Caesarean delivery without TOL
Caesarean delivery without TOL risks were higher for SGA <5th percentile (25.2% versus 23.7%) and SGA <3rd percentile (26.2% versus 23.7%;Table 1 ) compared to AGA. A more nuanced picture emerges when this outcome is stratified by gestational age. For example, 77.5% of women who delivered at <34 weeks with SGA <5th percentile were delivered by caesarean without TOL compared to 48.3% of AGA births whereas the risk was substantially lower at 41 weeks and beyond (10.9% versus 8.0%). Although risks of caesarean delivery without TOL decreased across gestational age for SGA and AGA births, the caesarean risks without TOL were higher for SGA <5th percentile at each gestational age strata except for 39-40 weeks’ gestation (when 22.9% of SGA <5th percentile births were Caesarean without TOL compared to 24.5% of AGA births). These same trends, though more pronounced, were observed for SGA <3rdpercentile births.
The results of the mediation analysis for caesarean delivery without TOL and exposure to SGA <5th percentile are described in Table 2 . For SGA versus AGA births, the adjusted excess risk of caesarean delivery without trial of labour had a “U” shaped association, with increased risk at preterm gestations, lowest risk at 39-40 weeks, and increased risk at ≥41 weeks. The decomposition analysis revealed that the driver of this excess risk was the CDE with minimal mediation and interaction components. The mediation analysis for this outcome and exposure to SGA <3rd percentile revealed similar, but more pronounced, patterns (Table 3 ).