Cardiovascular system during the postpartum state in women with a history of preeclampsia

Abstract

In subjects with previous preeclampsia, differences in cardiovascular and/or blood biochemical parameters are present in the nonpregnant state, and a simultaneous assessment of multiple derived indices better differentiates between women with or without previous preeclampsia. We examined 18 previous preeclamptic and 50 previous uncomplicated pregnancies, ≈16 months postpartum. Cardiovascular assessment included the following: (1) systemic hemodynamics and mechanics (Doppler echocardiography, tonometry, and oscillometric sphygmomanometry); (2) endothelial function (plethysmography); (3) left ventricular properties (echocardiography); and (4) blood biochemical analyses. Compared to women with previous uncomplicated pregnancies, previous preeclamptics had higher mean (80±1 versus 86±3 mm Hg; P=0.04) and diastolic (64±1 versus 68±2 mm Hg; P=0.04) pressures and total vascular resistance (1562±37 versus 1784±114 dyne · s/cm(5); P=0.03). Systolic blood pressure, arterial compliance, and left ventricular properties were not different. Although heart-to-femoral pulse wave velocity was not different, heart-to-brachial pulse wave velocity tended to be faster in previous preeclamptics (374±8 versus 404±20 cm/s; P=0.06). Stress-induced increase in forearm blood flow was less in previous preeclamptics (245%±21% versus 136%±22%; P=0.01), indicating impaired endothelial function. No significant differences were observed in markers of endothelial activation, dyslipidemia, or oxidative stress; previous preeclamptics tended to have higher glucose level (58.7±1.9 versus 95±5.2 mg/dL; P=0.06). Logistic regression analysis indicated that a simultaneous evaluation of multiple derived indices better discriminated between the 2 groups. The differences in the previous preeclamptic group are in directions known to be associated with greater cardiovascular disease risk later in life.

Figure 1

Endothelial function by venous occlusion plethysmography. FBF indicates forearm blood flow. Data are mean ± SEM. *P<0.05, prior preeclampsia versus prior uncomplicated pregnancy by univariate logistic regression.

Figure 2

Receiver operating characteristic (ROC) curves for various logistic regression analyses. The multiple logistic regression equation used to construct the ROC curve (solid line) was logit(y) = −5.56 + (−0.551•excess FBF) + (0.141•MAP) + (−1.972•BSA), with the area under the curve of 0.82. Univariate logistic regression equation used to construct the ROC curve for excess FBF (dash-to-dot line) was logit(y) = 0.573 + (−0.364•excess FBF); with the area under the curve of 0.71. Univariate logistic regression equation used to construct the ROC curve for MAP (dash line) was logit(y) = −6.269 + (0.063•MAP), with the area under the curve of 0.66.