The changing relationship between HbA1c and FPG according to different FPG ranges.

Abstract

Since the American Diabetes Association included hemoglobin A1c (HbA1c) in the diagnostic criteria for diabetes in 2010, the clinical use of HbA1c has remained controversial. We explored the use of HbA1c for diagnosing diabetes and intermediate hyperglycemia in comparison with fasting plasma glucose (FPG). We screened 3710 adult subjects (mean age = 55.24 years) comprising 1704 males and 2006 females. We drew an receiver operating characteristic (ROC) curve to evaluate the ability of HbA1c to diagnose diabetes and intermediate hyperglycemia according to FPG. We used Kappa coefficient and Pearson's correlation coefficient to evaluate the relationship between HbA1c and FPG in different FPG ranges. The areas under ROC curve to diagnose diabetes and intermediate hyperglycemia were 0.859 (95 % CI 0.827-0.892) and 0.633 (95 % CI 0.615-0.651). The kappa coefficients between FPG and HbA1c for diagnosis of diabetes and intermediate hyperglycemia were 0.601 (P < 0.001) and 0.104 (P < 0.001). The Pearson's correlation coefficient of FPG and HbA1c was 0.640 (P < 0.001), but when we classified FPG as normal, intermediate hyperglycemia and diabetes, the coefficients became 0.07 (P = 0.002), 0.185 (P < 0.001) and 0.760 (P < 0.001), respectively. The relationship between HbA1c and FPG changed according to the different FPG ranges. When FPG was higher, the relationship was stronger. HbA1c and FPG were highly consistent in diagnosing diabetes, but they were not in predicting intermediate hyperglycemia.