Abstract

Objective: To analyse the short and long term prognostic significance of admission glycaemia in a large registry of non-diabetic patients with acute myocardial infarction.

Methods: Assessment of short and long term prognostic significance of admission blood glucose in a consecutive population of 1604 non-diabetic patients admitted to intensive care units in France in November 2000 for a recent (⩽ 48 hours) myocardial infarction.

Results: In-hospital mortality, compared with that of patients with admission glycaemia below the median value of 6.88 mmol/l (3.7%), rose gradually with each of the three upper sextiles of glycaemia: 6.5%, 12.5% and 15.2%. Conversely, one year survival decreased from 92.5% to 88%, 83% and 75% (p < 0.001). Admission glycaemia remained an independent predictor of in-hospital and one year mortality after multivariate analyses accounting for potential confounders. Increased admission glycaemia also was a predictor of poor outcome in all clinical subsets studied: patients without heart failure on admission, younger and older patients, patients with or without reperfusion therapy, and patients with or without ST segment elevation.

Conclusion: In non-diabetic patients, raised admission blood glucose is a strong and independent predictor of both in-hospital and long term mortality.