Predictive ability of fetal growth charts in identifying kindergarten-age developmental challenges: a cohort study

Abstract

Background: The Society for Maternal-Fetal Medicine recommends defining fetal growth restriction as an estimated fetal weight or abdominal circumference <10th percentile of a population-based reference. However, because multiple references are available, an understanding of their ability to identify infants at increased risk due to fetal growth restriction is critical. Previous studies have focused on the ability of different population references to identify short-term outcomes, but fetal growth restriction also has longer-term consequences for child development.

Objective: This study aimed to estimate the association between estimated fetal weight percentiles on the INTERGROWTH-21^st and World Health Organization fetal growth charts and kindergarten-age childhood development, and establish the charts' discriminatory ability in predicting kindergarten-age developmental challenges.

Study design: We conducted a retrospective cohort study linking obstetrical ultrasound scans conducted at BC Women's Hospital, Vancouver, Canada, with population-based standardized kindergarten test results. The cohort was limited to nonanomalous, singleton fetuses scanned at ≥28 weeks' gestation from 2000 to 2011, with follow-up until 2017. We classified estimated fetal weight into percentiles using the INTERGROWTH-21^st and World Health Organization charts. We used generalized additive modeling to link estimated fetal weight percentile with routine province-wide kindergarten readiness test results. We calculated the area under the receiver-operating characteristic curve and other measures of diagnostic accuracy with 95% confidence intervals at select percentile cut-points of the charts. We repeated analyses using the Hadlock chart to help contextualize findings. The main outcome measure was the total Early Development Instrument score (/50). Secondary outcomes were Early Development Instrument subdomain scores for language and cognitive development, and for communication skills and general knowledge, as well as designation of "developmentally vulnerable" or "special needs".

Results: Among 3418 eligible fetuses, those with lower estimated fetal weight percentiles had systematically lower Early Development Instrument scores and increased risks of developmental vulnerability. However, the clinical significance of differences was modest in magnitude (eg, total Early Development Instrument score -2.8 [95% confidence interval, -5.1 to -0.5] in children with an estimated fetal weight in 3rd-9th percentile of INTERGROWTH-21^st chart [vs reference of 31st-90th]). The charts' predictive abilities for adverse child development were limited (eg, area under the receiver-operating characteristic curve <0.53 for all 3 charts).

Conclusion: Lower estimated fetal weight percentiles on the INTERGROWTH-21^st and World Health Organization charts indicate increased risks of adverse kindergarten-age child development at the population level, but are not accurate individual-level predictors of adverse child development.

Keywords: child development; estimated fetal weight; fetal growth; growth charts.