Résumé
Evaluation of diabetic control was performed by using fasting plasma glucose, hemoglobin A1 and fructosamine in 139 patients with diabetes mellitus, and 36 normal controls. A linear correlation of fasting plasma glucose with fructosamine and hemoglobin A1 was found. Using fasting plasma glucose alone was found to be inadequate to define good control. HbA1 and fructosamine had an acceptable sensitivity and specificity in assessment of diabetic control, although fructosamine was slightly less sensitive than HbA1. In patients with thalassemia, hemoglobin A1 levels were elevated in 18 of 19 patients. Fructosamine levels also gave misleading results since 6 to 19 patients had an elevated level and one patient had a decreased level. Patients with hypoproteinemia had a decreased fructosamine and hemoglobin A1 level compared to normal control. HbA1 and fructosamine should be cautiously interpreted in patients with thalassemia and hypoproteinemic states. Using these methods in combination with other measure such as home monitoring of blood glucose would be more precise particularly in diabetic patients with hypoproteinemia, abnormal hemoglobin and other hemolytic disorders.