EFFECTIVENESS OF GLYCOSYLATED HEMOGLOBIN, FASTING PLASMA GLUCOSE, AND A SINGLE POST LOAD PLASMA GLUCOSE LEVEL IN POPULATION SCREENING FOR GLUCOSE INTOLERANCE

Five shortcut methods of population screening for glucose intolerance (impaired glucose tolerance and non-insulin-dependent diabetes mellitus) were assessed for effectiveness: 1) glycosylated hemoglobin concentration (HbA1), 2) fasting plasma glucose level, 3) combinations of fasting plasma glucose and HbA1, 4) plasma glucose one hour post oral glucose load, and 5) plasma glucose two hour post oral glucose load. In a sample of the Israeli Jewish population aged 40-70 years, 2040 participants in the Israel Study of Glucose Intolerance, Obesity and Hypertension, who were not known to be diabetic, underwent an oral glucose tolerance test based on three blood samples (fasting, one hour, and two hour post oral glucose load). In 1058 of the subjects, HbA1 was also measured, and was found to increase significantly (P less than 0.001) with increasing glucose intolerance, but with extensive overlap of ranges, even between normals and newly found diabetics. Fasting plasma glucose was more effective than HbA1 in screening for both impaired glucose tolerance and diabetes by its higher specificity and predictive value of a positive test at comparable sensitivity levels. Combinations of HbA1 and fasting plasma glucose did not improve prediction over fasting plasma glucose alone. As observed in other studies, the screening effectiveness of fasting plasma glucose was also unsatisfactory, either post load glucose level being more effective. Plasma glucose level two hour post load was better for detection of diabetes alone. Plasma glucose level one hour post load was more effective at detecting the total group of glucose intolerance, but did not discriminate well between impaired tolerance and diabetes. A cost-risk-benefit evaluation suggests that a full three-sample oral glucose tolerance test is the best method in screening for both intolerance categories.