Trajectory modeling of gestational weight: A functional principal component analysis approach

Area of Science:

• Obstetrics and Gynecology
• Maternal-Fetal Medicine
• Biostatistics

Background:

• Suboptimal gestational weight gain (GWG) poses risks to both mother and infant.
• Prevalence of inadequate or excessive GWG necessitates effective monitoring strategies.
• Accurate tracking of GWG is crucial for identifying at-risk pregnancies.

Purpose of the Study:

• To estimate individual weight growth trajectories during pregnancy using sparse data.
• To identify key factors influencing GWG, including prepregnancy BMI, diet, and physical activity.
• To develop a novel, adaptable tool for real-time GWG monitoring and intervention.

Main Methods:

• Functional Principal Component Analysis (FPCA) by conditional expectation for trajectory modeling.
• Linear regression analysis to determine factors affecting total weight gain.
• Comparison of FPCA model performance against traditional nonlinear mixed-effect models.

Main Results:

• FPCA demonstrated superior adaptability and a significantly lower root mean square error (RMSE) for trajectory modeling compared to nonlinear mixed-effect models.
• Prepregnancy body mass index (BMI) was identified as a highly predictive factor for weight changes during pregnancy.
• Findings align with existing guidelines for recommended gestational weight gain.

Conclusions:

• FPCA offers a powerful and accurate method for modeling individual GWG trajectories from sparse data.
• Prepregnancy BMI is a critical determinant of GWG, supporting its use in clinical guidelines.
• This novel approach can enhance real-time monitoring and facilitate timely interventions for suboptimal GWG.