Birth weight percentiles in relation to different GA estimates among singleton stillbirths.

Birth weight percentiles for stillbirths ≥20 wk gestation using three GA estimates: GA at delivery, GA at delivery minus 2 d, and GA at death estimated using the SCRN algorithm [17]. Percentages may add to slightly more or less than 100% because of rounding.aUnadjusted OR for stillbirth for infants with birth weight in the percentile group shown compared to infants in the reference group from a logistic regression model that included effects for percentile group only.bAdjusted OR for stillbirth for infants with birth weight in the percentile group shown compared to infants in the reference group from a logistic regression model that in addition to the percentile group indicators included study site number; paternal age (Table 1 or as noted): maternal age (cAnalysis weights that accounted for the basic study design plus other aspects of the sampling were used.Unweighted sample sizes were 527, 561, and 570 stillbirths for GA at death, GA at delivery minus 2 d, and GA at delivery, respectively, and 1,821 live births. Unweighted (weighted) sample sizes included in computation of adjusted ORs were 491 (489) stillbirths and 1,665 (1,261) live births.dIndividualized norm percentiles were derived using the fetal weight for GA equation from Bukowski et al. [15].eUltrasound norm percentiles were derived using the fetal weight for GA equation and standard error from Hadlock et al. [19].fAlexander et al. population norm percentiles of birth weight for GA were used [18].Simple linear interpolation was used with the Alexander et al. birth weight percentiles reported for completed weeks of GA in whole weeks to derive birth weight percentiles for GA in weeks and days.