This study had some limitations including  cross-sectional nature of the data used for model building,  participant recruitment from the southwestern US among justice-involved youth who use alcohol and cannabis regularly, and  a limited sample size after restricting to users of all three substances. To partially address , we only used risk factors that would remain relatively stable over time. For , we note that regular alcohol and/or cannabis grow rack use was defined as at least once per month for the past 6 months, so the study inclusion criteria is rather representative of many adolescents. To mitigate , we used leave-one-out cross-validation to evaluate the predictive accuracy of our models, which guards against over fitting. In light of these limitations, we consider the proposed model to be preliminary. In summary, this study is a novel attempt to build a risk prediction model for adolescent users of multiple substances. It serves as an important step towards our ultimate goal of building a comprehensive risk prediction model based on a large nationally representative longitudinal sample of adolescent users.

Such a model can help us identify risk factors for hazardous substance use in adolescence and ideally flag factors that might help us understand the future transition into substance use disorders in adulthood.  The prevalence of cannabis use world-wide has been fairly stable over the past decade, with an estimated 3.8 per cent  of the global population aged 15–64 in 2018 having used cannabis at least once in the previous year, compared to 3.6% in 2008 . Cannabis use in the United States  has been increasing since 2002, adding an average of 7,000 new users daily. In 2014, 8.4% of the population were current cannabis users , representing an increase of 35% over 2002 rates . Cannabis use has increased across adult age groups , with 18–25 year-olds having the highest current-use rate , of whom a large portion  reported daily or almost daily cannabis use.Combustion of tobacco and cannabis is common  and produces heavy particulate matter emissions , which is a significant agent of cardiovascular harm .

Furthermore, the chemical composition of cannabis smoke is similar to that of tobacco smoke with the two having similar levels of cytotoxic, mutagenic, and carcinogenic substances.Cannabis grow system smoke contains higher levels  of ammonia, hydrogen cyanide, 1,3-butadiene, and acrylonitrile—constituents identified as harmful or potentially harmful by the FDA.The Global Adult Tobacco Survey found that of approximately one billion children under age 15 in 21 countries, 51% were exposed to SHS . However, efforts to restrict tobacco smoking are encouraging and lead to lower SHS exposure. As of 2017, more than 30 countries globally had instituted at the highest level at least 4 of the 6 most important tobacco control policies  that were first recommended in 2008 by the World Health Organization . At least one MPOWER tobacco control policy measure is in effect for approximately 65% of the world population . Details of tobacco smoking regulations by country are provided by the World Health Organization in Tables 6.1, 6.2, and 6.3 of their report on the global tobacco epidemic . The 3 tables reveal considerable variation in smoke-free legislation by country regarding the sorts of locations or establishments restricted, kinds of restrictions imposed, level of compliance, investigations of complaints, consequences for violations, etc.; there are also regulations that vary by region in some countries. Level of regulation likely influences home smoking rules and behaviors.

Tobacco SHS exposure most often occurs in the home  & U.S. Department of Health and Human Services, 2006; Wipfli et al., 2008; Zhang, Martinez-Donate, Kuo, & Jones, 2012.Fortunately, tobacco control efforts have successfully reduced the number of people who smoke tobacco in their home. In Spain, assessments conducted before implementation of two national smoking bans in public and work places and 9 years later  showed an increase in voluntary adoption of smoke-free home rules . England saw a trend to more smoke-free homes from 1996 to 2007, with a concomitant decrease in SHS exposure . In the U.S., the national prevalence of smoke-free home rules increased from 43.0% to 86.8% between the 1992–1993 and 2014–2015 surveys , while SHS exposure prevalence among nonsmokers declined from 52.5% to 25.3%  . In California, the proportion of smokers exposed to in-home cigarette smoke decreased from 24.0% in 2003 to 13.0% in 2011, while a similar relative decrease was observed among nonsmokers  .