadjusted and adjusted multinomial odds ratios [with 95% Confidence Intervals] for the association between smoked only, smokeless only and both smoked and smokeless forms of tobacco use and cannabis use and abuse/dependence