Stage-of-change of smoking acquisition in South African high-school adolescents : a cross-sectional study of decisional balance, temptation and perceived social norms.
This study aimed to use the Transtheoretical Model of behaviour change (Prochaska and DiClemente, 1983) to investigate the process of smoking initiation in a cross-section of high school adolescents in urban KwaZulu-Natal. This aim was informed by the increase in incidence of smoking among adolescents, who represent a particularly vulnerable population to smoking (Orlandi and Dalton, 1998). The Transtheoretical Model has proved successful in changing problem health behaviours (prochaska and DiClemente, 1983) and lends itself to be a suitable framework for investigating smoking acquisition in adolescents (Werch and DiClemente, 1994). An extensive review of the causes and correlates of smoking uptake and past intervention evaluations suggests that the core constructs of the TTM (Decisional Balance, Temptation and Stage of Acquisition) can be complemented by other another variable, Perceived Social Norms (informed by the Theory of Planned Behaviour, Fishbein and Ajzen, 1975, as cited in Pitts, 1996), to provide a more holistic explanation of smoking acquisition. Alcohol use, parental smoking and certain demographic factors are also found to be salient factors in this process. Findings showed that the perceived cons of smoking were constant across stages and seemed to have no effect on stage membership. Stage differences were explained almost entirely in terms of pros, which increased drastically with later stages. Perceived social norms increased with a later stage, confirming a tentative theoretical relationship between the Transtheoretical Model and Theory of Planned Behaviour. Lack of expressed intention to smoke by participants questions the validity of using the rational decision-making Transtheoretical Model to investigate a process not informed by decision-making. European language speakers were found to be a particularly vulnerable group to smoking, while African language speaking girls show very low rates of smoking. High religiosity was found to be a protective factor, while alcohol use was strongly associated with smoking. Maternal smoking was strongly associated with smoking, but only in girls. No difference in stage was found between schools and grades. The study should be replicated using a longitudinal design to determine the causal relationship between factors and smoking and to further investigate the applicability of the Transtheoretical Model in smoking acquisition.