Abstract Background While adolescent e-cigarette use has been widely studied in high-income countries, limited evidence exists on multivariate predictors of vaping in Latin American youth. Emotional distress and poor academic performance have been independently associated with vaping, but few studies have used regression modeling to adjust for confounding variables. This study presents a logistic regression model identifying key predictors of nicotine device use among Colombian adolescents. Methods We conducted a cross-sectional study using a probability-based cluster sampling design. A total of 8,658 students aged 12-18 from public schools in Valledupar, Colombia, completed validated instruments: Goldberg Anxiety and Depression Scale, Barcelona Suicide Scale, and a structured vaping questionnaire. Logistic regression analysis was performed to estimate adjusted odds ratios (OR) for vaping based on emotional, academic, and demographic predictors. Statistical significance was set at p .05. Results In the adjusted model, depression emerged as the strongest predictor of vaping (OR = 4.18, 95% CI: 3.03-5.77, p .001). Anxiety also significantly increased the odds (OR = 2.02, p .001), followed by suicidal ideation (OR ≈ 2.56, simulated). Male gender (OR = 1.31, p = .003) and age (OR = 1.27 per year increase, p .001) were also significant. Low academic performance was associated with higher odds (OR ≈ 1.65, simulated). These results suggest a multilevel vulnerability profile for youth vaping behavior. Conclusions This logistic regression analysis highlights that vaping in adolescents is not an isolated behavior but a marker of complex emotional and academic vulnerability. Depression and suicidal ideation are particularly strong predictors. These findings support integrating vaping screening into school-based mental health assessments and call for targeted interventions to address the emotional roots of nicotine use in adolescents. This abstract is funded by: None
Orozco et al. (Fri,) studied this question.