Concordance in incidence of diabetes among persons with prediabetes detected using either oral glucose tolerance test or glycated haemoglobin

•HbA1c or OGTT was used in identifying prediabetes and assessing incident T2DM.•Concordance in the conversion of prediabetes to T2DM with glucose and HbA1c studied.•Incidence of T2DM in both cohorts were similar with either of the two criteria.•Both the criteria had similar validity and utility in identifying diabetes risk.•Multivariate analysis showed baseline waist circumference to impact study outcome. To study the concordance in the incidence of type 2 diabetes (T2DM) between cohorts with prediabetes, selected either by oral glucose tolerance test (OGTT) or glycosylated haemoglobin (HbA1c) at two years in a real world situation. Two cohorts with impaired glucose tolerance (IGT) were selected from the non-interventional arm of the Indian diabetes prevention programmes; a group selected by using OGTT (Cohort 1, n = 498), another selected based on the HbA1c criterion (Cohort 2, n = 504). Clinical and biochemical data collected for 24 months at 6 monthly intervals were used in assessing the cumulative incidence of T2DM using the respective diagnostic criteria. Intra and inter group comparisons were analysed using appropriate statistical tests. A multiple logistic regression analysis was used to identify the variables significantly associated with the incidence of diabetes. Incidence of diabetes in both cohorts were similar at 12 and 24 months with either of the two criteria (25.3% with glucose and 27.5% with HbA1c, p = 0.41 at 24 months). The multivariate analysis confirmed the results. Only baseline waist circumference was positively associated with the incidence. Both OGTT and HbA1c have similar utility and validity in identifying persons with IGT. Persons identified with either of the criterion had similar incidence of T2DM among Asian Indians.