2.4. Statistical Analysis

The prevalence of socioeconomic, occupational, and lifestyle characteristics was compared with work schedule using the chi-square test. The generalized estimating equation was used to analyze the association between changes in work schedule and changes in smoking and/or drinking status after adjusting for age in the observation year, sex, education level, occupational classification, monthly wage income level, and self-rated health status. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. All analyses were performed using SAS, version 9.4 (SAS Institute, Cary, NC, USA). For all statistical calculations, a two-tailed p-value of < 0.05 was considered statistically significant.