A smoker's paradox in patients hospitalized for heart failure: findings from OPTIMIZE-HF.

Journal Article (Journal Article;Multicenter Study)

AIMS: Cigarette smoking is a well-established risk factor for cardiovascular disease yet several studies have shown lower mortality after acute coronary syndromes in smokers compared with non-smokers, the so called 'smoker's paradox'. This study aimed to ascertain the relationship between smoking and clinical outcomes in patients hospitalized with heart failure (HF). METHODS AND RESULTS: OPTIMIZE-HF (Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure) collected data on 48 612 patients from 259 hospitals. Characteristics, treatments, and outcomes were compared for current/recent smokers vs. those without current/recent smoking, and multivariable regression analyses with adjustment for hospital clustering were performed. There were 7743 (15.9%) smokers, 39 126 (80.5%) non-smokers, and 1743 (3.6%) missing. Smokers were younger, had similar renal function, but lower ejection fraction. The risk of in-hospital mortality was less in smokers (2.3 vs. 3.9%, P < 0.001). After extensive covariate adjustment, smokers still had lower in-hospital mortality risk OR (odds ratio) 0.70, 95% CI (confidence interval) 0.56-0.88, P = 0.002. Post-discharge, smokers (n = 998) had similar mortality risk (6.7 vs. 8.4%, P = 0.29) compared with those without current/recent smoking. CONCLUSION: Smokers hospitalized with HF had lower risk adjusted in-hospital mortality and similar early post-discharge mortality compared with non-smokers. The residual association of smoking and better prognosis, the 'smoker's paradox', was not fully explained by measured covariates.

Full Text

Duke Authors

Cited Authors

  • Fonarow, GC; Abraham, WT; Albert, NM; Stough, WG; Gheorghiade, M; Greenberg, BH; O'Connor, CM; Nunez, E; Yancy, CW; Young, JB

Published Date

  • August 2008

Published In

Volume / Issue

  • 29 / 16

Start / End Page

  • 1983 - 1991

PubMed ID

  • 18487210

Electronic International Standard Serial Number (EISSN)

  • 1522-9645

Digital Object Identifier (DOI)

  • 10.1093/eurheartj/ehn210


  • eng

Conference Location

  • England