Abstract
We assessed the possibility that the glycated albumin (GA) level measured eight years previously can be used to predict the present glycohemoglobin (HbA1c) level. Two hundred and thirty patients referred to our hospital for a general health checkup both this year and eight years earlier were selected. The average age was 56.6 years. A multiple regression analysis was performed using the present HbA1c level as the dependent factor. Independent factors used for the analysis included the levels of GA, fasting blood glucose and 1- and 2-hour blood glucose after 75-g glucose oral loading, as well as body mass index and age, all of which were measured eight years previously. The coefficient of determination for the analysis was 0.575, thus implying that the equation did not fit perfectly. The most stable levels of HbA1c predicted from the level of GA measured eight years earlier were 5.8 %, 5.9 % and 6.0 %; the GA levels predicting these HbA1c levels were 13.9 %, 15.6 % and 16.4 %, respectively. The proportion of patients whose present HbA1c level was over 5.8 % was 20.3 % greater than that observed for patients whose GA level measured eight years earlier was over 13.9 %. A GA level of 15.6 % and 16.4 % predicted an HbA1c level that was 10.6 % and 11.4 % less, respectively. A GA level of 15.6 % or 16.4 % may be therefore used to identify patients who may develop diabetes eight years later.
