Predicting the risk for hypertension and cardiovascular disease after pregnancy: A multistate model

Abstract Background Hypertensive disorders of pregnancy (HDP) are associated with an increased risk of future hypertension and cardiovascular disease (CVD). The aim of this study was to develop a prediction model to identify women at high risk for hypertension and CVD directly after pregnancy. Methods Using routine care data of 737,257 primiparous women who gave birth between 2006-2015, we developed a multistate model to predict hypertension (intermediate outcome), CVD (primary outcome: hospital admission or death from myocardial infarction or stroke), and non-CVD death as a competing risk. Follow-up started 6 weeks after delivery until CVD, non-CVD death, age 50, or December 2022. Predictors included pregnancy complications (HDP, small for gestational age, preterm birth, repeated miscarriages), diabetes, country of origin, and socioeconomic status (SES) proxies (income and education). We developed a time-dependent Cox model with forward stepwise selection, stratified by country of origin (Netherlands, Europe, outside Europe and Suriname). Model assumptions were checked by Schoenfelds residuals. Results During follow-up (median 11 years), 12,240 women (1.7%) developed hypertension and 2,171 (0.3%) experienced CVD. HDP was the strongest pregnancy-related predictor. Among women of Dutch origin with HDP, those with low SES had a higher risk of hypertension compared to those with high SES (15% vs. 5%). Similar trends were observed for other groups, with the highest risks among women of Surinamese origin. Graphical comparison of the Cox model curves with the Aalen-Johansen estimator indicated good model fit. Conclusions This model identifies women at high risk for hypertension and CVD directly after pregnancy, supporting early prevention strategies. HDP, especially in the presence of low SES and/or a Surinamese background, was associated with the highest risks. Targeted cardiovascular risk management for these groups could improve public health and reduce health disparities. Key messages • A multistate model can identify high-risk women for hypertension and CVD directly after pregnancy. • Preventive care should specifically target women directly after pregnancy, with HDP, low SES and/or from minority ethnic groups, such as women of Surinamese origin.