Screening for gestational diabetes mellitus in southern Asian women

AIM: To compare fasting plasma glucose (FPG) against 2-h postprandial plasma glucose (PPPG), following a carbohydrate meal, for screening of gestational diabetes mellitus (GDM) in southern Asian women with one or more risk factors. METHODS: A comparative study was conducted at a university obstetric unit in Sri Lanka. Two hundred and seventy one women undergoing oral glucose tolerance test (OGTT) according to the WHO criteria of 1999 had a 2-h PPPG performed within the following week. Sensitivity, specificity, predictive values and correlation coefficients for predicting a diagnosis of GDM and areas under receiver-operator curves (ROC) were calculated for FPG and PPPG. The ability to predict GDM and to reduce the need for OGTT were the main outcome measures. RESULTS: The mean period of gestation was 26.43 weeks (SD = 5.4) Seventy-five (27.7%) women were diagnosed with GDM. The optimal threshold for FPG was 4.4 mmol/L and for PPPG 4.7 mmol/L. At these, sensitivity was 92.0% and 90.7%, specificity 48.7% and 25.4% and the areas under the ROC 0.82 and 0.73 for FPG and PPPG, respectively. Nine (12%) women could be diagnosed as having GDM on the basis of the FPG being above the threshold. CONCLUSIONS: FPG is superior to 2-h PPPG for screening high-risk women for GDM. Nine women were diagnosed as having GDM on the basis of having an FPG above 7 mmol/L. FPG could reduce the number of OGTT needed by 40.9%, compared to 20.6% by PPPG. FPG is a less cumbersome and cost-effective screening test.