Page 19
knowledge on risky behaviours within the Canadian population, with a specific focus on younger populations. For the 2017 cycle of CTADS, data collection was initiated on February 1st, 2017 and ended on December 31st, 2017. The target population consisted of individuals aged 15 and above and residing in the 10 Canadian provinces. Full-time residents of institutions and those living in the three territories (Yukon, Northwest Territories and Nunavut) were not included in the survey. The survey was conducted using a specialized sample design which included a two-phase stratified random sampling of telephone numbers. Those without telephones or cellphones were not recruited for the study, however survey estimates have been applied to include those without landlines and cellular phones. Participation in the survey was voluntary and data were collected directly from the survey respondent. Additionally, missing or inconsistent information was handled by Statistics Canada through various types of editing to ensure logical relationship between responses. A more detailed breakdown of the survey’s design and sampling procedure can be found on the Statistics Canada website [33]. For the purposes of this study, only individuals who reported being 15–24 years of age were included. The total sample comprised of 10,322 individuals, representing approximately 4,443,600 Canadians of ages 15–24. The average age of the respondents was 19.5 years, whereas the median age was 20 years. Additionally, 48.6% of the sample were females and 51.4% were males. Main outcome. The main outcome of the study was ‘e-cigarette use’. This was measured based on the following question: “In the past 30 days did you use an electronic cigarette, also known as an e-cigarette?” The respondents were provided with the following options, “Yes”, “No”. CovariatesThe main independent variables for this study included, substance use variables comprising of alcohol use in the past 12 months (asked by the question, ‘How often did you drink alcoholic beverages during the past 12 months’ with ‘Never’ recoded as ‘no’ and any alcohol use recoded as ‘yes’), marijuana (asked by the question ‘During the past 12 months have you used marijuana?’ yes or no), and illicit drug use (including cocaine, speed/meth, ecstasy, hallucinogens, salvia, heroin, inhalants, abuse of pain relievers, stimulants and sedatives to get high in the past 12 months, assessed using questions asking participants whether they had used a specific substance in the past 12 months, yes or no). Tobacco use variables included household smoking (asked by the question, ‘Do you/Does anyone in your household smoke cigarettes? Yes or No) current smoking status (variable created by Statistics Canada based on multiple questions), and other tobacco products (comprised of use of either cigarillos, cigars, tobacco water pipe or smoke-less tobacco in the past 30 days, assessed using questions asking participants whether they had used any of the above specified tobacco products in the past 30 days, yes or no). In addition, a wide range of covariates including demographic and health factors, and tobacco exposure variables, were considered. Demographic and health factors included age, sex, province, indigenous status (asked by the question ‘Are you an Aboriginal person, that is, First Nations, Métis or Inuk (Inuit)?’) , community dwelling (rural or urban), current employment (asked by the question ‘Last week, did you work at a job or business?’), and self-perceived health (asked by the question ‘In general, would you say your health is...?’ Additional variables considered in the analysis also included whether nicotine was present in the last e-cigarette used, frequency of using e-cigarette (everyday, occasionally or not at all) and reasons for using e-cigarettes.Descriptive statistics of the main outcome and other variables were conducted. Odds Ratios (ORs) and 95% Confidence Intervals (CIs) were obtained using bivariate and multivariable logistic regression. When all the variables were considered, a maximum of 3.16% of the cases had missing information. These cases were excluded from the analysis. Population weights were applied to each calculated estimate and bootstrapping was