晚孕期母体血糖指标与胎儿胃超声参数预测GDM新生儿早期喂养障碍
多中心前瞻性队列研究纳入妊娠期糖尿病单胎孕妇,于孕32+0至36+6周测量母体血糖指标、胎儿胃超声参数及常规生物测量。以出生72小时内发生喂养障碍为主要结局,构建列线图模型。训练、验证、外部测试队列AUC分别为0.906、0.883、0.876,显示模型具有良好区分度与潜在临床效用。
要点
- 胎儿胃容积Z值、胃排空率Z值、胃围/腹围比值是预测新生儿喂养障碍的关键超声参数。
- 晚孕期平均餐后2小时血糖进入最终预测模型,超声报告中应关注母体血糖控制情况。
- 模型AUC达0.876–0.906,可用于产前识别喂养障碍高危新生儿。
核心结论
- 整合晚孕期母体血糖指标与胎儿胃超声参数的产前模型对GDM孕妇新生儿早期喂养障碍具有良好预测性能。
研究设计与样本
- 来自一个中心的病例随机分为训练队列和验证队列,其他三个中心病例构成外部测试队列。
- 结局为出生后72小时内发生早期喂养障碍。
- 候选预测因子包括母体特征、晚孕期血糖指标、常规胎儿生物测量及孕32+0至36+6周胎儿胃超声参数。
关键影像表现/征象
- 胎儿胃超声参数包括胃容积Z值、胃排空率Z值、胃围/腹围比值。
- 最终模型纳入的超声参数与实验室指标反映胎儿胃功能状态及母体血糖控制水平。
诊断标准与关键数值
- 训练队列早期喂养障碍发生率为21.3%,验证队列为21.4%。
- 最终模型纳入晚孕期平均餐后2小时血糖、胃容积Z值、胃排空率Z值、胃围/腹围比值。
- 训练、验证、测试队列的AUC分别为0.906、0.883、0.876。
- 训练与验证队列校准良好,测试队列校准可接受。
- 决策曲线和临床影响曲线分析支持模型的潜在临床效用。
临床应用与阅片要点
- 对GDM孕妇行晚孕期超声检查时,应规范测量胎儿胃容积、胃排空率及胃围/腹围比值。
- 产前超声报告提示胃功能参数异常时,应结合母体餐后血糖水平综合评估新生儿喂养障碍风险。
- 该列线图可辅助产科与新生儿科提前制定喂养管理方案。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: To develop and validate a prenatal prediction model for early feeding dysfunction in neonates born to women with gestational diabetes mellitus (GDM) using late-pregnancy maternal glycemic indices and fetal gastric ultrasound parameters.
MATERIALS AND METHODS: This multicenter prospective cohort study enrolled singleton pregnancies complicated by GDM from December 2019 to December 2025. Cases from one center were randomly divided into training and validation cohorts. Cases from three other centers formed the external test cohort. The primary outcome was early feeding dysfunction within 72 h after birth. Candidate predictors included maternal characteristics, late-pregnancy glycemic indices, routine fetal biometry, and fetal gastric ultrasound parameters measured at 32+0 to 36+6 weeks. A multivariable logistic regression model was developed and presented as a nomogram. Model performance was assessed by discrimination, calibration, and clinical utility in the training, validation, and test cohorts.
RESULTS: Early feeding dysfunction occurred in 21.3% and 21.4% of neonates in the training and validation cohorts, respectively. The final model included mean 2-hour postprandial glucose in late pregnancy, gastric volume Z score, gastric emptying rate (GER) Z score, and gastric circumference-to-abdominal circumference ratio (GC/AC). The nomogram showed good discrimination, with area under the curve (AUC) values of 0.906, 0.883, and 0.876 in the training, validation, and test cohorts, respectively. Calibration was good in the training and validation cohorts and acceptable in the test cohort. Decision curve and clinical impact curve analyses supported potential clinical utility.
CONCLUSION: This prenatal model integrating late-pregnancy maternal glycemic indices and fetal gastric ultrasound parameters showed good performance for predicting early feeding dysfunction in neonates born to women with GDM.
引用
- Lian, Lin, Bai, et al. Prediction of Early Feeding Dysfunction in Neonates of Women with Gestational Diabetes Mellitus Using Late-pregnancy Maternal Glycemic Indices and Fetal Gastric Ultrasound Parameters: A Multicenter Prospective Cohort Study. Acad Radiol. 2026. PMID: 42469042. doi:10.1016/j.acra.2026.06.049
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。