Maternal urine beta-core hCG fragment level and small for gestational age neonates

OBJECTIVE: To determine whether second-trimester urine beta-core fragments of hCG predict small for gestational age (SGA) neonates. METHODS: Spot urine beta-core levels were measured in 733 nonhypertensive women with singleton pregnancies who presented for amniocentesis and had karyotypically normal fetuses. The beta-core level was standardized to urine creatinine and expressed as multiples of the median. The area under a receiver operating characteristics curve was used to determine the screening efficiency of the urine analyte for prediction of small for gestational age (SGA) births. In a subgroup of cases, serum markers (alpha-fetoprotein [AFP], hCG, and unconjugated estriol) were compared using stepwise regression analysis to urine beta-core fragment for SGA prediction. RESULTS: There were 23 (3.0%) SGA neonates. The mean +/- standard deviation (SD) gestation at urine collection was 16.4 +/- 1.3 weeks and collection to delivery interval was 23.0 +/- 2.2 weeks. Mean beta-core (+/- SD) fragment levels were significantly higher in those who later had SGA infants compared with appropriately grown infants (2982.8 ng/mg creatinine versus 1447.4 ng/mg creatinine, P <.001). Stepwise logistic regression found that urine beta-core fragment and serum AFP were the only significant predictors of SGA, with statistically significant chi(2) values (P <.001 and P =.038, respectively). The urine analyte was significantly superior. Second-trimester urine beta-core fragment had a 78.3% sensitivity and 70% specificity for SGA prediction. Exclusion of preeclamptic cases resulted in a sensitivity of 84.2% and a specificity of 71.2%. CONCLUSION: Second-trimester elevated maternal urine beta-core fragment of hCG predicted SGA infants, and was superior to other serum analytes in that prediction.