Atkinson, Christina2006-08-222006-08-2220052005http://hdl.handle.net/10012/714Smoking rates among senior students have been related to smoking initiation in younger students. Opportunities to select smoking friends may be one explanation, however our understanding of this process has been limited by cross-sectional designs. <br ><br /> The purpose of this longitudinal study was to determine whether senior student smoking rates a) predict smoking initiation in younger elementary school students, controlling for individual exposure to family and friends who smoke and b) are related to the selection of smoking friends, increasing risk of smoking initiation as a result. <br ><br /> This study involved secondary data analysis of 2798 students from 84 Ontario elementary schools involved with the Third Waterloo Smoking Prevention Project (WSPP3). Grade 8 students completed a questionnaire at baseline to obtain the percentage of senior students who smoke in each school. Students in grade 6 completed a similar questionnaire at baseline, and were surveyed again in grades 7 and 8. Multilevel regression analyses were used to examine school and individual characteristics simultaneously. <br ><br /> Each 5% increase in the senior student smoking rate at a school increased the risk that a non-smoking grade 6 student would try smoking more than once by grade 8 (OR 1. 05) and that a non-smoking grade 6 student with no smoking friends would gain a smoking friend by grade 7 (OR 1. 10). Students who remained non-smokers in grade 7 but gained a smoking friend were more likely to try smoking more than once by grade 8 (OR 4. 31). <br ><br /> In schools where a high proportion of senior students smoked, younger students were more likely to initiate smoking, and gain a smoking friend. Anti-smoking policies and interventions may be more urgently required in these schools to lower senior student smoking rates and reduce initiation among younger students. Tailoring the intensity and content of programs to match the needs of schools is one way to potentially maximize effectiveness.application/pdf427115 bytesapplication/pdfenCopyright: 2005, Atkinson, Christina. All rights reserved.Health Sciencessmoking initiationschool prevalencesocial influencespeer selectionyouthlongtitudinalmultilevel logistic regressionMaster Thesis