Diagnostic value of hemoglobin A1c for type 2 diabetes mellitus in a population at risk

Objective: Because the American Diabetes Association has recently included HbA1c as the primary diagnostic test for the detection of diabetes mellitus (HbA1c ≥6.5%) we investigated its use as screening parameter for diabetes in a cohort at increased risk for the disease.

Research design and methods: During the last 10 years 2 036 Caucasians at risk to develop type 2 diabetes but not having this diagnosis yet, consecutively underwent a 75 g oral glucose tolerance test (OGTT). HbA1c was determined with the HPLC method (Tosoh A1c 2.2), external and internal quality controls were well within the allowed ranges.

Results: The oral glucose tolerance test classified 1 523 individuals as normal glucose tolerant (NGT), 387 as impaired glucose tolerant (IGT) or having impaired fasting glycemia (IFG) and 126 as diabetic. The 6.5% cut-off value of HbA1c classified 47% of the diabetic individuals correctly. Of the remaining 53% diabetic individuals (HbA1c <6.5%) 35% had increased fasting glucose levels, while 65% were only diagnosed by their increased 2 h glucose values.

Conclusion: A cut-off value of 6.5% HbA1c classifies diabetic subjects with a specificity of 98.7%. However, the sensitivity of 46.8% is low, indicating that more than half of diabetic subjects are missed when using this test. The present data shows that the use of HbA1c as a the primary diagnostic test will reduce diabetes prevalence. Furthermore, it suggests, that HbA1c and OGTT measurements cannot simply be exchanged, but most probably detect and define different categories of diabetes, i. e., categories with different risk of cardiovascular disease.