OBJECTIVETo compare the impact of the first 24 hours mean blood glucose (MBG) level and admission glucose (AG) during hospitalization on the short term mortality and combined end point events in patients with ST-segment elevation acute myocardial infarction (STEMI).

METHODSA total of 7446 Chinese STEMI patients hospitalized within 12 hours of symptom onset were included. Plasma glucose was measured at admission, 6 and 24 hours after admission, respectively. The MBG level through the first 24 hours for each patient was calculated. Patients were stratified into six groups according to their MBG levels: < 4.5, 4.5 - 5.5, 5.6 - 7.0, 7.1 - 8.5, 8.6 - 11.0 and > 11.0 mmol/L. The incidence of all-cause mortality and combined end point of death, re-infarction, cardiogenic shock, recurrence ischemia, and stroke at 7 days and 30 days post hospitalization were analyzed. Nested models were compared to determine whether logistic regression models that included MBG provided a significantly better fit than logistic regression models included AG.

RESULTSCompared with the MBG of 4.5 - 5.5 mmol/L group, 7-day and 30-day mortality and combined end point events increased in proportion to plasma MBG level increase. Multivariate logistic regression analysis showed that elevated MBG (equal or greater than 7.1 - 8.5 mmol/L) level is an independent predictor of 7-day and 30-day mortality and combined end point events. Nested models analysis showed that the prognostic impact of MBG is superior to AG (P < 0.001) on predicting 7-day and 30-day mortality and combined end point events in this patient cohort.

CONCLUSIONElevated MBG (≥ 7.1 mmol/L) level is an independent predictor of 7-day and 30-day mortality and combined end point events. MBG is superior to AG on predicting short-term prognosis in this patient cohort.