Cigarette smoking and drug use in schoolchildren. II. Factors factors associated with smoking

In 1979 as the first stage in the evaluation of an educational strategy to discourage children from smoking, we surveyed over 6000 10-12 year old primary school children in the Hunter Health Region (New South Wales, Australia) to assess the prevalence of smoking and to identify some of the social and personal factors associated with this phenomenon. Logistic regression was used to determine the relative importance of the key factors in relation to the proportion of smokers in the 'recent' and 'regular' smoking categories. Friends' smoking behaviour was the best predictor of the child's smoking status followed in order by approval of advertising, siblings' smoking behaviour, amount of money available to spend weekly, sex, age and finally parents' smoking habits. Other factors such as size of family, position of the child in the family, country of birth of parents, father's occupation, having a single parent and urban or non-urban location of schools were all significantly associated with smoking prevalence in these children. Linear modelling techniques used in the analysis showed girls and non-smokers to have less favouring attitudes to smoking and tobacco advertising than boys and smokers. Significant differences in knowledge occurred between categories of smokers. Most children claiming not to smoke disapproved of advertising and did not intend to smoke in the future. Among current smokers, some with neutral or favouring attitudes were ambivalent about future smoking. Our data suggest that intervention in the field of cigarette advertising may be the most direct and effective means of reducing prevalence of smoking in children