| Sumario: | Objective: Approximately one-third of women in the U.S. experience an adverse pregnancy outcome (APO), which are recognized as sex-specific cardiovascular disease (CVD) risk factors. We examine if APOs confer additional CVD risk beyond that of traditional CVD risk factors. Methods: Women, age 40–79, with a pregnancy history and no pre-existing CVD were identified in the electronic health record of one health system (n = 2306). APOs included any APO, hypertensive disease of pregnancy (HDP), and gestational diabetes (GDM). Hazard ratios of time to CVD event were estimated from survival models using Cox proportional hazard regression. Discrimination, calibration, and net reclassification of re-estimated CVD risk prediction models including APOs were examined. Results: There was no significant association between any APO, HDP, or GDM and time to CVD outcome in survival models (95% confidence intervals all include 1). Including any APO, HDP, GDM in the CVD risk prediction model did not significantly improve discrimination and there were no clinically relevant changes in net reclassification of cases and non-cases. The strongest predictor of time to CVD event in the survival models was Black race, with hazard ratios ranging from 1.59 to 1.62, statistically significant for all three models. Conclusion: Women with APOs did not have an additional risk of CVD, controlling for traditional risk factors in the PCE and this sex-specific factor did not improve risk prediction. Black race was consistently a strong predictor of CVD even with data limitations. Further study of APOs can help determine how to best use this information for CVD prevention in women. Significance: Adverse pregnancy outcomes (APOs), such as gestational diabetes or hypertensive disease of pregnancy, are recognized as sex-specific cardiovascular disease (CVD) risk factors. While professional societies recommend considering APOs in cardiovascular risk assessment, CVD risk prediction calculators only include traditional risk factors. This study adds to the small body of research on the quantitative contribution of APOs to CVD risk prediction by using a contemporary, diverse cohort derived from an electronic health record. Including APOs did not improve risk prediction beyond traditional risk factors. How to best incorporate a history of APOs into CVD prevention requires further study.
|
|---|