Aim: To determine the HbA1c cut-offs for diagnosis of prediabetes based on isolated IFG, isolated IGT and both IFG and IGT.
Methods: Cross-sectional randomized population survey involving 2368 adults, aged ≥20 years. HbA1c was measured on a Bio-Rad 10 system in 1972 subjects.
Results: Among 1972 studied subjects, 329 were detected to have prediabetes based on isolated IFG in 125 (6.3%), isolated IGT in 141 (7.1%) and both in 63 (3.2%) of subjects. The optimal HbA1c cut-off for diagnosis of prediabetes based on IFG criteria (6.1 to <7 mmol/l) was 5.5% which had a sensitivity of 67%, specificity of 52% with accuracy of 45%, and for 2 h PG (7.8 to <11.1 mmol/l), it was 5.7% with sensitivity, specificity and accuracy of 67, 67 and 59%, respectively. Subjects who had both IFG and IGT had optimal HbA1c cut-off of 5.7% with sensitivity, specificity and accuracy of 69, 60 and 57%, respectively. One hundred and thirty-four subjects (6.8%) were newly detected to have diabetes and HbA1c cut-off of 6.1% had an optimal sensitivity and specificity of 81% each with accuracy of 77% for diagnosis of diabetes.
Conclusions: HbA1c between 5.5 and 6% corresponds to the presence of prediabetes in Asian Indians.