Center for Genomic and Personalized Medicine, Guangxi Medical University, Nanning 530021, Guangxi, China; School of Public Health of Guangxi Medical University, Nanning 530021, Guangxi, China.
Department of Nutrition and Food Hygiene, School of Public Health, Guangxi Medical University, Nanning 530021, Guangxi, China.
Cytokine. 2019 May;117:91-97. doi: 10.1016/j.cyto.2019.02.007. Epub 2019 Mar 1.
Current biomarkers such as fetal fibronectin and cervical length are accurate predictors of spontaneous preterm birth (sPTB) in women with clinically suspected preterm risk; however, these are not effective for predicting the risk of sPTB in asymptomatic women. Therefore, we performed this study with the objective of determining whether the combinations of specific serum cytokines could accurately predict the sPTB risk in asymptomatic women.
We conducted a nested case-control study with 129 incident sPTB cases and 258 individually matched controls who participated in an ongoing birth cohort study. The maternal serum levels of the selected 35 cytokines were measured. We evaluated the relationship between the multiple cytokines and sPTB risk using conditional logistic regression and elastic net model.
A panel of cytokines was significantly associated with an increased risk of sPTB. The odds ratio (OR) of sPTB per standard deviation (SD) increase of the predictive model score was 1.57 (95% CI 1.25-1.97) for the cytokines model. The combination of the selected serum cytokines was substantially more effective in predicting the risk for sPTB, as the receiver-operator characteristic curve (AUC) values were 0.546 and 0.559 in the single cytokine model and it improved to 0.642 in the multiple cytokines model (P = 0.02 for TNF-α vs. multiple cytokines; P = 0.05 for TRAIL vs. multiple cytokines). Moreover, the prediction was more accurate in overweight pregnant women, with an AUC = 0.879.
The current study suggested that the combination of selected serum cytokines can more effectively predict the risk of sPTB in asymptomatic women compared with the use of single cytokine.
目前的生物标志物，如胎儿纤维连接蛋白和宫颈长度，是对有临床早产风险的孕妇发生自发性早产（sPTB）的准确预测指标；然而，它们对预测无症状孕妇发生 sPTB 的风险并不有效。因此，我们进行了这项研究，旨在确定特定血清细胞因子的组合是否可以准确预测无症状妇女发生 sPTB 的风险。
我们进行了一项嵌套病例对照研究，纳入了 129 例 sPTB 病例和 258 例个体匹配的对照，这些参与者参加了一项正在进行的出生队列研究。测量了选定的 35 种细胞因子的母血清水平。我们使用条件逻辑回归和弹性网络模型评估了多种细胞因子与 sPTB 风险之间的关系。
一组细胞因子与 sPTB 风险增加显著相关。预测模型评分每增加一个标准差（SD），sPTB 的优势比（OR）为 1.57（95%CI 1.25-1.97）。所选血清细胞因子的组合在预测 sPTB 风险方面更有效，因为在单个细胞因子模型中，受试者工作特征曲线（AUC）值分别为 0.546 和 0.559，而在多个细胞因子模型中则提高至 0.642（TNF-α 与多个细胞因子相比，P=0.02；TRAIL 与多个细胞因子相比，P=0.05）。此外，在超重孕妇中，预测更为准确，AUC 值为 0.879。
本研究表明，与使用单个细胞因子相比，选定的血清细胞因子组合可以更有效地预测无症状妇女发生 sPTB 的风险。
