Second- to Third-Trimester Longitudinal Growth Assessment for the Prediction of Largeness for Gestational Age and Macrosomia in an Unselected Population.

BACKGROUND: Prenatal detection of excessive growth remains inaccurate. Most strategies rely on a single cross-sectional evaluation of fetal size during the third trimester.

OBJECTIVES: To compare second- to third-trimester longitudinal growth assessment with cross-sectional evaluation at the third trimester in the prediction of largeness for gestational age (LGA) and macrosomia.

METHODS: A cohort of 2,696 unselected singleton pregnancies scanned at 21 ± 2 and 32 ± 2 weeks was created. Abdominal circumference (AC) measurements were transformed to z values according to the INTERGROWTH-21st standards. Longitudinal growth assessment was performed by calculation of z velocity and conditional growth. Both methods were compared to cross-sectional assessment at 32 ± 2 weeks. Predictive performance for LGA and macrosomia was determined by receiver operating characteristic curve analysis.

RESULT: A total of 188 (7%) newborns qualified for LGA and 182 (6.8%) for macrosomia. The areas under the curve (AUCs) for 32-week AC z score, AC z velocity, and conditional AC were 0.78, 0.61, and 0.55, respectively, for the prediction of LGA, and 0.75, 0.61, and 0.55, respectively, for the prediction of macrosomia. Both AUCs of AC z velocity and conditional AC were significantly lower (p < 0.001) than the AUC of cross-sectional AC z scores.

CONCLUSIONS: In the general population, second- to third-trimester longitudinal assessment of fetal growth is inferior to third-trimester cross-sectional evaluation of size in the prediction of LGA and macrosomia.