THE SOCIAL AND DEMOGRAPHIC CORRELATES OF SMOKING AMONG YOUNG-ADULTS IN MARYLAND

Objectives: We examine the utility of pooling data from the 1987 throu gh 1994 Maryland Behavioral Risk Factor Surveillance System (BRFSS) su rveys in order to increase sample size, and we investigate the social and demographic correlates of smoking in Maryland among 18-24-year-old s. Methods: The data are from 1,714 subjects who were between 18 and 2 4 years of age and the principle method was multiple logistic regressi on. Results: African Americans are less likely to smoke than Caucasian s, and people with some college or more are less likely to smoke than those with a high school diploma/GED or less. Men who are in an inform al sexual union (cohabitation) are more likely to smoke than men who a re not. There is a negative association between being a student and sm oking among all young women, which does not exist for men enrolled in postsecondary schools. There is no gender difference in the level of s moking, despite the fact that the predictors of smoking are somewhat d ifferent for men and women. Conclusions: Pooling data from several BRF SS studies is a useful way to increase sample size. Analyses of the co rrelates of smoking among people from a narrow age range is a useful w ay of highlighting the unique correlates of smoking across the life co urse.