Incorporating first trimester analytes to predict delivery of a large for gestational infant in women with impaired glucose tolerance

Aim: To investigate first trimester maternal characteristics of women with impaired glucose tolerance that deliver large for gestational age (LGA) infants. Methods: Nested analysis from a prospective study of singleton pregnancies enrolled at first trimester. We studied women with an abnormal 1 h glucose challenge test that had normal follow-up oral glucose tolerance testing. Maternal characteristics, ultrasound parameters and serum analytes were stratified by subsequent delivery of an LGA infant. Parameters identified as significant on univariate analysis were used for a derivation of prediction by logistic regression. Odds ratio and prediction performance was determined using receiver operator curve (ROC) statistics. Results: A total of 33/114 (28.9%) women meeting the criteria delivered LGA infants. Maternal height (cm), and first trimester pregnancy-associated protein-A and free beta-hCG (MoM) predicted delivery of an LGA infant (ROC area under curve 0.73; 95% CI 0.63-0.83). At a cutoff value of 0.172 the prediction rule achieved 91% sensitivity, 44% specificity, 41% positive predictive and 92% negative predictive value. Conclusion: Maternal height and first trimester high free beta-hCG and PAPP-A levels may be used as predictors for delivery of LGA infants in women with impaired glucose tolerance.