Advertisement

399: 1,5-Anhydroglucitol: A new diagnostic test for gestational diabetes?

      Objective

      Current screening for gestational diabetes (GDM) relies on glucose tolerance testing, which is burdensome, expensive, and unreliable. 1,5-Anhydroglucitol (AG) is an unmetabolized monosaccharide excreted in the urine during hyperglycemia. Its steady state concentration is unaffected by fasting or pregnancy; results <10 μg/mL are abnormal outside of pregnancy. We tested the hypothesis that AG can be used as a screening test for GDM in obese women.

      Study Design

      Prospective observational study as a sub-study of an RCT (n=954) that enrolled obese women (BMI≥30 kg/m2) with non-anomalous, singleton gestations <20wk. Women were included if serum samples were obtained at 14-20 wk and/or 24-28 wk. GDM screening was performed using a 50-g, 1-hr glucose challenge test followed by a 100-g, 3-hr glucose tolerance test if ≥135 mg/dL. GDM was diagnosed using Carpenter Coustan criteria at 24-28 wk (gold standard). Serum AG levels were measured with a commercially available assay. AG levels of women with and without GDM were compared and an ROC curve was created to assess the association of AG and GDM. A cutoff for AG was selected using the Liu method, and the test characteristics of AG at this cutoff determined. Association of AG with a perinatal composite outcome of macrosomia (>4000g), primary cesarean, hypertensive disease of pregnancy (PIH), shoulder dystocia, neonatal hyperbilirubinemia, and neonatal hypoglycemia was also examined.

      Results

      Of 954 enrolled, 519 (54.4%) had AG at 14-20 wks and 517 (54.2%) had AG at 24-28 wks. The mean gestational age at blood draw was 17.5 wk and 26.3 wk, respectively. The mean AG value was 13.3 ±5.7 μg/mL at 14-20 wk and 11.1 ± 5.1 μg/mL at 24-28 wk. AG value was significantly different by GDM status at both time points (Table). On ROC analysis, the 14-20 wk AG had an area under the curve (AUC) of 0.61 and the 24-28 week AG had an AUC of 0.67, demonstrating moderate association with a GDM diagnosis. Using a cutoff of 10 μg/mL at 14-20 wk, the sensitivity of an AG≥10 for diagnosing GDM was 45% with a specificity of 70%. Using a cutoff of 11.8 μg/mL at 24 -28 wk the sensitivity was 76% at a specificity of 43%. The primary composite outcome was not associated with AG at either time point (58.3% vs 62.3%, p=0.41 at 14-20 wk; 55.8% vs 63.4%, p=0.08 at 24-28wk).

      Conclusion

      AG at 14-20 wk or 24-28 wk is not sufficiently predictive of a diagnosis of GDM. The sensitivity for GDM is inadequate early in pregnancy and later the specificity is poor.
      Figure thumbnail fx1
      Figure thumbnail fx2