Smoking in acute coronary syndromes--the "smoker's paradox" revisited.

INTRODUCTION Although a well-known risk factor for coronary disease, smoking has long been associated with lower short-term mortality in acute coronary syndromes (ACS). There are few recent works on Portuguese populations examining all aspects of smoking in ACS, particularly the interaction between smoking and other risk factors, and the management and prognosis of patients according to smoking status. OBJECTIVE We sought to examine clinical characteristics, presentation, in-hospital treatment, angiographic features and prognosis of patients with and without smoking history admitted with ACS. METHODS A total of 1228 patients consecutively admitted with ACS from January 2004 to March 2007 were analyzed. Patients were classified into two groups, those with present or past smoking habits (n=450) making up Group I and those without smoking habits (n=778), Group II. The main outcome analyzed was overall mortality during hospital stay and at 6 months. RESULTS Smokers and former smokers were younger and more frequently male (odds ratio [OR] = 22.46; 95% confidence interval [CI]: 12.94-38.96), and less often had diabetes (OR = 0.41; 95% CI: 0.30-0.54), hypertension (OR = 0.31; 95% CI: 0.24-0.39) and renal insufficiency (OR = 0.26; 95% CI: 0.18-0.36). Patients with smoking habits more frequently presented with ST elevation (OR = 1.32; 95% CI: 1.04-1.67), more often received evidence-based medical therapy, namely beta blockers (during hospital stay, OR = 2.42; 95% CI: 1.63-3.56 and at discharge, OR = 1.45; 95% CI: 1.03-2.1) and statins (at discharge, OR = 2.48; 95% CI: 1.2-6.1), and more frequently underwent coronary angiography (OR = 2.15; 95% CI: 1.63-2.84). Although smokers and former smokers had lower in-hospital mortality on univariate analysis (OR = 0.54; 95% CI: 0.31-0.96), this association was not confirmed on multivariate analysis, with adjustment for known short-term mortality predictors (OR = 1.25; 95% CI: 0.61-2.54). Similarly, multivariate analysis failed to confirm lower 6-month mortality for smokers and former smokers (OR = 2.0; 95% CI: 1.17-3.41). CONCLUSIONS Clinical characteristics and management options differed between ACS patients with and without smoking habits. These differences explained the lower shortterm mortality initially observed between the two groups. In our population of patients admitted with ACS, we did not find a real "smoker's paradox".