1159-P: One-Hour Glucose during Oral Glucose Tolerance Test Predicts Hyperglycemia Relapse in Black Obese Patients with Hyperglycemic Crises

Abstract

Approximately 50% of Black individuals with unprovoked diabetic ketoacidosis (DKA) or severe hyperglycemia (SH, blood glucose [BG]>400 mg/dl without ketosis) at new-onset of diabetes achieve near-normoglycemia remission (FBG < 130 mg/dl, HbA1c < 7%) with intensive insulin treatment. After remission, many patients experience hyperglycemia relapse. Previous studies showed that one-hour post glucose load (1-h PG) correlated with insulin sensitivity (Si) and secretion. We tested whether 1-h PG at remission predicts HR in obese Black individuals presenting with DKA or SH. Seventy-three Black individuals presenting with DKA (n=40) or SH (n=33) underwent a 2-h 75-gm OGTT after achieving remission and followed for a median of 58 weeks until hyperglycemia relapse (FBG>130 mg/dl, random BG>180 mg/dl, or HbA1c> 7%). Si and was calculated from the oral minimal model. Disposition index (DI) was calculated as the product of Si and Area under the curve for insulin. Mean age was 47±10 years, with 36% (n=26) women, mean BMI 36.1±9.5 kg/m2 and 1-h PG 192±48 mg/dl. One-h PG was moderately correlated with Si (r=-0.26; P=0.0294) and DI (r=-0.28; P=0.023). Highest Youden index for 1-h PG was 0.38, corresponding to an optimal cut-point of 199mg/dl with 64% sensitivity and 71% specificity. Cox proportional hazards model adjusted for age, sex and BMI showed that 1-h PG level>199 mg/dl was independently associated with increased hyperglycemia relapse (HR: 2.61 [95% CI:1.13-6.01]). In a multivariate model with fasting BG, adding 1-h PG level improved prediction of hyperglycemia relapse than 2-h PG level, with significant improvements in C index (Δ: +0.07; P=0.04), net reclassification index, and integrated discrimination index. A 1-h PG at the time of remission is a better predictor of hyperglycemia relapse than fasting or 2-h PG among obese Black individuals presenting with DKA or SH. 1-h PG can be used to identify individuals at high risk of developing hyperglycemia relapse. Disclosure R. Jagannathan: None. D. Smiley-byrd: Consultant; Self; LifeScan, Other Relationship; Self; Novo Nordisk, Speaker’s Bureau; Self; Merck & Co., Inc. D. Stefanovski: None. G. E. Umpierrez: Research Support; Self; AstraZeneca, Dexcom, Inc., Novo Nordisk. P. Vellanki: Consultant; Self; Boehringer Ingelheim Pharmaceuticals, Inc. Funding National Institute of Diabetes and Digestive and Kidney Diseases (K23DK11324-01A1, UL1TR002378, K08DK0830361)