202-OR: Prediction of Late Gestational Diabetes Mellitus in Those without Early GDM

Aim: To compare glycemic test characteristics for predicting GDM at 24-28 weeks’ gestation (LGDM) among those with a normal OGTT 20 weeks’ gestation. Methods: Pregnant women with GDM risk factors were enrolled into a multicenter early GDM (EGDM) treatment trial. Women with EGDM are excluded from this analysis. A 2-h 75g OGTT with HbA1c occurred on entry. The OGTT was repeated at 24-28 weeks’ gestation. GDM was defined by WHO 2013 criteria. Receiver Operator Curve (ROC) assessment compared early fasting (FBG), 1-h (1HBG), 2-h (2HBG) glucose and HbA1c prediction of LGDM. ROCs95%CI were calculated within European, Asian/Pasifika and other ethnic groups. The proportions with LGDM ≥90th and 90th centile, and their odds ratio (OR 95%CI) using Mantel-Haenszel testing were calculated for each early glycemic test and a composite of any OGTT measure≥90th centile (F12BG). Results: Those with LGDM (n=467) vs no GDM (n=2218) were older (31.4±4.9 vs 30.7±5.1 yrs p=0.009), more likely to be overweight/obese (67.0% vs 57.3% p0.001), non-European (63.6% vs 57.0% p=0.030) and to have a diabetes family history (52.1% vs 40.7% p0.001). The ROCs for early FBG, 1HBG, 2HBG, HbA1c were 0.630.60-0.66, 0.730.70-0.76, 0.660.63-0.69, 0.630.60-0.66 respectively. ROC patterns were similar across ethnic groups. The ≥90th centile cut-offs (FBG, 1HBG, 2HBG, HbA1c) were ≥4.9 mmol/l, ≥8.8 mmol/l, ≥7.3 mmol/l; ≥5.3% including 11.1%, 10.4%, 10.1%, 15.0% of the cohort. The F12BG composite included 34.0% of the cohort. Comparing ≥90th centile vs 90th centile, LGDM was present in 32.6% vs 15.4%, 47.3% vs 13.9%, 36.4% vs 15.3%, 30.1% vs 14.6%, 35.3% vs 11.4% across the 5 measures with OR of 2.62.0-3.4, 5.64.3-7.2, 3.22.4-4.2, 2.52.0-3.2, 4.23.4-5.2 respectively. Sensitivities were 21.4%, 28.3%, 21.2%, 27.7%, 50.7% respectively. Conclusions: After excluding those with EGDM, the 1HBG was the best single test, but no single or combined early glycemic measure was sufficiently discriminatory to avoid the second OGTT. Disclosure D. Simmons: Speaker's Bureau; Abbott. Consultant; Sanofi. Speaker's Bureau; Ascensia Diabetes Care. Other Relationship; Ascensia Diabetes Care, Boehringer-Ingelheim. Research Support; Dexcom, Inc. Speaker's Bureau; Novo Nordisk. Research Support; Wests Club, NSW, Australia, Roche Diabetes Care. W. Hague: None. A. Sweeting: None. N. Cheung: None. C.J. Nolan: None. H. Teede: None. M.J. Peek: None. A. Kautzky-Willer: None. E. Hibbert: None. H.E. Fadl: None. V.W.M. Wong: None. J. Harreiter: None. J. Immanuel: None. E.J. Gianatti: None. V. Mohan: None. Funding National Health and Medical Research Council (grants 1104231 and 2009326), the Region Örebro Research Committee (grants Dnr OLL-970566 and OLL-942177), Medical Scientific Fund of the Mayor of Vienna (project numbers 15205 and 23026), the South Western Sydney Local Health District Academic Unit (grant 2016), and a Western Sydney University Ainsworth Trust Grant (2019).