Using Latent Class Analysis to Examine Susceptibility to Various Tobacco Products Among Adolescents


Introduction: Most tobacco use is initiated during adolescence and susceptibility is an important predictor of initiation. Research is needed to understand susceptibility across tobacco products among youth to inform prevention efforts.

Aims and Methods: We surveyed 2945 students from four Southeastern Connecticut high schools in 2017. We assessed susceptibility to future use of seven different tobacco products (ie, cigarettes, e-cigarettes, hookah, cigars, cigarillos, blunts, and smokeless tobacco) among youth who had never used a tobacco product. Susceptibility was coded as any response other than “definitely not” to three items assessing (1) curiosity, (2) willingness to experiment with the product in the future, and (3) willingness to try if offered by a friend. Latent class analysis was used to identify patterns of susceptibility across multiple tobacco products and sociodemographic predictors of these classes.

Results: Among n = 1246 youth who had never tried a tobacco product, 51.2% (n = 638) were susceptible to at least one product and 29.1% (n = 363) were susceptible to two or more products (average among susceptible youth: M = 2.4 products, SD = 1.7, range 1–7). Latent class analysis identified three classes: youth who were: (1) susceptible to all products (5.6%), (2) susceptible to e-cigarettes, hookah, and blunts (23.6%), and (3) mostly non-susceptible to all products (70.8%). In a multivariable model, demographic variables and peer tobacco use predicted class membership.

Conclusions: We observed high rates of susceptibility to non-traditional tobacco products among youth, and many youth were susceptible to more than one product. Targeting susceptibility across multiple tobacco products may be important for reducing youth tobacco use.

Implications: Patterns of polytobacco use susceptibility were identified among youth who had never previously used tobacco products. Many youth were susceptible to more than one product and we observed three distinct classes of susceptibility among youth: those who were mostly non-susceptible, those who were susceptible to alternative or non-traditional tobacco products (e-cigarettes, hookah, and blunts), and those who were susceptible to all tobacco products. These findings suggest the importance of addressing susceptibility across multiple tobacco products.


Oxford University Press

Publication Date


Publication Title

Nicotine & Tobacco Research



Document Type




Adolescent, Demography, Tobacco, Tobacco use, Waterpipes, Electronic cigarettes, Latent class analysis