INTRODUCTION: It is held that younger smoking initiates are more likely to become regular smokers. The definitions of smoking initiation (a puff, part of a cigarette, a whole cigarette) are inconsistent and raise questions about the robustness of the view. We sought to re-examine the relationship using adolescent smoking data from 3 European countries. METHODS: A stratified secondary, logistic regression analysis of Global Youth Tobacco Survey data was conducted using a design-based analysis. Subgroup analyses were conducted of 13- to 15-year olds from Latvia (high smoking prevalence), Slovenia (moderate prevalence), and Montenegro (low prevalence) who had initiated smoking. The outcome was current smoking--smoking everyday for the past 30 days, or smoking 10 or more days in the past 30 days. Smoking initiation was operationalized as a single puff of a cigarette, and age of smoking initiation was a derived continuous measure. RESULTS: In Latvia, there was a significant association between age of smoking initiation and current smoking for males (p < .05) and females (p < .001) when smoking was operationalized as smoking every day. It was only significant in female adolescents (p < .001) for smoking 10 or more days. In Slovenia and Montenegro, there was no significant relationship between age of smoking initiation and current smoking for either males or females. CONCLUSIONS: The evidence about the relationship between age of smoking initiation and current smoking is not clear. Explanations for the findings may relate to a lack of power, the specificity of the measure, or problems with the theory.
INTRODUCTION: It is held that younger smoking initiates are more likely to become regular smokers. The definitions of smoking initiation (a puff, part of a cigarette, a whole cigarette) are inconsistent and raise questions about the robustness of the view. We sought to re-examine the relationship using adolescent smoking data from 3 European countries. METHODS: A stratified secondary, logistic regression analysis of Global Youth Tobacco Survey data was conducted using a design-based analysis. Subgroup analyses were conducted of 13- to 15-year olds from Latvia (high smoking prevalence), Slovenia (moderate prevalence), and Montenegro (low prevalence) who had initiated smoking. The outcome was current smoking--smoking everyday for the past 30 days, or smoking 10 or more days in the past 30 days. Smoking initiation was operationalized as a single puff of a cigarette, and age of smoking initiation was a derived continuous measure. RESULTS: In Latvia, there was a significant association between age of smoking initiation and current smoking for males (p < .05) and females (p < .001) when smoking was operationalized as smoking every day. It was only significant in female adolescents (p < .001) for smoking 10 or more days. In Slovenia and Montenegro, there was no significant relationship between age of smoking initiation and current smoking for either males or females. CONCLUSIONS: The evidence about the relationship between age of smoking initiation and current smoking is not clear. Explanations for the findings may relate to a lack of power, the specificity of the measure, or problems with the theory.