Bayesian Joint Modeling of Longitudinal Blood Pressure and Time-to-Preeclampsia in Pregnant Women: A Retrospective Cohort Study

Background Hypertensive disorders of pregnancy, particularly preeclampsia, remain major contributors to maternal morbidity and mortality. This study jointly modeled longitudinal systolic and diastolic blood pressure trajectories with time to onset of preeclampsia. Methods A retrospective cohort study was conducted among 549 pregnant women receiving antenatal care at Bishoftu General Hospital. Descriptive statistics, profile plots, and Kaplan–Meier estimates were used for preliminary analyses. A Bayesian multivariate joint model with shared random effects simultaneously evaluated repeated blood pressure measurements and time-to-preeclampsia, adjusting for maternal covariates. Results Preeclampsia occurred in 10.9% of women. Each one-unit increase in systolic and diastolic blood pressure was associated with 6% and 7% higher hazards of preeclampsia, respectively. Higher baseline blood pressure levels were linked to shorter time to onset. Additional risk factors included maternal age ≥ 30 years, higher maternal weight, family history of hypertension, prior induced abortion, and primigravidity. Adverse outcomes, including stillbirth and pregnancy-related complications, were more frequent among women who developed preeclampsia. Conclusion Bayesian joint modeling demonstrates the prognostic importance of longitudinal blood pressure trajectories for early identification of preeclampsia, supporting integrated monitoring strategies to improve maternal outcomes.