Results in the context of what is known
The literature on fetal growth and child cognitive development is highly heterogeneous in terms of how fetal growth restriction is defined (antenatal and/or postnatal) and how development is assessed (various measures of IQ and development are used); however, findings are consistent that infants with fetal growth restriction and/or small size are more likely to experience developmental challenges. Our finding that lower estimated fetal weight percentile is associated with greater developmental challenges is consistent with this literature. However, our study extends this work to move beyond population-level associations and evaluate the utility of estimated fetal weight charts as a clinical prediction tool, by examining measures such as sensitivity, specificity, positive predictive value, and negative predictive value. Our analyses highlight that, despite the population-level associations, EFW percentiles have limited ability as a tool to predict which specific children will go on to be developmentally vulnerable.
The magnitude of the association between EFW and developmental vulnerability observed in our study can be contextualized by comparing it to the effects of other determinants of developmental vulnerability. The link between low EFW percentile observed in our study (RR= 1.15 [95%CI: 0.65 to 1.86) and 1.49 [95%CI: 1.00 to 2.12] for fetuses <3rd percentile using the INTERGROWTH and WHO charts, respectively) are within the range of the magnitudes of increased risk associated with being born at late preterm gestation: Guthridge et al. found a two-fold increase of EDI developmental vulnerability in infants born between 34 and 36 weeks’ gestation (OR= 2.08 [95%CI: 1.27 to 3.39]), while Bentley et al. likewise found that EDI developmental vulnerability increased as gestational age decreased: 26% higher risks [95% CI: 1.18 to 1.34] at 34-36 weeks’ gestation compared with those born at 40 weeks.
Guthridge et al. found that a 5 minute Apgar <7 was associated with an 18% increased risk of developmental vulnerability on the EDI test [95% CI 0.46 to 3.01] whereas Razaz et al. reported that infants with 5-minute Apgar scores of 5,6,7, or 8 had increases in risk of EDI developmental vulnerability ranging from risk ratios of 1.1 to 1.5, compared with those with a 5-minute Apgar score of 10.
With regards to socio-economic factors, Guthridge et al found a two-fold increase of developmental vulnerability in infants of care-givers with 9 years or less of education [2.16 [95% CI: 1.40-3.33]). Likewise, Chittleborough et al. found that the most important early-life risks factors for developmental vulnerability at kindergarten age were socio-economic in nature (mother’s age, mother’s marital status, mother’s occupation, father’s occupation, number of previous pregnancies and smoking in the second half of pregnancy). Thus while risks of adverse child developmental outcomes are increased with lower EFW percentiles, they do not appear to be a dominant influence.
Not surprisingly, the association between low EFW percentile and developmental vulnerability is markedly lower than the association between low EFW and adverse neonatal outcomes. In previous work from this cohort of fetuses, we found that fetuses with an EFW less than the 10th centile were 3.1 fold more likely to have perinatal morbidity/mortality compared with fetuses between the 10th and 90th centile on the IG and WHO charts. However, this work, as well as that of others, also found that the charts have a poor ability to predict adverse outcomes at the individual level (i.e., poor sensitivity, specificity, and low AUC).