The predictive validity of self-rated alcohol susceptibility to alcohol use in early adolescents in Latin America

Abstract

Background: Measurement of substance use susceptibility was first conceived as part of the preparatory stage preceding youth smoking initiation and is defined as the lack of a firm commitment not to smoke in the future. Despite being a consistent and validated predictor of smoking initiation, little research has investigated whether susceptibility can predict alcohol use. This study aimed to assess the predictive validity of an adapted alcohol susceptibility measure among early adolescent students in Argentina and Mexico who had not previously consumed alcohol. Methods: A school-based longitudinal study was conducted among first-year students in 33 secondary schools in Argentina and 57 in Mexico. The baseline sample included 1,504 never-drinker adolescents in Argentina and 5,264 in Mexico, of whom 1,055 and 3,540, respectively, completed a follow-up survey a year and a half later. Logistic regression with school as a random effect was used to estimate adjusted odds ratios for the transitions from never-drinker to ever-drinker, current drinker, and binge drinker. Results: At baseline, 34% and 23.6% of adolescent never-users were susceptible to alcohol in Argentina and Mexico, respectively. After controlling for other known predictors of alcohol use initiation, alcohol susceptibility was positively associated with ever-drinking (OR = 3.23, 95% CI 2.38-4.36 in Argentina, OR = 1.73, 95% CI 1.43-2.10 in Mexico), current drinking (OR = 2.41, 95% CI 1.71-3.4 in Argentina, OR = 1.77, 95% CI 1.37-2.28 in Mexico), and binge drinking (OR = 2.27, 95% CI 1.78-2.91 in Argentina, OR = 1.89, 95% CI 1.32-1.99 in Mexico). Conclusions: The adapted susceptibility measure for alcohol appears valid for identifying individuals or groups at risk of drinking initiation and problematic drinking among early adolescents in Argentina and Mexico.