Objective: Preterm premature rupture of the membranes (PPROM) is believed to be caused, in part, by abnormalities of collagen and increased levels of oxidative stress. Elevated homocysteine levels have been shown to induce these same pathophysiologic changes. We tested the hypothesis that serum homocysteine levels would be higher in women with PPROM when compared with matched control women.
Study design: A secondary analysis derived from 2 previously completed studies performed in the National Institutes of Child Health and Human Development Maternal-Fetal Medicine Units (MFMU) Network. We identified 99 study cases with PPROM (24 to 32 weeks' gestation) and matched them with 99 asymptomatic control women from an observational study of preterm birth prediction. Cases and control women were matched for race, gestational age at sampling, and MFMU Network center. Serum homocysteine levels were determined by immunoassay in batch fashion by personnel masked to study arm and clinical outcomes. Serum homocysteine levels were compared between groups, as were the baseline characteristics of maternal age, cigarette smoking, nulliparity, infections during pregnancy, and body mass index (BMI) <19.8 kg/m 2. Serum homocysteine levels were dichotomized as >75th, 90th, and 95th %ile of control women, and the likelihood of elevated homocysteine levels was determined in women who smoked, had a BMI <19.8 kg/m 2, or who had PPROM. Statistical analyses included the Wilcoxon rank sum, chi-square, and Pearson correlation coefficient, where appropriate. Baseline characteristics were controlled with a logistic regression model.
Results: Serum homocysteine levels measured in patients with PPROM were not significantly different from matched control women: median and (25th to 75th %ile): 4.9 (3.5-6.2) vs 4.8 (3.9-6.2 micromol/L), P =.73. In our population, neither the number of cigarettes smoked ( r = -0.08, P =.57), nor BMI ( r = -0.08, P =.24) correlated with serum homocysteine levels. The strongest association was seen in women with PPROM having serum homocysteine levels >95th %ile of control women (odds ratio [OR] 2.7, P =.10). After adjusting for baseline characteristics, no correlation between serum homocysteine level and the presence of PPROM was seen, OR 1.0 (.9-1.1); P =.99.
Conclusion: Women presenting with PPROM did not have significantly increased serum homocysteine levels when compared with control women.