Refined fetal abdominal growth assessment in normal pregnancy: Part I. Abdominal anteroposterior diameter.

BACKGROUND: The assessment of the adequacy of fetal growth by parameters other than the abdominal anteroposterior diameter (AAPD) of the fetus has been studied extensively. However the designs of these studies and the statistical methods used appears to deserve some criticism, based on present knowledge. Noncross-sectional cases selection, uncertainty of the fetal normality and inadequacy in statistical method, mostly ignored the changing property of each standard deviation (SD) of each gestational age (GA) which was proposed by Altman et. al. in 1993, are the three most common flaws in previous publishes. We tried to use AAPD with a strict study design as well as a reasonable statistical method to evaluate the fetal growth.

METHODS: This study was performed in the Division of Maternal Fetal Medicine, National Cheng Kung University Hospital, Taiwan. The prenatal sonographic data of the fetuses were collected prospectively based on the following criteria: (1) accurate dating by knowing the maternal last menstration period (LMP) and early ultrasonography of the fetuses, (2) singleton pregnancy, (3) no fetal structural or chromosomal abnormality confirmed by prenatal ultrasonography and postnatal examination, (4) GA at birth was between 37 and 41 weeks' gestation, (5) no birth asphyxia, (6) appropriate birth body weight, and (7) no maternal medical disease or obstetrical complication which might predictably interfere with fetal growth. The collected data were analyzed by polynomial regression test and the best-fit equation for prediction of fetal growth was selected. The standard deviation (SD) of each GA was modeled before constructing the fetal growth centile charts by Altman's method.

RESULTS: A total of 2077 cross-sectional sonographic data meeting the above criteria were collected for statistical analysis. The best-fit equation for the prediction of fetal GA by AAPD is GA = 20.8539 - 3.36743 x AAPD + 0.86927 x (AAPD)2 - 0.03789 x (AAPD)3 +/- k x [1.2533 x (0.36772 + 0.10938 x AAPD)], (r = 0.97287, p < 0.0001). The best-fit equation for prediction of fetal AAPD by GA is AAPD = -2.49495 + 0.38247 x GA - 1.07071 x 0.001 x (GA)2 +/- k x [1.2533 x (0.01760 + 0.01372 x GA)], r = 0.97122, p < 0.0001. The SD of AAPD for each complete GA was not the same. The fetal growth centile charts in the study are presented in this article.

CONCLUSIONS: The SD of each complete GA changed with each specific GA. The GA of the fetus can be assessed accurately by measuring the AAPD alone. The utilization of these growth centile charts for evaluation of fetal growth is recommended.