Hemoglobin A1c less than 6.5% on admission increases risk for in-hospital and 1-year mortality in patients with diabetes and chest pain

Abstract

BackgroundAs the association between low hemoglobin A1c (HbA1c) and adverse clinical outcomes in patients with chest pain and myocardial infarction has not been extensively examined, a prospective cohort study was conducted to examine whether low HbA1c magnifies mortality risk. Patients and methodsA total of 717 consecutive patients (284 with diabetes and 433 without known diabetes) admitted to a public hospital over a 6-month period with chest pain and troponin measurement were included. HbA1c was obtained upon admission. Mortality was confirmed by National Death Index search. Cox proportional hazards models were constructed for the sample stratified by diabetes status. ResultsThe unadjusted incidence of mortality during hospitalization in patients with diabetes and HbA1c less than 6.5% (n=55) was 3.84 compared with those with HbA1c of 6.5% or more (n=229). In Cox proportional hazard models, patients with diabetes and HbA1c less than 6.5% had higher risk for in-hospital (hazard ratio 3.51, 95% confidence interval 1.25–9.83, P=0.02) and 1-year mortality (hazard ratio 3.50, 95% confidence interval 1.84–6.64, P<0.001) compared with those with HbA1c of 6.5% or more. No increase in risk was seen for those without diabetes and HbA1c less than 6.5%. Hypoglycemia (serum glucose<3.89 mmol/l) was more frequent in patients with diabetes and HbA1c less than 6.5% (44%) than those with HbA1c of 6.5% or more (22%) (P=0.001). Patients with hypoglycemia and diabetes had significantly higher rate of in-hospital mortality than those without [13.33% (10/75) vs. 2.87% (6/209), P<0.001]. ConclusionPatients with diabetes admitted for chest pain with HbA1c less than 6.5% may be at increased risk for in-hospital and 1-year mortality. The mechanism for death may be mediated in part by hypoglycemia.