Key points are not available for this paper at this time.
OBJECTIVE: To systematically consolidate the most consistently applicable risk factors and assess their predictive performance for gestational diabetes mellitus (GDM) prediction. DESIGN: Systematic review and meta-analysis. DATA SOURCES: A comprehensive search included several databases (PubMed, Web of Science, Scopus, EMBASE via OVID and CINAHL) from inception to 30 October 2024. REVIEW: Methods prediction studies conducted in pregnant women during the first and second trimesters were included. Predictive parameters, including true positive, false positive, false negative and true negative values for each factor, were extracted from the 2×2 table. The Prediction model Risk Of Bias ASsessment Tool (PROBAST) was used to evaluate the risk of bias and the applicability of the included studies. A predictive meta-analysis was performed using Stata V.17. Variability was assessed through subgroup analysis. RESULTS: Seventy-seven eligible studies involving a total of 477 673 participants were included. Among these, 52 had an overall low risk of bias, and 65 demonstrated low concern regarding applicability. Common risk factors used to predict GDM included maternal age, body mass index (BMI), ethnicity, family history of diabetes mellitus (DM), previous GDM, previous macrosomia, multiparity, hypertensive disorders of pregnancy (HDP) and polycystic ovarian syndrome (PCOS). Among all factors, previous macrosomia showed the highest discriminatory power (area under the curve (AUC)=0.77), followed by previous GDM (AUC=0.73), PCOS (AUC=0.66), BMI (AUC=0.65), family history of DM (AUC=0.64), HDP (AUC=0.64), age over 30 (AUC=0.60), Asian ethnicity (AUC=0.59) and multiparity (AUC=0.56). A combination of age, BMI, family history of DM, previous GDM, ethnicity and multiparity achieved the highest AUC of 0.88, followed by the combination of age, BMI, family history of DM and previous GDM, with an AUC of 0.75. CONCLUSIONS: Previous macrosomia and previous GDM are clinically essential risk factors with the highest discriminatory power, indicated by an AUC >0.7. Combining risk factors improves predictive performance, reaching an AUC of 0.88. Our findings highlight the need to enhance existing risk-based screening tools for early detection of GDM and to prevent its adverse pregnancy outcomes and future cardiometabolic risks. PROSPERO REGISTRATION NUMBER: CRD 42024598399.
Ayele et al. (Tue,) studied this question.