Education tracking and adolescent smoking: a counterfactual and prospective cohort study

Abstract

Background and aims: Adolescent smoking is a health issue and a potential health inequality issue. Education tracking, which is the placement of students into different school types and curricula based on their learning needs or abilities, is an indicator of inequality and risk factor of adolescent smoking. We examined the effect of educational tracking, dividing students into vocational and academic high school tracks, on adolescent smoking in Taiwan.

Design and setting: Longitudinal panel data, collected annually from 2000 over a period of 6 years as part of the Taiwan Youth project, were used.

Participants: Adolescents (aged 13-18 years) from the first six waves of the Taiwan Youth Project were included in the project, of whom 2147 had clear information on track attendance in 10th grade, control variables in 7th/8th grades and smoking behavior in 8th grade (before track placement). Post-track smoking behavior was measured at 10th, 11th and 12th grades.

Measurements: The outcome variable was the self-reported smoking status in the 8th grade and between 10th and 12th grades. The treatment variable of interest was education tracking (vocational versus academic), which was conducted when the student was in 10th grade. Several important confounders were used for the difference-in-differences propensity score matching (e.g. parents' education and same classroom peer smoking).

Findings: Placement of a student in the vocational track increased the proportion of smokers by 3.3 percentage points in 10th grade (P = 0.039). The effect was even more pronounced in 11th grade (6.2 percentage points; P = 0.000) and 12th grade (5.9 percentage points; P = 0.003).

Conclusions: Education tracking (placement of students into different school curricula based on learning needs or abilities) appears to be a risk factor for adolescent smoking among Taiwanese adolescents.

Keywords: Adolescent; education tracking; health inequality; propensity score matching; smoking; vocational education.