Objective To estimate the ability of maternal serum markers and uterine artery Doppler in predicting preeclampsia. Methods In this nested case-control study, maternal serum concentrations of cystatin C, beta2-microglobulin, serum amyloid A, C-reactive protein (CRP), and neopterin were measured, and resistance index of uterine artery blood flow was assessed in 45 women in whom preeclampsia subsequently developed and in 125 women with normal pregnancy outcome. Univariable regression analysis was performed to estimate correlations between serum markers and resistance index for the development of preeclampsia. Significant variables were identified using multiple logistic regressions. Results Maternal serum markers were measured at a median gestational age of 14.7 weeks (interquartile range 3.1) in control group members and 16.3 weeks (interquartile range 4.8) in the case group members, and uterine resistance index was measured at the second-trimester scan. Univariable logistic regression showed that women with subsequent preeclampsia had increased levels of cystatin C, beta2-microglobulin, neopterin, CRP, and resistance index. Cystatin C, CRP, and resistance index remained independently associated with preeclampsia when multiple logistic regression was applied. Receiver-operating characteristic curve analysis showed that a combination of markers had a better area under the curve (AUC; 0.825) than when used in isolation (cystatin C, AUC 0.725; CRP, AUC 0.634; resistance index, AUC 0.728). Sensitivity of uterine artery resistance index, cystatin C, and CRP combined for predicting preeclampsia was 69.2% for a screen-positive rate of 15%. Conclusion Maternal serum cystatin C, CRP, and uterine artery mean resistance index are independent predictors of preeclampsia. There is improved prediction of preeclampsia when serum markers are combined with Doppler indices.
