Prediction model for hypertension in first decade after pre-eclampsia in initially normotensive women

OBJECTIVE

To develop a prediction model for the development of hypertension in the decade following pre-eclampsia in women who were normotensive shortly after pregnancy.

METHODS

This was a longitudinal cohort study of formerly pre-eclamptic women attending a university hospital in The Netherlands between 1996 and 2019. We developed a prediction model for incident hypertension using multivariable logistic regression analysis. The model was validated internally using bootstrapping techniques.

RESULTS

Of 259 women, 185 (71%) were normotensive at the first cardiovascular assessment, at a median of 10 (interquartile range (IQR), 6-24) months after a pre-eclamptic pregnancy, of whom 49 (26%) had developed hypertension by the second visit, at a median of 11 (IQR, 6-14) years postpartum. The prediction model, based on birth-weight centile, mean arterial pressure, total cholesterol, left ventricular mass index and left ventricular ejection fraction, had good-to-excellent discriminative ability, with an area under the receiver-operating-characteristics curve (AUC) of 0.82 (95% CI, 0.75-0.89) and an optimism-corrected AUC of 0.80. The sensitivity and specificity of our model to predict hypertension were 98% and 34%, respectively, and positive and negative predictive values were 35% and 98%, respectively.

CONCLUSIONS

Based on five variables, we developed a good-to-excellent predictive tool to identify incident hypertension following pre-eclampsia in women who were normotensive shortly after pregnancy. After external validation, this model could have considerable clinical utility in tackling the cardiovascular legacy of pre-eclampsia. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Metabolic syndrome and pre-eclampsia

OBJECTIVE

To evaluate the association between different pre-eclampsia (PE) phenotypes and the development of metabolic syndrome postpartum, in order to identify the subgroup of formerly pre-eclamptic women with a worse cardiovascular risk profile requiring tailored postpartum follow-up.

METHODS

This was a cohort study of 1102 formerly pre-eclamptic women in whom cardiovascular and cardiometabolic evaluation was performed at least 3 months postpartum. Women were divided into four subgroups based on PE resulting in delivery before 34 weeks (early-onset (EO)) or at or after 34 weeks (late onset (LO)) of gestation and whether they delivered a small-for-gestational-age (SGA) neonate. Metabolic syndrome was diagnosed as the presence of hyperinsulinemia along with two or more of: body mass index ≥ 30 kg/m² ; dyslipidemia; hypertension; and microalbuminuria or proteinuria. Data were compared between groups using ANOVA after Bonferroni correction. Odds ratios (OR) were calculated using logistic regression to determine the association between metabolic syndrome and the four subgroups. We constructed receiver-operating characteristics curves and computed the area under the curve (AUC) to quantify the ability of different obstetric variables to distinguish between women who developed metabolic syndrome and those who did not.

RESULTS

The prevalence of metabolic syndrome was higher in women with EO-PE and SGA (25.8%) than in those with EO-PE without SGA (14.7%) (OR 2.01 (95% CI, 1.34-3.03)) and approximately five-fold higher than in women with LO-PE with SGA (5.6%) (OR 5.85 (95% CI, 2.60-13.10)). In women with LO-PE, the prevalence of metabolic syndrome did not differ significantly between women with and those without SGA. Multivariate analysis revealed that a history of SGA, a history of EO-PE and systolic blood pressure at the time of screening are the best predictors of developing metabolic syndrome postpartum. The AUC of the model combining these three variables was 74.6% (95% CI, 70.7-78.5%). The probability of the presence of metabolic syndrome was calculated as: P = 1/(1 + e^-LP ), where LP is linear predictor = -8.693 + (0.312 × SGA (yes = 1)) + (0.507 × EO-PE (yes = 1)) + (0.053 × systolic blood pressure).

CONCLUSIONS

The incidence of metabolic syndrome postpartum was associated more strongly with EO-PE in combination with SGA as compared with LO-PE or EO-PE without SGA. Both time of onset of PE and fetal growth affect the risk of metabolic syndrome after a pre-eclamptic pregnancy. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.