Gender Differences in the Longitudinal Association of Internalizing and Externalizing Behaviors on Cigarette Smoking Among Chilean Adolescents

Background: Cigarette smoking has been consistently declining among U.S. adolescents (Johnston et al., 2014), however, in many middle and low-income countries smoking has not only remained steady but also risen (Ng. et al., 2014).  In Chile, where the present study takes place, the prevalence of cigarette use is the highest in South America and a greater percent of adolescent girls smoke than boys (CICAD, 2015).  Although factors associated with the onset of cigarette smoking span many domains (e.g., genetic, psychological, cultural), we focus on internalizing and externalizing behaviors because of the general lack of attention gender differences have received in middle and low-income countries. We present the results of a prospective study that examined main and interaction effects of internalizing and externalizing behaviors by gender on the onset and present use of cigarettes among Chilean adolescents.

Methods: Sample: Data are from the Santiago Longitudinal Study, a NIDA-funded study of substance use and mental health behaviors among middle to low-income urban adolescents in Chile. Data are from two waves of assessment, when adolescents were 10 years old and at follow up, 12-17 years of age.

 

Measures: Youth and their parents were administered a battery of measures at both points in time. Primary predictors are parental reports of their 10 year old internalizing and externalizing behaviors (CBCL). The outcomes of interest were assessed at follow up when adolescents answered many questions about cigarette use such as age first offered, if smoked when first offered, lifetime and past month use.

 

Analyses: The analytic sample consists of 607 adolescents who reported ever being offered a cigarette at age 11 or later and who had data on all of the study variables at both time points. Data were analyzed with logistic regression. Continuous variables were mean centered to aid in the interpretation of interaction effects and analyses controlled for numerous covariates.

Results: There was no difference in the probability (P) of accepting a first cigarette offer between boys (P=.44) and girls (P=.42) who were high in internalizing symptoms (+1SD) but girls were significantly more likely to have smoked when first offered (P=.60) than boys (P=.40) if they were low in internalizing symptoms (-1SD). Similarly, current smoking was more likely among girls with low vs. high internalizing symptoms but among boys it did not vary by degree of internalizing. Externalizing symptoms at age 10 predicted smoking at first offer and being current smoker but gender did not moderate the associations.

 

Conclusions and Implications:

The findings that externalizing behaviors operate similarly among Chilean boys and girls while low-internalizing is a risk factor for girls but not boys provides a more nuanced explanation of smoking behaviors among adolescents. From a social learning model, internalizing symptoms may be indirectly protective for girls by decreasing their exposure to peer smoking. The findings further contribute to the field’s understanding of gender differences in smoking risk factors among understudied populations and highlights the importance of conducting cross-cultural research to further develop context-informed prevention interventions.