Accelerated Biological Age as an Early Pregnancy Risk Factor for Preeclampsia

The inability to predict risk in early pregnancy for preeclampsia represents a major limitation in prenatal care.

We used machine-learning approaches to identify early pregnancy features that distinguish women who develop preeclampsia from those with normotensive pregnancies, specifically focusing on the influence of maternal biological age.

Data were analyzed from a prospective cohort of pregnant women living in the upper Midwest (Midwest cohort) and from a cohort using publicly available data from the Prenatal Exposures & Preeclampsia Prevention Project (PEPP3). In both data sets, DNA methylation (DNAm) was quantified from blood samples collected in early pregnancy. Biological aging was estimated using established epigenetic clocks including Hannum, Horvath, and PhenoAge. Maternal data across pregnancy were collected via medical record abstraction. Predictors of preeclampsia were identified using LASSO regression and Random Forest. The potential predictive capacity of the selected features was evaluated by building logistic regression models with leave-one-out cross-validation and reporting performance metrics.

In the Midwest cohort, accelerated biological aging per Hannum's epigenetic clock as well as higher chronological age and higher gestational weight gain were identified as important predictors of preeclampsia. Interestingly, only Hannum's epigenetic clock showed predictive power for preeclampsia. In the PEPP3 cohort, accelerated biological aging, as measured by Hannum's epigenetic clock, and chronological age were also identified as significant predictors of preeclampsia.

Among women who developed preeclampsia, accelerated biological aging during early pregnancy may represent a risk biomarker that can be leveraged in clinical care. These findings identify a promising clinical indicator for preeclampsia risk, addressing a critical gap in screening and early diagnosis.