Aims The aim of this study was to estimate the incidence of type 2 diabetes using newly proposed haemoglobin A_1C (HbA_1c) and current oral glucose tolerance test (OGTT) definition in an Iranian non-diabetic population.

Methods A total of 923 non-diabetic first-degree relatives of patients with type 2 diabetes 20–70 years old in 2003–2005 were followed through 2009 for the occurrence of type 2 diabetes. At baseline and through follow-ups, participants underwent a standard 75 g 2-h OGTT and HbA_1c measurements. Prediction of progression to type 2 diabetes by OGTT-defined or HbA_1c-defined was assessed with area under the receiver-operating characteristic curves based upon measurement of fasting plasma glucose, 2-h post-load glucose values and HbA_1c.

Results The prevalence of type 2 diabetes was 9.2% (95% CI 8.2 to 10.2) by OGTT-defined diabetes and 7.9% (95% CI 6.9 to 9.0) by HbA_1c ≥6.5. The incidence of type 2 diabetes was 2.0% (95% CI 1.6 to 2.4) (1.8% men and 2.1% women) per year by the current OGTT definition, whereas the incidence rates were 1.7% (95% CI 1.3 to 2.0) (1.6% men and 1.7% women) per year by HbA_1c ≥6.5%. Of those diagnosed with type 2 diabetes by OGTT, 69.6% had HbA_1c<6.5% and therefore would not have been classified as having type 2 diabetes.

Conclusions The incidence and prevalence of diabetes using newly proposed HbA_1c threshold in this first-degree relatives of patients with type 2 diabetes was slightly lower than using current OGTT definition.