• Adolescent;
  • child;
  • longitudinal studies;
  • parent;
  • risk factor;
  • smoking


Aims  To examine the long-term effects of childhood smoking experimentation and exposure to parental smoking on adult smoking risk.

Methods  Data were from a 20-year follow-up of 9–15-year-olds who completed questionnaires in the 1985 Australian Schools Health and Fitness Survey (n = 6559). The relative risks (RR) of adult current smoking in 2004–05 for childhood exposure to smoking experimentation (never, a few puffs, < 10 cigarettes, > 10 cigarettes) and parental smoking (none, father, mother, both parents) in 1985, with adjustment for confounders, were estimated by log binomial modelling. Analyses were stratified by age (9–13 and 14–15 years) and sex.

Findings  Participation at follow-up was 54% (n = 3559). Childhood smoking experimentation increased the risk of being a current smoker particularly for 14–15-year-old experimenters of more than 10 cigarettes [males, RR 2.72, 95% confidence interval (CI) 1.74–4.25; females, RR 6.39, 95% CI 2.85–14.33]. Parental smoking was associated with adult current smoking risk, particularly for 9–13-year-olds with two smoking parents (males, RR 1.53, 95% CI 1.19–1.96; females, RR 1.99, 95% CI 1.52–2.61) and older males with smoking mothers (RR 1.82, 95% CI 1.22–2.73). Parental smoking was not associated with childhood smoking experimentation.

Conclusions  These findings suggest that any childhood smoking experimentation increases the risk of being a smoker 20 years later. As exposure to parental smoking predicted current smoking, parents should be aware of the association between their own smoking behaviour and that of their children.