Joint Modeling of Preconception Nutrition, Fetal Growth, and Adverse Birth Outcomes: A Longitudinal Study in Padangsidimpuan

Preconception undernutrition remains a major determinant of impaired fetal growth and adverse birth outcomes in low- and middle-income countries. In Padangsidimpuan, North Sumatra, the stunting prevalence among children under five reached 32.01% in 2021, substantially above the national average. However, evidence integrating nutritional status before conception, longitudinal fetal growth, and neonatal outcomes within a single analytic framework is limited in Indonesia. This study aims to propose a longitudinal cohort design and joint modeling framework to identify high-risk preconception nutritional profiles, quantify their impact on fetal growth trajectories and adverse birth outcomes, and provide an evidence base for preconception nutrition programs within local stunting-reduction strategies. This prospective longitudinal cohort study protocol will enroll 1,000 non-pregnant women aged 18-35 years who intend to conceive within 12 months and will be recruited through community health centers in Padangsidimpuan. Exclusion criteria include chronic diseases and multiple gestations. estimation. Preconception nutrition status will be measured by anthropometry (BMI, MUAC), hemoglobin measurement, dietary diversity scores and supplement use into a composite nutrition score using principal component analysis. Fetal growth will be measured by ultrasound at 3 time points ≤14 weeks, 20–24 weeks and 32–36 weeks with INTERGROWTH-21st standards; birth outcomes are low birth weight (<2500g); preterm birth (<37 weeks) and small for gestational age (<10th percentile). Shared parameter joint model is used to relate longitudinal fetal growth trajectories to binary adverse outcomes via random effects (maximum likelihood estimation). Illustrative analysis using hypothetical data consistent with Indonesian patterns (N=1,000) demonstrated that women in the highest nutrition tertile had higher mean fetal weight z-scores at 24 weeks (0.18 vs -0.32) and greater growth velocity (0.11 vs -0.13) than those in the lowest tertile. Adverse outcomes were substantially lower in the high tertile (10%) than in the low tertile (27%). Joint modeling found that each standard deviation increase in nutrition score was associated with a 0.22 higher fetal weight z-score (95% CI: 0.14–0.30) and 36% lower odds of adverse outcomes (OR=0.64; 95% CI: 0.45–0.85). The joint modeling framework proposed here provides a holistic view of pathways from preconception nutrition to adverse birth outcomes. If implemented, this study could provide high value evidence for targeted preconception interventions and stunting reduction strategies in Padangsidimpuan and other settings in Indonesia.