BACKGROUND: Predictors of the clinical characteristics of pregnant women at risk of peripartum cardiomyopathy (PPCM) remain unclear. METHODS AND RESULTS: We enrolled 450 cases of high-risk pregnancy with any risk factor from among ≥35 years of age, hypertensive disorders of pregnancy (HDP), multiple gestation, or other systemic or obstetric complications except for a history of cardiac disease. All the women underwent echocardiography and plasma B-type natriuretic peptide (BNP) measurement during the second/third trimester and the early/late postpartum period. Logistic regression analyses identified clinical factors associated with left ventricular (LV) dysfunction. The incidence of PPCM was 0.89%, which was higher than in previous reports. Early diastolic mitral annular velocity (e') was significantly negatively associated with the occurrence of PPCM, and the BNP level showed a significant positive association with the occurrence of PPCM. The percentages of cases of e' <
7 cm/s, and BNP level ≥100 pg/mL were 25.3%, and 20.4%, respectively. Multivariate regression analysis revealed that HDP was independently associated with e'. A negative correlation between e' and BNP level was observed in HDP. CONCLUSIONS: High-risk pregnancy was associated with an increased incidence of PPCM. Measurement of BNP levels and echocardiographic assessment of LV diastolic function during pregnancy may be useful in predicting PPCM.