Birth weight and gestational age standards based on regional perinatal network data: an analysis of risk factors.

Abstract

The most frequently used set of gestational age-birth weight curves in the United States is the Colorado (C) standard published in 1963. To investigate the usefulness of this standard in an urban population at sea level, we examined the birth weight vs gestational age data from 56,675 singleton liveborn infants born between 1982 and 1985 in the University of Illinois (UI) perinatal network of 13 hospitals. Between 32 and 42 weeks, the UI 10th, median, and 90th percentile weights were significantly higher than those of Colorado. At term gestations the Colorado 10th and 90th percentile weights were the same as the UI 3rd and 80th percentile weights, respectively. Using the UI and Colorado standards for 10th and 90th percentile weights, the study sample was divided into five subgroups. To evaluate the risk prevalence, we examined the frequency of neonatal death, low, and very low Apgar scores (below seven and three, respectively), respiratory distress, maternal hypertension, and diabetes in the five subgroups. The highest frequency of adverse factors was seen in infants classified as small for gestational age (SGA) by both standards, but overall, the size-for-gestation grouping was better accomplished using the UI rather than the C standard. In 9188 infants (16.2%) classified into wrong weight-gestation subgroups using the Colorado standard, the prevalence of actual risk factors was at variance with the group to which they were assigned. This included 3632 (6.4%) SGA infants who were grouped as appropriate for gestational age (AGA), and 5556 (9.8%) AGA infants grouped as large for gestational age.(ABSTRACT TRUNCATED AT 250 WORDS)