First-trimester maternal serum alpha-fetoprotein is not a good predictor for adverse pregnancy outcomes: a retrospective study of 3325 cases.

BACKGROUND: It is well known that second-trimester maternal serum alpha-fetoprotein (MS-AFP) is a predictor for adverse pregnancy outcomes (APOs), such as preterm birth, stillbirth, preeclampsia and small for gestational age (SGA). However, it is unknown whether first-trimester MS-AFP is also predictive of APOs. METHODS: We retrospectively reviewed the data on the first-trimester MS-AFP levels and pregnancy outcomes of 3325 singleton pregnant women. The cutoff value of 2.5 multiple of the median (MoM) was used to evaluate the risks of APOs regarding MS-AFP. The receiver operating characteristic (ROC) curves were used to evaluate the predictive efficiencies of MS-AFP to these disorders. RESULTS: A total of 181 pregnancies resulted in preterm birth, 32 in stillbirth, 81 in preeclampsia, and 362 in SGA. Compared to women with MS-AFP /= 2.5MoM had increased risks (odds ratio, 95% confidence interval) of preterm birth (2.53, 1.65~3.88), preeclampsia (3.05, 1.71~5.43) and SGA (1.90, 1.34~2.69), and had an earlier distribution of gestational weeks at delivery (P = 0.004) and a lower distribution of neonatal birth weights (P = 0.000), but the actual between-group differences were minuscule. The areas under ROC curves were 0.572 (P = 0.001), 0.579 (P = 0.015) and 0.565 (P = 0.000) for preterm birth, preeclampsia and SGA, respectively. Subdivisions for the disorders did not obviously improve the performances of MS-AFP. CONCLUSIONS: Elevated first-trimester MS-AFP is associated with increased risk of preterm birth, preeclampsia and SGA. However, the predictive efficiencies were low and it is not a good predictor for these APOs.