Thus, we included measures of demographics in the analyses presented below, as well as factors that have been related to alcohol use and selleck kinase inhibitor smoking. Demographic variables. Demographic variables included age, gender, race, and mother��s and father��s educational levels (some college education or less vs. 4-year college degree). Lifestyle factors. Lifestyle factors included residence (on- or off-campus), Greek pledge or member, and living in a smoke-free dorm (yes/no). College-level correlates. With only 10 schools participating in the study, we had limited ability to include college-level factors in statistical analyses. Consequently, the college-level factors that were included were whether the school was a private or public institution, the overall smoking rate on the campus (based on participants�� past-30-day smoking rate), and whether the campus was an intervention or a comparison school in the SPARC trial (the survey was conducted ca.
2.5 years into the intervention period). Tobacco use. Tobacco use was measured by one item assessing past 30 day use of cigarettes. Responses options were as follows: 0, 1�C2, 3�C5, 6�C9, 10�C19, 20�C29, or all 30 days. Categories were collapsed to form three mutually exclusive categories: nonsmokers (who smoked on none of the past 30 days), nondaily smokers (who smoked on at least 1 but fewer than 30 days), and daily smokers (who reported smoking on all the past 30 days). Alcohol use. Alcohol use was assessed with one item measuring heavy episodic drinking in the past 30 days: four or more drinks in a row for females and five or more drinks in a row for males (coded as yes/no).
Data analyses The goal of the statistical analyses was to identify correlates associated with any exposure to SHS and context-specific exposure in our sample of students attending 4-year colleges. Context-specific exposure was considered using survey items related to three contexts: Entinostat in a car, at home or in the same room, and in a bar or restaurant. School was treated as a random effect in mixed-effects logistic regression analyses to take into account within-school clustering (where students were considered nested within schools). Adjusted odds ratios and 95% CIs were calculated for predictor variables in the multivariable logistic regression analyses. All predictor variables were included in multivariable analyses simultaneously in each model. A two-sided p value of less than .05 was considered to be statistically significant. All analyses were performed using Stata version 10. Results A total of 4,275 students completed the survey; however, data for the items assessed in this paper were available from 4,223 students (98.8%).