Adolescent smoking and tobacco use has been linked to a variety of adverse and long-term health consequences, and thus represents a behavior that is an important target of a large number of past and current health communications campaigns bent on inhibiting its growth. “Susceptibility” is an important predictor of adolescent smoking behavior. Until now, most studies pertaining to “susceptibility” to smoking have treated the construct as a dichotomous measure, one classifying adolescents as either “Susceptible” or “Not Susceptible.” This study proposes an adolescent susceptibility to smoking hierarchy in which the construct is divided into distinct degrees of susceptibility: adolescents low on the hierarchy may be expected to be less susceptible than those higher on the hierarchy and those who have Never Smoked may be expected to be less susceptible than those who are Former Smokers.
It is initially proposed here that variables commonly used to predict susceptibility will be systematically related to a hierarchical measure of susceptibility. It is also proposed that multivariate models and the relationships between the predictor variables and susceptibility to smoking will be stronger when using the hierarchical measure of susceptibility than the traditional construct.
Six commonly used predictor variables of susceptibility (age, household member smoking, best friends’ smoking, exposure to pro-tobacco media messages, and receptivity to pro-tobacco promotional items) were used to examine how relationships between these variables and the traditional and hierarchical models changed. Results suggested that household smoking, best friends’ smoking, exposure to pro-tobacco media messages, receptivity to owning pro-tobacco promotional items, and receptivity to wearing pro-tobacco promotional items were all positively related to the hierarchical level of adolescent susceptibility to smoking. Furthermore, multinomial regression procedures revealed that the multivariate model of the hierarchy of susceptibility featured stronger relationships with certain of these predictor variables among adolescents who were susceptible than did the traditional, 2-level susceptibility construct, as evidenced by both overall model fitting information as well as multivariate odds ratios and their confidence intervals. Finally, it was discovered that the percentage of adolescents at either of the four level hierarchical of susceptibility (No, Low, Intermediate, and High) decreased from 2000 to 2002 and from 2002 to 2004 and the percentage of adolescents who were Not Susceptible increased significantly during this time. These findings differed from those obtained when the traditional construct was examined over this time period.
The above findings show that specific risk variables of susceptibility are systematically related to an adolescent’s degree or level of susceptibility and are more strongly related to a hierarchy of susceptibility than they are to the traditional and widely used dichotomous susceptibility construct. These findings are important to health communicators and social marketers. They show that a hierarchy of susceptibility can be predicted by common predictor variables, allowing for the segmentation of adolescents into various audience groups based on their degree of susceptibility. Such segmentation should allow for the more efficient targeting of adolescents for anti-tobacco media messaging and should allow more persuasive messages to be directed at those adolescents most in need (or most susceptible to smoking behavior).