Birth weight charts for a Chinese population: an observational study of routine newborn weight data from Chongqing

Background To construct birth weight charts for the Chongqing municipality, China and to identify whether differences in birth weight exist across urban/rural populations, thereby warranting separate charts. Methods Secondary analysis of routinely collected data from 338,454 live infants between 2014 and 2017 in Chongqing municipality. Sex-specific birth weight-for-gestational age centiles were constructed by the lambda-mu-sigma method via the GAMLSS R-based package. This method remodels the skewed birth weight distribution to estimate a normal distribution, allowing any birth weight centile to be generated. A separate set of centiles were created, accounting for urban/rural differences in birth weight. Results The centiles performed well across all gestational ages. For example, 2.37% (n = 4176) of males and 2.26% (n = 3656) of females were classified as below the 2nd centile (expected percentage = 2.28%), 49.75% of males (n = 87,756) and 50.73% of females (n = 82,203) were classified as below the 50th centile (expected proportion = 50%) and 97.52% of males (n = 172,021) and 97.48% of females (n = 157,967) were classified as below the 98th centile (expected proportion = 97.72%). The overall estimated centiles of birth weight for rural infants were higher than the centiles for urban infants at the earlier gestational ages (< 37 gestational weeks). However, this trend was reversed in infants born at term. Conclusion We have constructed a readily utilizable set of birth weight references from a large representative sample of births in Chongqing. The method used to construct the references allows for the calculation of the exact centile for any infant delivered between 28 and 42 completed weeks, which was not possible with previous charts.