Author Affiliations: Endocrinology and Diabetes Section (Dr Chalew) (email@example.com) and Research Institute for Children (Dr Hempe), Children's Hospital, New Orleans, Louisiana; and Bioinformatics and Statistics Unit, Children's National Medical Center, Washington, DC (Dr McCarter).
To the Editor: Hemoglobin A1c (HbA1c) measures contain information regarding mean blood glucose (MBG) as well as MBG-independent factors.1 In their Commentary, Drs Rubinow and Hirsch pointed out that the presence of considerable between-patient variation in HbA1c values not due to MBG poses an important challenge for clinical interpretation of HbA1c tests.2 The hemoglobin glycation index (HGI) was created specifically to quantify and assess this MBG-independent component of HbA1c measurement.1 HGI is the difference between a patient's observed HbA1c value and the predicted HbA1c value (calculated by inserting the patient's observed MBG into the population regression equation of HbA1c vs MBG1,3). HGI is highly correlated with HbA1c values but statistically independent of MBG values.1,3 In an analysis of data from the Diabetes Control and Complications Trial, we used HGI to demonstrate that HbA1c level was a predictor of retinopathy and nephropathy that was statistically independent of MBG level.3 The authors referenced a critique by Lachin et al4 that questioned the value of HGI and its strong association with complications because it is not independent of HbA1c level. As HGI is a component of HbA1c measurement, it cannot be independent of HbA1c level. But it is independent of MBG values and thus can be used to assess the importance of MBG-independent variation in HbA1c values as a predictor of complications.3
Chalew S, Hempe J, McCarter R. Metrics for Monitoring Glycemic Control. JAMA. 2011;305(24):2522-2523. doi:10.1001/jama.2011.845