Amer J Perinatol
DOI: 10.1055/s-0037-1608791
Original Article
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

Effect of Population-Specific Birthweight Curves on Disparities in Perinatal Mortality in Small-for-Gestational Age Pregnancies

Jacob C. Larkin
Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania
Ashley I. Naimi
Department of Epidemiology, University of Pittsburgh School of Public Health, Pittsburgh, Pennsylvania
› Author Affiliations
Further Information

Publication History

24 August 2017

14 October 2017

Publication Date:
13 December 2017 (eFirst)


Objective To determine the effect of adopting sex or race/ethnicity-specific birthweight curves on small-for-gestational age (SGA)-associated mortality rates for specific populations.

Materials and Methods Analyzing 20,095,735 singleton pregnancies, we compared rates of perinatal death associated with SGA in distinct sex and racial/ethnic groups when SGA was defined using nonspecific, sex-specific, and race/ethnicity-specific birthweight curves.

Results With use of a nonspecific birthweight curve, the rate of perinatal death was higher for SGA males (20.4/1,000 [95% confidence interval (CI), 20.1, 20.7]) than SGA females [14.6/1,000 (95% CI, 14.4, 14.8)]. With a sex-specific curve, this disparity was reduced, measuring 17.7/1,000 (95% CI, 17.4, 17.9) for SGA males and 16.3/1,000 (95% CI, 16.1, 16.6) for females. Using a nonspecific birthweight curve, perinatal death rates were higher for non-Hispanic blacks (20.4/1,000 [95% CI, 20.0, 20.8]) than for all other racial/ethnic groups (15.9/1,000 [95% CI, 15.7, 16.1]). This difference increased with use of a race-specific birthweight curve: perinatal mortality was 29.7/1,000 (95% CI, 29.0, 30.3) for SGA blacks and 14.7/1,000 (95% CI, 14.6, 14.9) for all other racial/ethnic groups.

Conclusion Population-based differences in SGA-associated mortality are reduced with adoption of a sex-specific birthweight curve, but widen with use of a race/ethnicity-specific curve. These findings highlight the importance of outcomes analysis in the selection of diagnostic criteria for SGA.


This project is supported by NIH grant K12HD063087 (J.C.L.).

Supplementary Material