Clustering Longitudinal Blood Pressure Trajectories to Examine Heterogeneity in Outcomes Among Preeclampsia Cases and Controls

Abstract

Preeclampsia is a heterogeneous disease characterized by new onset of hypertension along with signs of organ damage, affects 2% to 8% of pregnancies, and can result in serious complications to the mother and her child. There is little empirical evidence on the clinical importance of differences in blood pressure trajectories over the course of pregnancy, particularly in pregnancies affected by preeclampsia. We undertook an investigation of longitudinal changes in gestational blood pressure in a nested case-control study of preeclampsia in MoBa (Norwegian Mother, Father and Child Cohort Study). We included 1906 validated preeclampsia cases and 1413 validated controls. We derived blood pressure trajectory clusters using longitudinal k-means clustering and examined demographic and early-pregnancy predictors and birth outcomes, in relation to clusters. Maternal age, prepregnancy body mass index, and parity were substantially different across blood pressure clusters of cases. Pregnancy outcomes, including preterm birth, small for gestational age, and birthweight Z score, were meaningfully worse for individuals with a more rapid increase in blood pressure, as well as for individuals with a high starting blood pressure. For example, risk of preterm birth was 11-fold to 35-fold higher for steep and high trajectory clusters, and risk of small for gestational age was 2-fold higher compared with the reference cluster. Future studies may leverage these trajectories to differentiate preeclampsia cases in relation to circulating biomarkers, which may help in the development of preeclampsia prediction tools.