Search

1. The aim of the present study was to assess the validity of glycated albumin (GA) and fasting plasma glucose (FPG) as a screening tool for the early detection of diabetes in Chinese subjects. 2. A total of 1971 outpatient subjects underwent a 75 g oral glucose tolerance test (OGTT) and GA measurement. The receiver operating characteristic curve (ROC) was plotted to examine the sensitivity, specificity, and positive and negative predictive values of GA and FPG in detecting undiagnosed diabetes at the different cut-off levels. 3. The prevalence of impaired glucose regulation and diabetes was 27.40% and 38.30%. For these diabetic individuals, 4.64% had isolated fasting hyperglycemia, 50.86% had isolated postprandial hyperglycemia and 44.50% had both. Using ROC analysis, a GA of 17.1% gave an optimal sensitivity of 76.82% (95% confidence interval: 73.64-79.79%) and specificity of 76.89% (74.42-79.23%) for the diagnosis of diabetes. Likewise, a FPG of 6.1 mmol/L gave an optimal sensitivity of 80.93% (77.94-83.67%) and specificity of 85.94% (83.86-87.84%). If subjects met both criteria, they were regarded as having diabetes; the positive predictive value of the combined criteria, FPG ≥ 6.1 mmol/L and GA ≥ 17.1%, was relatively high (84.79% (81.62-87.60%)), and this would have avoided 76% of the OGTT in our survey. 4. In conclusion, a GA value of 17.1%, an optimal cut-off in Chinese subjects, identified a high proportion of potential diabetic individuals. Simultaneous measurement of FPG and GA would enhance the sensitivity of diabetes screening in our population and avoid 76% of OGTT. [ABSTRACT FROM AUTHOR]