The longitudinal analysis for the association between smoking and the risk of depressive symptoms

Background Despite high smoking rate in people with depressive symptoms, there is ongoing debate about relationship between smoking and depressive symptoms. Methods Study participants were 57,441 Korean men. We collected their baseline data between 2011 and 2012, and conducted follow-up from 2013 to 2019. They were categorized by smoking status (never: < 100 cigarettes smoking in life time, former: currently quitting smoking, and current smoker: currently smoking), smoking amount (pack/day and pack-year) and urine cotinine excretion. The development of depressive symptoms was determined in CES-D score >= 16. Cox proportional hazards model was used to analyze the multivariable-adjusted hazard ratio (HR) and 95% confidence intervals (CI) for depressive symptoms in relation to smoking status, smoking amount, and urine cotinine excretion. Results During 6.7 years of median follow-up, the risk of depressive symptoms increased in order of never (reference), former (HR = 1.08, 95% CI: 1.01-1.15) and current smoker (HR = 1.24, 95% CI: 1.16-1.32). Among current smoker, the risk of depressive symptoms increased proportionally to daily smoking amount (< 1 pack; HR = 1.21, 95% CI: 1.13-1.29, and >= 1 pack; HR = 1.34, 95% CI: 1.23 - 1.45). This pattern of relationship was consistently observed for pack-year in former smoker and current smoker. Additionally, urine cotinine excretion was proportionally associated with the risk of depressive symptoms. Conclusion Exposure to smoking was associated with the increased risk of depressive symptoms. Dose dependent relationship was observed between smoking amount and the risk of depressive symptoms.