A conditional standard for the customized fetal growth trajectory in twin pregnancy

Abstract

Background: The difference in the birth weights between twins and singletons grows with advancing gestation. Although many fetal weight standards based on ultrasonographic measurements have been created for tracking fetal growth in twin pregnancies, their applicability to other groups is limited by the fact that they are population specific.

Objective: This study aimed to generate conditional centiles for growth assessment of twin fetuses and to compare them with other population-based growth standards for singleton and twin fetuses.

Study design: This was a retrospective, longitudinal study of ultrasound-based estimated fetal weight data of normal twin fetuses delivered after 34 weeks of gestation. Linear mixed effects models that adjusted for maternal characteristics, fetal gender, and chorionicity were used to evaluate the relationship between estimated fetal weight and gestational age. The estimated fetal weight reference values were calculated using conditional centile based on the estimated fetal weight at an earlier gestational age. To compare our results with previous studies, fetal growth curves were generated using a formula we created that included maternal characteristics and the estimated fetal weight at 24 weeks of gestation in these studies. In a subgroup analysis of our low-risk twin babies born at full term, we calculated the number of fetuses who were reclassified as being in the bottom 10th percentile using each of the previous population-based standard.

Results: A total of 2644 ultrasounds with a median of 4 scans per fetus from 572 twin pregnancies were included in this analysis. In the cohort, 36% of the fetuses were monochorionic. Maternal age, body mass index, and the interaction between fetal gender and chorionicity were significantly associated with estimated fetal weight. The predicted growth curves matched the growth standard for twins. In our low-risk group, when the singleton standard was used, the incidence of estimated fetal weight <10th percentile was above 20% from gestational week 24 to 38, and this incidence varied when reclassified using other population-based twin standards.

Conclusion: This conditional growth chart was specifically designed to assess fetal growth in twin pregnancies, but it is generalizable to other populations.

Keywords: chorionicity; conditional percentile; estimated fetal weight; fetal growth; nomogram; twin pregnancy.