[The predictive value of admission hyperglycemia in hospital and long-term mortality of diabetics versus non-diabetics with acute coronary syndrome].

OBJECTIVE

To evaluate whether the predictive value of admission hyperglycemia for mortality differs between diabetics and non-diabetics with acute coronary syndrome (ACS).

METHODS

A total of 1534 consecutive patients admitted with ACS to Central Control Unit of Beijing Friendship Hospital were analyzed. Among these patients, 1332 patients who had a venous plasma glucose record at admission were retrospectively enrolled and were stratified according to admission glucose levels with or without diabetes. The primary end point was in-hospital and 2-year all-cause mortality.

RESULTS

In patients without a history of diabetes, 649 patients had glucose level ≤ 7.8 mmol/L(group A), 204 patients with 7.8-11.0 mmol/L (group B) and 142 patients > 11.0 mmol/L (group C). Of those with diabetes, 124 patients were in group A, 142 in group B and 126 in group C. By Cox regression analysis, admission hyperglycemia (group B, HR 1.48, P = 0.05; group C, HR 2.78, P < 0.001) was an independent predictor of long-term all-cause mortality in patients without diabetes. In patients with diabetes, blood glucose > 11.0 mmol/L (group C, HR 2.68, P < 0.001) also independently predicted long-term all-cause mortality.

CONCLUSIONS

Compared with the diabetic patients, elevation of blood glucose at admission is common in ACS patients without definite history of diabetes. Admission hyperglycemia should be considered as a risk factor strongly correlated with in-hospital and 2-year all-cause mortality in patients with ACS.