Predicting Birth Outcomes Based on Maternal Anthropometry Using Ordinal Logistic Regression Approach: A Hospital Based Cross-Sectional Study in Sri Lanka.

Abstract

Background: The state of maternal nutrition, denoted by the maternal anthropometric parameters (MAPs), plays a pivotal role in determining the birth outcomes.

Objectives: This study was conducted to determine whether MAPs could predict selected birth outcomes.

Materials and methods: A cross-sectional study was conducted among randomly selected 333 pregnant mothers admitted for delivery after 28 weeks of period of amenorrhea at a tertiary care maternity hospital in Sri Lanka. Pregnant mothers who were having multiple pregnancies, awaiting elective cesarean section due to past section, registered for booking visits after 12 weeks of gestation, and had pre-existing disease conditions that might affect anthropometric parameters were excluded from the sample. Information on MAPs and birth outcomes were extracted from medical records. Data were analysed using ordinal logistic regression.

Results: Mean ± standard deviation of pre-pregnancy weight (PPW), maternal height (MH), pregnancy weight gain (PWG), and pre-pregnancy body mass index (BMI) were 55.1 ± 12.8 kg, 154.7 ± 5.7 cm, 9.6 ± 4.1 kg, and 22.9 ± 4.9 kg/m2, respectively. Nearly half of the mothers had unsatisfactory prepregnancy BMI, while 68.5% of mothers had unsatisfactory PWG. Low birth weight was reported in 24.6%, while 18.3% and 12.1% had prematurity and APGAR scores less than nine at birth, respectively. Higher PPW and PWG predicted better birth weight. Satisfactory PWG was predictive of maturity at birth. Both a satisfactory PPW and MH emerged as predictors of a good APGAR score at birth.

Conclusions: PPW, PWG, and MH can significantly predict selected birth outcomes in singleton pregnancies. These predictions will be useful for the provision of better perinatal care through early identification of high-risk newborns.