The Role of Diabetes Mellitus as an Effect Modifier of the Association Between Smoking Cessation and Its Clinical Prognoses: An Observational Cohort Study.

The smoker's paradox refers to an increased risk of adverse clinical outcomes after smoking cessation in patients with coronary artery disease. The mechanisms involved are controversial. The present study evaluated the effect of delay in smoking cessation on clinical outcomes among patients after percutaneous coronary intervention (PCI) stratified by diabetes mellitus (DM). Patients included in this study came from an established Fu Wai hospital PCI cohort. Smoking behavior was recorded; clinical end points included all-cause mortality and repeat revascularization. The analyses were based on 8489 smokers who underwent PCI. Patients with and without DM were examined separately. Multivariable model analysis suggested that smoking cessation was associated with significant lower all-cause mortality both for non-DM and DM patients. The smoking paradox was observed for revascularization. However, the increased risk of repeat revascularization correlated with quitting time among non-DM patients only, especially if they stopped smoking late (>90 days) after their index procedure (adjusted hazard ratio, 3.40; 95% CI: 2.45-4.72). In conclusion, smoking cessation is associated with a lower mortality rate for PCI patients. However, the relative benefit on repeated revascularization was only observed among non-DM patients if they quit smoking early.