943-P: Glycated Albumin as a Predictor of Adverse Neonatal Events in Pregnant Women with Diabetes Free

Control of plasma glucose is effective to reduce adverse neonatal events (ANE) for pregnant women with diabetes, including gestational diabetes (GDM); however, compliance can be poor. We examined the utility of glycated serum protein (GSP) and glycated albumin (%GA) to predict risk of ANE. Published studies have shown greater risk if %GA of ≥ 12 -15.8%, but population and trimester-specific cut-offs are needed. Additionally, BMI has been reported to affect %GA. We prospectively recruited pregnant women with type 2 diabetes or GDM and collected serum during the second and/or third trimesters. No exclusion criteria were used, but an attempt was made to recruit similar numbers of overweight and Class I, II, or III obese BMI. Enrolled subjects included 52% African-American and 41% Caucasian with median BMI 34.0 (25.0 - 70.4). Thirty bio-banked samples, matched for ethnicity, BMI and gestational age, from women with no evidence of diabetes were used as controls. GSP and %GA as well as various lipid markers were measured using methods from Roche or Diazyme (GSP, %GA) on automated analyzers. Our goals were to determine preliminary cut-off values for GSP and %GA to predict ANE and identify whether additional lipid markers may improve prediction. Of those delivered to date (n=55), two controls and four subjects experienced one or more ANE (macrosomia, hyperbilirubinemia, hypoglycemia or RSD). Using receiver operator curve analysis, trimester two values of %GA or GSP were much better at predicting ANE (AUC: 0.896, 0.910, respectively) than trimester three values (AUC: 0.662, 0.713, respectively). Our data suggests that a cut-off ≥ 11.8 %GA can effectively predict those at greatest risk in this population. Statistical analysis showed no differences in %GA or GSP between ethnicities or BMI subgroups. Additional data is needed to determine whether or not incorporation of lipid markers will improve risk prediction. However, only beta-hydroxybutyrate shows significant differences between subjects and controls.