Possibility to predict early postpartum glucose abnormality following gestational diabetes mellitus based on the results of routine mid-gestational screening

INTRODUCTION Women with previous gestational diabetes mellitus (GDM) have increased risk of developing glucose abnormality, but current diagnostic criteria are evidence-based for adverse pregnancy outcome. THE AIMS OF OUR STUDY WERE (i) to ascertain a frequency of early conversion of GDM into permanent glucose abnormality, (ii) to determine predictive potential of current GDM diagnostic criteria for prediction of postpartum glucose abnormality and (iii) to find optimal cut-off values of oral glucose tolerance test (oGTT) to stratify GDM population according to postpartum risk. MATERIALS AND METHODS Electronic medical records of an ethnically homogenous cohort of women diagnosed and treated for GDM in a single medical centre during the period 2005-2011 who completed postpartum oGTT up to 1 year after the index delivery were retrospectively analysed (N=305). RESULTS Postpartum glucose abnormality was detected in 16.7% subjects. Mid-trimester oGTT values, respective area under the curve and HbA1c were significantly associated with early postpartum glucose abnormality (P<0.05, Mann-Whitney) and exhibited significant predictive potential for postpartum glucose abnormality risk assessment. Optimal cut-off values for discrimination of at-risk sub-population were identified using ROC analysis and their comparison with WHO and IADPSG criteria exhibited superiority of IADPSG for risk-stratification of GDM population. CONCLUSION Risk-based stratification at the time of GDM diagnosis could improve efficiency of the post-gestational screening for diabetes. IADPSG criteria seem to optimally capture both perinatal and maternal metabolic risks and are therefore medically and economically justified.