The Korean prediction model for adolescents' future smoking intentions.
Lee, Sungkyu;
Yun, Ji Eun;
Lee, Ja Kyoung;
Kim, Il Soon;
Jee, Sun Ha;
(2010)
The Korean prediction model for adolescents' future smoking intentions.
Journal of preventive medicine and public health = Yebang Uihakhoe chi, 43 (4).
pp. 283-291.
ISSN 1975-8375
DOI: https://doi.org/10.3961/jpmph.2010.43.4.283
Permanent Identifier
Use this Digital Object Identifier when citing or linking to this resource.
OBJECTIVES: The purpose of this study was to develop a prediction model for future smoking intention among Korean adolescents aged 13 to 15 in order to identify the high risk group exposed to future smoking. METHODS: The data was collected from a total of 5940 students who participated in a self-administrated questionnaire of a cross-sectional school-based survey, the 2004 Korea Global Youth Tobacco Survey. Chi-square tests and logistic regression analyses were carried out to identify the relevant determinants associated with intentions of adolescents' future smoking. Receiver Operation Characteristic (ROC) assessment was applied to evaluate the explanation level of the developed prediction model. RESULTS: 8.4% of male and 7.2% of female participants show their intentions of future smoking. Among non-smoking adolescents; who have past smoking experience [odds ratio (OR) 2.73; 95% confidence interval (CI) 1.92 - 3.88]; who have intentions of smoking when close friends offer a cigarette (OR 31.47; 95% CI = 21.50 - 46.05); and who have friends that are mostly smokers (OR 5.27; 95% CI = 2.85 - 9.74) are more likely to be smokers in the future. The prediction model developed from this study consists of five determinants; past smoking experience; parents smoking status; friends smoking status; ownership of a product with a cigarette brand logo; and intentions of smoking from close friends' cigarette offer. The area under the ROC curve was 0.8744 (95% CI=0.85 - 0.90) for current non-smokers. CONCLUSIONS: For efficiency, school-based smoking prevention programs need to be designed to target the high risk group exposed to future smoking through the prediction model developed by the study, instead of implementing the programs for all the students.