Glycosylated albumin and transferrin: short-term markers of blood glucose control

Serum was collected from children with type 1 diabetes mellitus before and 10 days after attending a camp session in which blood glucose concentrations were carefully controlled. Glycosylated and nonglycosylated proteins were separated, and levels of albumin and transferrin were determined on each of these fractions. Glycosylated hemoglobin was also determined and ranged from 4.6% to 14.6% (mean +/- SEM 8.1% +/- 0.2%). Mean initial glycosylated albumin in 73 children was 16.4% +/- 0.6%, which was elevated compared with the mean of levels in 20 nondiabetic controls (8.7% +/- 0.3%) and correlated well with levels of glycosylated hemoglobin (r = 0.71). After 10 days mean glycosylated albumin fell to 14.6% +/- 0.5% (P less than 0.00001), near the predicted final value of 13.4% if control had been ideal. Initial levels of glycosylated transferrin in 44 of these children ranged from 4.5% to 22.3% (11.4 +/- 0.6%) and was significantly higher than the mean of 3.8% +/- 0.3% in 20 nondiabetic controls. Mean final glycosylated transferrin fell to 8.2% +/- 0.3% (P less than 0.00001), near the predicted final mean of 7.0% +/- 0.2%. The mean of each subject's blood glucose determinations performed throughout the study period correlated with final levels of both glycosylated albumin (r = 0.55, P less than 0.001) and glycosylated transferrin (r = 0.54, P less than 0.001). Both glycosylated albumin and glycosylated transferrin appear to be reliable markers of short-term glycemic control; glycosylated transferrin (half-life 8 days) was more sensitive than glycosylated albumin over this 10-day period.