Mediated, moderated and direct effects of country of residence, age, and gender on the cognitive and social determinants of adolescent smoking in Spain and the UK: A cross-sectional study
Background. European trans-national adolescent smoking prevention interventions based on social influences approaches have had limited success. The attitudes-social influences-efficacy (ASE) model is a social cognition model that states smoking behaviour is determined by smoking intention which, in turn, is predicted by seven ASE determinants; disadvantages, advantages, social acceptance, social norms, modelling, perceived pressure, self-efficacy. Distal factors such as country of residence, age and gender are external to the model. The ASE model is, thus, closely related to the Theory of Planned Behaviour. This study assessed the utility of the ASE model using cross-sectional data from Spanish and UK adolescents. Methods. In 1997, questionnaires were simultaneously administered to Spanish (n = 3716) and UK adolescents (n = 3715) who were considered at high risk of smoking. Participants' age, gender, smoking intentions and ASE determinant scores were identified and linear regression analysis was used to examine the mediated, moderated and direct effects of country of residence, age and gender on participants' smoking intentions. Results. All UK participants were aged 12 or 13 and most Spanish participants were aged between 12 and 14 (range 12-16 years). Amongst 12 and 13 year olds, regular smoking was more common in Spain. Almost half the participants were female (47.2% in Spain; 49.9% in the UK). Gender did not vary significantly according to age. The distribution of ASE determinant scores varied by country and predicted intention. The influence of each ASE determinant on intention was moderated by country. Country had a large direct influence on intention (1.72 points on a 7 point scale) but the effects of age and gender were mediated by the ASE determinants. The findings suggest resisting peer pressure interventions could potentially influence smoking amongst UK adolescents but not Spanish adolescents. Interventions that promote self-efficacy, on the other hand, would possibly have a greater influence on smoking amongst Spanish adolescents. Conclusion. The ASE model may not capture important cultural factors related to adolescent smoking and the relative contribution of particular ASE determinants to adolescent smoking intentions may differ between countries. Future European trans-national adolescent smoking prevention programmes may benefit from greater undestanding of country-level cultural norms. © 2009 Markham et al; licensee BioMed Central Ltd.