Introduction: Simulation modeling is an innovative approach for understanding the social dynamics of adolescent cigarette smoking. Numerous studies demonstrate positive relationships between adolescent cigarette smoking and both parental smoking and access to cigarettes in the home. We utilize a simulation approach to observe the effects of experimental manipulations that alter parental smoking, peer influence, and youths’ access to cigarettes to understand the consequences for youths’ networks and smoking behavior. This approach allows for the identification of risk and protective targets for preventive interventions.
Methods: We utilize Stochastic Actor-Based Models to examine the co-evolution of smoking behavior and friendship tie choice, accounting for parental smoking and the level of cigarette availability in the home. We next manipulated the parameters and levels of parental smoking, peer influence, and access to cigarettes in the home, and simulated the network forward 1000 times in each condition and assessed the consequences for the network and smoking behavior. Data are from the two largest schools (n= 2,178 and n=976, respectively) across three waves of the National Longitudinal Study of Adolescent Health saturation sample surveys, including Wave 1 interviews occurring in school (1994 – 1995), Wave 2 interviews occurring at youths’ homes from April-December 1995, and Wave 3 interviews also occurring at home one year later.
Results: Whereas cigarette availability in the home has a positive effect on smoking behavior, parental smoking does not relate to adolescent smoking in the models. A one unit increase in home cigarette availability increases the odds of increased smoking behavior 14% or 23% in both schools. The simulation results for a hypothetical intervention that reduced the percentage of homes with cigarette availability found stronger effects in the high substance use school. The percentage of non-smokers rises from 73% to 74% for a 50% reduction in home cigarette availability in the low smoking school, but it rises from 48% to 54% in the high smoking school (and the percentage of heavy smokers falls from 30% to 25%).
Conclusions: Our findings indicate that cigarette availability in the home not only has a direct positive relationship with the smoking, but our simulation demonstrates that this then propagates through the network and has consequences for school levels of smoking.