955-P: Postpartum Carbohydrate Intake Is Not Associated with Oral Glucose Tolerance Test Results

Abstract

Background: In pregnant and postpartum (PP) women, an oral glucose tolerance test (OGTT) is the diagnostic test of choice for glucose intolerance. Consistent with classic studies outside of pregnancy, we recently found that lower carbohydrate intake on the day before an OGTT in pregnancy was associated with higher post-load OGTT glucose. Here, we test for a relationship between carbohydrate intake and OGTT glucose in PP women. Methods: PP women (N=84, median [IQR] = 9 [8-14] weeks PP) completed a 75 g OGTT and a validated self-administered 24-hr dietary recall. We tested for an association between carbohydrate intake in the day before the OGTT and 1-hr OGTT glucose using linear regression, with adjustment for age, race, BMI, time postpartum, and caloric intake. We also tested for associations with 30 minute and 2-hr OGTT glucose and glucose area under the curve (AUC). A subset of women (N=72) had completed the same protocol during pregnancy (median=26 weeks gestation). Results: PP women with lower carbohydrate intake (<median=230 g/day) had similar age, BMI, race, education, GDM status, time postpartum, and time fasting before OGTT to women with higher intake. We found a trend toward higher fasting glucose in those with lower carbohydrate intake (median=87 vs. 81 mg/dl, P=0.09), but no association between carbohydrate intake and post-load OGTT glucose (1-hr: β=1.2 mg/dl, P=0.74, 30 min: β=-2.0 mg/dl, P=0.39, 2-hr: β=1.0 mg/dl, P=0.74, AUC: β=66, P=0.83). Post-load results did not change with fasting glucose adjustment. In women with data available in pregnancy, we found similar results to those previously reported: at ~26 weeks gestation, lower carbohydrate intake was associated with higher OGTT glucose. Conclusion: While carbohydrate intake in the day prior to an OGTT may alter gestational diabetes diagnosis, we found no relationship between carbohydrate intake and OGTT glucose among PP women. Metabolic factors that differentiate the pregnant and postpartum state may account for our findings. Disclosure E. A. Rosenberg: None. E. W. Seely: None. K. James: None. M. J. Callahan: None. S. Nelson: None. C. E. Powe: None. Funding National Institutes of Health (K23DK113218); Robert Wood Johnson Foundation (T32DK007028); Massachusetts General Hospital