Smoking Paradox in Patients Hospitalized With Coronary Artery Disease or Acute Ischemic Stroke: Findings From Get With The Guidelines.
Smoking is a potent risk factor for coronary artery disease (CAD) and acute ischemic stroke (AIS), but there are numerous reports of lower in-hospital mortality among smokers versus nonsmokers hospitalized for these events.We analyzed all consecutive patients hospitalized with a first index CAD (n=158 054) or AIS (n=899 295) event in Get With The Guidelines from 2002 to 2012; 20.4% of AIS and 30.4% of patients with CAD were past-year smokers. Multivariable models and age-stratified analyses were used to estimate the adjusted odds ratio of in-hospital mortality in smokers versus nonsmokers. Smokers were younger, more often male, with fewer vascular risk factors, and were more likely to be admitted to hospitals that were large, academic, or in the South. In-hospital mortality was significantly lower among smokers in both CAD (2.7% versus 5.2%; P<0.0001) and AIS (3.5% versus 5.8%; P<0.0001). The difference between unadjusted and adjusted odds ratios for smoking (0.57 versus 0.86 in CAD; 0.56 versus 0.86 in AIS) indicates the presence of substantial confounding by age and other covariates, but a significant association of past-year smoking remained.Among patients hospitalized with CAD and AIS, smoking is a risk factor for early age of onset, even among those with few vascular risk factors. The persistent association with lower in-hospital mortality after adjusted and stratified analyses probably represents residual unmeasured confounding, although a biological effect of smoking cannot be excluded. Further clinical and prospective population-based studies are needed to explore variables that contribute to outcomes in these patients.
Ali, SF; Smith, EE; Reeves, MJ; Zhao, X; Xian, Y; Hernandez, AF; Bhatt, DL; Fonarow, GC; Schwamm, LH
Volume / Issue
Start / End Page
Electronic International Standard Serial Number (EISSN)
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)