BACKGROUND: Although some professional societies recommend target-driven glucose control for all hospitalized patients, the association between elevated glucose and adverse outcomes has not been well established in patients hospitalized with heart failure. METHODS AND RESULTS: We evaluated a nationally representative cohort of 50,532 elderly patients hospitalized with heart failure in the United States between April 1998 and June 2001. The association between admission glucose and all-cause mortality at 30 days and 1 year was analyzed with multivariable Cox regression models, both in the entire cohort and in patients with and without diabetes mellitus. After multivariable adjustment, no significant relationship was found between glucose and 30-day mortality (for glucose groups of >110 to 140, >140 to 170, >170 to 200, and >200 mg/dL; hazard ratios for 30-day mortality were 1.09 (95% confidence interval, 0.98 to 1.22), 1.27 (95% confidence interval, 1.11 to 1.45), 1.16 (95% confidence interval, 0.98 to 1.37), and 1.00 (95% confidence interval, 0.87 to 1.15), respectively, versus glucose < or =110 mg/dL; P for linear trend=0.53). Results were similar for 1-year mortality and did not differ between patients with and without known diabetes mellitus (for diabetesxglucose interaction, P=0.11 and 0.55 for 30-day and 1-year mortality, respectively). CONCLUSIONS: We found no significant association between admission glucose levels and mortality in a large cohort of patients hospitalized with heart failure. Our findings suggest that the relationship between hyperglycemia and adverse outcomes seen in acute myocardial infarction cannot be automatically extended to patients hospitalized with other cardiovascular conditions.
BACKGROUND: Although some professional societies recommend target-driven glucose control for all hospitalized patients, the association between elevated glucose and adverse outcomes has not been well established in patients hospitalized with heart failure. METHODS AND RESULTS: We evaluated a nationally representative cohort of 50,532 elderly patients hospitalized with heart failure in the United States between April 1998 and June 2001. The association between admission glucose and all-cause mortality at 30 days and 1 year was analyzed with multivariable Cox regression models, both in the entire cohort and in patients with and without diabetes mellitus. After multivariable adjustment, no significant relationship was found between glucose and 30-day mortality (for glucose groups of >110 to 140, >140 to 170, >170 to 200, and >200 mg/dL; hazard ratios for 30-day mortality were 1.09 (95% confidence interval, 0.98 to 1.22), 1.27 (95% confidence interval, 1.11 to 1.45), 1.16 (95% confidence interval, 0.98 to 1.37), and 1.00 (95% confidence interval, 0.87 to 1.15), respectively, versus glucose < or =110 mg/dL; P for linear trend=0.53). Results were similar for 1-year mortality and did not differ between patients with and without known diabetes mellitus (for diabetesxglucose interaction, P=0.11 and 0.55 for 30-day and 1-year mortality, respectively). CONCLUSIONS: We found no significant association between admission glucose levels and mortality in a large cohort of patients hospitalized with heart failure. Our findings suggest that the relationship between hyperglycemia and adverse outcomes seen in acute myocardial infarction cannot be automatically extended to patients hospitalized with other cardiovascular conditions.