Background: In general having a parent who smokes or smoked is a strong and consistent predictor of smoking initiation among their children while authoritative parenting style, open communication that demonstrates mutual respect between child and parent, and parental expectations not to smoke are protective. It has been hypothesized that parental smoking affects their children's smoking initiation through both imitation of the behavior and effects on attitudes toward smoking. The goals of the current analysis were to examine these two potential mechanisms.
Methods: In 2003, 1,417 high school students in Houston, Texas, completed a cross-sectional survey as part of the evaluation of an interactive smoking prevention and cessation program delivered via CD-ROM. To assess the relationship between number of parents who currently smoke and children's smoking status, we completed an unconditional logistic regression. To determine whether the attitudes that children of smokers hold toward smoking are significantly more positive than the attitudes of children of non-smokers we examined whether the parents smoking status moderated the relationship between children's attitudes toward smoking and their ever smoking using unconditional logistic regressions.
Results: Compared to participants whose parents did not currently smoke, participants who reported one or both parents currently smoke, had increased odds of ever smoking (OR = 1.31; 95% CI: 1.03-1.68; Wald chi2 = 4.78 (df = 1) p = 0.03 and OR = 2.16; 95% CI: 1.51-3.10; Wald chi2 = 17.80 (df = 1) p < 0.001, respectively). In addition, the relationship between attitudes and ever smoking was stronger among participants when at least one parent currently smokes (OR = 2.50; 95% CI: 1.96-3.19; Wald chi2 = 54.71 (df = 1) p < 0.001) than among participants whose parents did not smoke (OR = 1.72; 95% CI: 1.40-2.12; Wald chi2 = 26.45 (df = 1) p < 0.001).
Conclusion: Children of smokers were more likely to smoke and reported more favorable attitudes toward smoking compared to children of non-smokers. One interpretation of our findings is that parental smoking not only directly influences behavior; it also moderates their children's attitudes towards smoking and thereby impacts their children's behavior. Our results demonstrate a continued need for primary prevention smoking interventions to be sensitive to the family context. They also underscore the importance of discussing parental smoking as a risk factor for smoking initiation, regardless of ethnicity, and of tailoring prevention messages to account for the influence that parental smoking status may have on the smoking attitudes and the associated normative beliefs.