355-OR: Is Classification of GDM Phenotypes Dependent on Ethnicity?

Abstract

Clinical experience suggests that diverse clinical phenotypes exist within the broader diagnosis of GDM. Recent published analysis from a single center outlined heterogeneity in GDM with respect to insulin secretion and sensitivity, defining four GDM phenotypes: 1) GDMsecr (<25th centile HOMA-β for non-GDM); 2) GDMsens (<25th centile Matsuda Index for non-GDM); 3) GDMmixed (both GDMsecr and GDMsens); 4) no defect (neither GDMsecr and GDMsens). Classification using these phenotypes is associated with adverse outcomes. It is unknown whether adjustment for ethnicity may provide stronger associations with outcome prediction. Using data from 5 centers in the HAPO study, GDM women were classified into one of four phenotypes in two ways: 1) whole sample cut-offs; and 2) ethnic-specific glycemic thresholds. Descriptive characteristics were compared between the five groups (four GDM phenotypes + non-GDM), and multivariable analysis adjusted for BMI, center and key demographic and clinical characteristics. Of the 6128 women, 2442 (40%) were caucasian, 294 (4.8%) black, 1556 (25.4%) Hispanic, 1616 (26.4%) Asian and 220 (3.6%) other ethnicity; ethnicity was closely related to study center. Seventeen percent were classified as GDM (n=1047), with slightly higher GDM frequency for black (20.1%) and Hispanic women (23.3%). There was 83.9% agreement between whole-sample and ethnic-specific classification into GDM phenotypes. There was a significant relationship (p<0.05) between one or more GDM phenotypes and LGA, primary CS, neonatal hyperinsulinemia and/or fetal adiposity. Despite minor changes in the adjusted odds ratios, there were no statistically significant changes in the relationship between GDM phenotype and outcome when comparing whole sample cut-offs and ethnic-specific cut-offs. Where available, ethnic-specific cut-offs should be used to define GDM phenotype to assist in identifying GDM women at higher risk of adverse outcome, however whole-sample cut-offs can be used with confidence in their absence. Disclosure K.S. Gibbons: None. D. McIntyre: None. R.C. Ma: Advisory Panel; Self; Boehringer Ingelheim International GmbH. Research Support; Self; AstraZeneca, Bayer AG, Pfizer Inc. Stock/Shareholder; Self; GemVCare. Other Relationship; Self; AstraZeneca. W. Tam: None. L.R. Madsen: None. D.A. Sacks: None. P. Catalano: None.