Objective: To determine the association of, and predictive ability of, pregnancy-associated plasma protein A (PAPP-A), free beta-human chorionic gonadotrophin (beta-hCG), and nuchal translucency (NT) with preterm birth (PTB).
Methods: A 5-year retrospective cohort study of women who underwent first-trimester combined screening was performed. Maternal medical, antepartum, and pregnancy outcome data were obtained. PAPP-A and beta-hCG were converted to multiples of the median (MoM), and primary exposure was defined as < or =10th percentile MoM for PAPP-A. Secondary exposures were defined as > or = 90th percentile MoM for beta-hCG and NT values of > or = 20 and 25 mm. The primary outcome was PTB before 35 weeks and the secondary outcome was PTB before 32 weeks. Univariate, bivariate, multivariate, and receiver-operator analyses were used.
Results: Of the 2231 patients meeting inclusion criteria with complete outcome data available, 222 had a PAPP-A level < or =10th percentile MoM. Abnormally low PAPP-A was associated with an increased risk for PTB < 35 weeks [adjusted odds ratio (aOR) 2.0, 1.0-3.8] and < 32 weeks (aOR 2.7, 1.1-6.4), even after adjusting for prior PTB, tobacco exposure, chronic hypertension, and body mass index. PAPP-A < or =10th percentile was not sufficiently predictive of PTB < 35 weeks (area under curve = 0.63, 95% CI 0.53-0.72). Neither abnormally high beta-hCG nor increased NT was associated with an increased risk for PTB.
Conclusions: PAPP-A < or =10th percentile is associated with an increased risk for PTB, but is not sufficiently predictive to be used clinically.
Copyright (c) 2010 John Wiley & Sons, Ltd.