Gaps in the glycation gap hypothesis.

Patients with diabetes mellitus are at risk for developing severe debilitating complications, including retinopathy, nephropathy, neuropathy, myocardial infarction, and stroke. Randomized trials have documented that lowering hemoglobin A1c (Hb A1c) 4 concentrations significantly reduces the onset and rate of progression of microvascular complications (1, 2 ). Moreover, Hb A1c, which is formed by the nonenzymatic attachment of glucose (termed glycation) to hemoglobin and reflects the mean glucose concentration over the preceding 8 to 12 weeks (3, 4 ), correlates with the risk for developing microvascular and cardiovascular complications in diabetes (5, 6 ). Hb A1c measurement is widely used to monitor glycemic control and to adjust therapy. More recently, its use has been advocated for diagnosing diabetes (7 ). Although intraindividual variation in Hb A1c values is minimal in nondiabetic persons (8 ), considerable interindividual variation has been reported (9, 10 ). Evidence that has accumulated over the last few years suggests that race influences Hb A1c. Mexican Americans and blacks have higher mean Hb A1c values than whites (9, 10 ). Similarly, Hb A1c values increase with age (11 ). In addition to these factors, some investigators have proposed that interindividual variation is due to differences in the glycation rate (9, 12 ), a premise that is not accepted by all (13 ). A large retrospective observational analysis that appears in this issue (14 ) provides additional information to consider. Cohen et al. suggested that a “glycosylation gap,” defined as the difference between the Hb A1c concentration and that predicted by the fructosamine concentration, could explain the excess interindividual variation in Hb A1c (15 ). They also suggested that the glycosylation gap predicted the progression of nephropathy, although Hb A1c in their small cohort was not itself associated with such progression, in contrast to the compelling evidence from the Diabetes Control and Complications Trial (DCCT) (1, 5 ). Similarly, McCarter et al. proposed the hemoglobin glycation index (HGI) to explain observations that some individuals have Hb A1c values that are higher or lower than expected from measurements of blood glucose (12 ). The HGI was calculated as the difference between the measured Hb A1c value and that predicted from the mean blood glucose concentration. In analyzing publicly available data from the DCCT, McCarter et al. found that the HGI, which was computed from infrequent glucose measurements, was a statistically significant predictor of retinopathy and nephropathy and inferred that this finding supported the concept that biological variation in glycation contributes to the risk of microvascular complications (12 ). Accurate measurement of mean blood glucose is difficult, and surrogate methods have been used. Fructosamine is the marker most widely used in studies that have supported the glycation gap (GG) concept. In addition to hemoglobin, glucose can attach nonenzymatically to free amino groups of other proteins to form ketoamines. The fructosamine assay measures glycated serum proteins, the most common of which is albumin. The half-life of albumin in the blood is 14 to 20 days, and the fructosamine concentration reflects the mean glucose concentration over 10 to 14 days, a much shorter period than that reflected by the Hb A1c assay. In a study described in this issue of Clinical Chemistry, Rodrı́guez-Segade et al. (14 ) followed 2314 patients with type 2 diabetes for a mean of 6.5 years. Values for the GG, which was calculated as the measured Hb A1c value minus that predicted from fructosamine, were used to divide the patients into tertiles. The authors observed that the risk of progression of nephropathy in the mediumand high-GG groups was significantly greater than that in the low-GG group. In contrast to the study of Cohen et al. (15 ), Rodrı́guezSegade et al. did observe that higher Hb A1c values were significantly associated with the progression of nephropathy. Furthermore, the GG predicted the progression of nephropathy, even after adjustment for Hb A1c. The DCCT (13 ), however, demonstrated statistically that the HGI must be strongly correlated with Hb A1c and not with the mean blood glucose concentration, owing to the fundamental relationship between the correlation of a residual from the regression of y on 1 Department of Pathology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA; 2 current affiliation: Department of Laboratory Medicine, NIH, Bethesda, MD; 3 Diabetes Center, Massachusetts General Hospital and Harvard Medical School, Boston, MA; 4 The Biostatistics Center, George Washington University, Rockville, MD. * Address correspondence to this author at: Department of Laboratory Medicine, 10 Center Drive, Building 10, Room 2C427, Bethesda, MD 20892-1508. Received November 18, 2010; accepted November 19, 2010. Previously published online at DOI: 10.1373/clinchem.2010.158071 4 Nonstandard abbreviations: Hb A1c, hemoglobin A1c; DCCT, Diabetes Control and Complications Trial; HGI, hemoglobin glycation index; GG, glycation gap. Clinical Chemistry 57:2 150–152 (2011) Editorial