A Multicenter Randomized Controlled Trial of a Digital Health Intervention for Risk Identification, Intelligent Early Warning, and Prevention of Gestational Diabetes Mellitus Based on Multi-Source Data Integration

Gestational diabetes mellitus (GDM) has become an increasingly important public health problem among pregnant women in China. GDM is associated with short-term adverse pregnancy outcomes and long-term health risks for both mothers and their offspring. Current routine screening for GDM is mainly based on oral glucose tolerance testing performed at 24 to 28 weeks of gestation, which may miss the optimal window for earlier risk identification and preventive intervention.

This study is designed to improve early identification and prevention of GDM through multi-source data integration and digital health intervention. The study is conducted in the context of maternal health management in China and includes a retrospective model development phase and a prospective randomized controlled trial phase.

In the model development phase, retrospective clinical data from maternal health information systems, including demographic characteristics, obstetric history, physical examination records, laboratory indicators, and pregnancy follow-up data, will be used to identify predictors of GDM and to construct an early risk prediction model. Statistical and machine learning methods, including random forest and other predictive modeling approaches, will be used to evaluate model performance and optimize discrimination, calibration, robustness, and interpretability.

In the intervention phase, pregnant women in early pregnancy who are identified as being at high risk for GDM will be recruited from participating maternal and child health hospitals and randomly assigned in a 1:1 ratio to either a digital health intervention group or a usual care group. Participants in the intervention group will receive additional support through a digital platform, including structured health education, individualized recommendations on diet, physical activity, gestational weight management, and sleep, as well as online consultation and self-management tools. Participants in the control group will receive routine antenatal care and standard health education materials.

Follow-up will be conducted during pregnancy and at delivery. The primary outcome is the incidence of GDM diagnosed according to routine clinical criteria during pregnancy. Secondary outcomes include gestational weight gain, cesarean delivery, macrosomia, and selected maternal and neonatal outcomes. Safety monitoring will focus on possible adverse events related to lifestyle intervention, such as hypoglycemic symptoms or other discomforts.