Abstract
IMPORTANCE: Cardiovascular disease (CVD) is the leading cause of death among women worldwide. Pregnancy serves as a natural cardiovascular stress test and universal clinical encounter, yet few approaches leverage its insights to inform long-term cardiovascular risk. OBJECTIVE: To determine whether clinical measures and biomarkers obtained during pregnancy may identify women at risk of long-term CVD. DESIGN, SETTING, AND PARTICIPANTS: This was a registry-linked, population-based cohort study of all pregnancies reaching at least 22 weeks between June 2010 and October 2013 in Southern Denmark. Primary analyses were performed in a nested prospective subcohort of Odense Child Cohort participants with available pregnancy biomarker data. Women with preexisting CVD were excluded (n = 114). Follow-up was done through December 31, 2023. Among 38 455 eligible women, 2056 had biomarker data at week 12 or week 29. Analytic subsets with complete data were used for prognostic modeling at week 12 (n = 1379) and week 29 (n = 1389). EXPOSURES: Clinical characteristics, obstetric outcomes, and pregnancy biomarkers including soluble fms-like tyrosine kinase-1 (sFlt-1), placental growth factor, high-sensitivity cardiac troponin I (hs-cTnI), and N-terminal pro-B-type natriuretic peptide. MAIN OUTCOMES AND MEASURES: Incident maternal CVD, evaluated using Cox proportional hazards models. RESULTS: In the biomarker cohort (median [IQR] age, 30.4 [27.4-33.8] years), 28 women (1.4%) developed CVD during a median (IQR) follow-up of 11.9 (11.2-12.5) years. Maternal age, hypertensive disorders of pregnancy (HDPs), and third-trimester concentrations of hs-cTnI and sFlt-1 were each independently associated with higher long-term CVD risk. A combined model including age and sFlt-1 measured at week 29 improved discrimination for CVD compared with a base model of age alone (ΔAUC, 0.16; 95% CI, 0.02-0.30), whereas a clinical model consisting of age, systolic blood pressure, and non-high-density lipoprotein cholesterol did not. Results were consistent in women without prior hypertension or HDPs and in nulliparous women. CVD incidence and the predictive value of the base model were comparable between the biomarker and the contemporaneous background cohorts (n = 36 274). CONCLUSIONS AND RELEVANCE: These findings support pregnancy as an opportunistic window for sex-specific cardiovascular risk assessment and prevention throughout a woman's life course. Further studies are warranted to validate these findings.