Other risk factors and infant birthweight
Table 2 and Table 3 presents the three quantile regression models for SGA and LGA infants, and Figure S3 presents model estimates for continuous covariates. Infants born to older mothers were more likely to be SGA (10th percentile: β=-15.2, 95%CI=[-23.0, -7.34], p=0.001), but not LGA (p=0.59) and had a reduced median birthweight (50th percentile: β=-7.5, 95%CI=[-14.9, -0.05], p=0.05). While older fathers had a reduced risk of fathering an SGA infant (10th percentile: β=9.5, 95%CI=[4.15, 14.8], p=0.0005). Higher SEIFA scores were associated with increased median infant birthweight (β=44.3, 95%CI=[16.7, 71.8], p=0.002) and a reduced risk of an SGA infant (β=32.0, 95%CI=[7.92, 56.1], p=0.009.
Male infants had higher birthweight compared with females by a similar amount in all regression analysis (all p≤0.03). Birthweight increased non-linearly with gestational age with smaller increases in weight for gestational ages >40 weeks. This tapering in the increase in birth weight was more extreme for SGA (10thpercentile: p=0.003 and 5th percentile: p=0.004) than for LGA (90th percentile: p=0.16 and 95th percentile: p=0.03). Infants delivered by caesarean section were more likely to be heavier (90thpercentile: β=133, 95%CI=[62.3, 203], p<0.0002), however their risk for SGA was not different to babies delivered vaginally (p=0.82). There was no detectable influence of insemination method on infant birthweight.