Mid-gestational factors predict small-for-gestational age

Risk prediction models including factors such as maternal smoking status, body mass index (BMI), and parity, can help predict small-for-gestational age (SGA) during pregnancy, a recent Japan study has found.

Through a prospective cohort study of 17,073 consenting pregnant women, the researchers aimed to construct risk scores for SGA during the early (11–17 weeks) and mid-gestational (18–21 weeks) periods of pregnancy. Participant data were drawn from the Tohoku Medical Megabank Project and the Three-Generation Cohort Study.

A total of 1,126 infants were deemed to be SGA, yielding an overall prevalence rate of 6.6 percent. Multiple logistic regression analysis yielded an early gestational predictive model that looked at the following factors: maternal age, BMI, height, parity, smoking status, blood pressure (BP), maternal birthweight, and the use of assisted reproductive technology (ART) with frozen-thawed embryo transfer (FET).

After tenfold cross-validation, the resulting C-statistic for the early gestation model was 0.659 (95 percent confidence interval [CI], 0.642–0.676).

During mid-gestation, the most pertinent factors for SGA risk prediction also included maternal age, height, and BMI, as well as parity, smoking status, BP, and the use of ART with FET. In addition, weight gain and estimated foetal weight during mid-gestation were additionally included in the mid-gestational model, which had a tenfold cross-validated C-statistic of 0.726 (95 percent CI, 0.710–0.741).

“Our prediction model for SGA infants, particularly during mid-gestation, may aid in the detection of pregnant women at a high risk of delivering SGA infants in Japan. Further studies for its external validation and improvement of its predictive ability are necessary,” the researchers said.