Abstract
BACKGROUND: Higher smoking prevalence in less educated persons and manual workers is well known. This study examines the independent relationship of education and occupation with tobacco use. METHODS: We used anonymized data from a nationwide population survey (30,617 men and 33,934 women). Education was divided into junior high school, high school, or university attainment. Occupation was grouped into upper non-manual, lower non-manual, and manual. Poisson regression models stratified by age and gender were used to estimate adjusted prevalence ratio (PR) and 95% confidence interval (CI) for current smoking. RESULTS: After adjustment for covariates, education, and occupation, education was significantly related to current smoking in both genders; compared to university graduates, PRs of junior high school graduates aged 20-39, 40-64, and ≥65 were 1.74 (95% CI, 1.53-1.98), 1.50 (95% CI, 1.36-1.65), and 1.28 (95% CI, 1.08-1.50) among men, and 3.54 (95% CI, 2.92-4.30), 2.72 (95% CI, 2.29-3.23), and 1.74 (95% CI, 1.14-2.66) among women, respectively. However, significantly higher smoking prevalence in manual than in upper non-manual was found only in men aged 20-64; compared to upper non-manual, the PRs of manual workers aged 20-39, 40-64, and ≥65 were 1.11 (95% CI, 1.02-1.22), 1.18 (95% CI, 1.10-1.27), and 1.10 (95% CI, 0.89-1.37) among men, and 0.95 (95% CI, 0.75-1.20), 0.92 (95% CI, 0.75-1.12), and 0.46 (95% CI, 0.22-0.95) among women, respectively. CONCLUSIONS: Independent of occupation, educational disparities in smoking existed, regardless of age and gender. Occupation-smoking relationship varied with age and gender. Our study suggests that we should pay attention to social inequality in smoking as well as national smoking prevalence.