The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #54334

Ying-Yeh Chen, MD, MS¹, S.V. Subramanian, PhD², Dolores Acevedo-Garcia, PhD², and Ichiro Kawachi, PhD³. (1) Department of Health and Social Behavior, Harvard University, School of Public Health, Kresge Building 7th floor, 677 Huntington Avenue, Boston, MA 02115, 617-5471844, ychen@hsph.harvard.edu, (2) School of Public Health, Dept of Society, Human Development & Health, Harvard University, Kresge Building, 7th Floor, 677 Huntington Avenue, Boston, MA 02115, (3) Professor, Department of Health and Social Behavior, School of Public Health, Harvard University, Kresge Building, 7th Floor, 677 Huntington Avenue, Boston, MA 02115

Objective: To examine the influence of U.S. state contexts on the prevalence and intensity of smoking and binge-drinking behaviors among young adults. Specifically, we tested whether states that have higher prevalence smoking and binge-drinking were also associated with higher levels of cigarette and alcohol consumption. Method: We developed a mixed multivariate multilevel structure of health behaviors at level-1 (smoking status, cigarettes consumption among smokers, binge-drinking status and binge-drinking occasions among binge-drinkers) nested within16,227 individuals (level-2) within 50 states (level-3). We applied this structure to the 1999 Behavior Risk Factor Surveillance System data on subjects aged between 18~25. Results: States that have higher proportions of smokers have higher proportion of binge-drinkers after controlling for individual characteristics, and vice versa (r=0.50, p<.05). Smokers who reside in states with higher number of smokers tend to consume higher number of cigarettes (r=0.61, P<.05); however, binge-drinkers living in states that have higher proportions of binge-drinkers are not associated with more binge-drinking occasions (r=0.34, p=0.95). At the individual level, higher educational attainment is associated with lower probability of smoking, lower intensity of cigarettes consumption and higher probability of binge drinking. Level of education is not associated with intensity of binge drinking. Conclusion: The empirical results suggest the need to consider the health behaviors, not simply within their context, but do so by explicitly recognizing and modeling their co-occurrence.

Learning Objectives:

• 1. Recognize that the so-called ˇ§life-style behaviorsˇ¨, including smoking and binge-drinking, is not merely the attributes of individuals; instead, social-contextual factors can facilitate the co-occurrence of the unhealthy collective behavioral pattern.
• 2. Apply a mixed multivariate-multilevel methodology to analyze the social determinants to the quantitative and qualitative aspects of health-related behaviors.

Keywords: Alcohol Use, Smoking

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.