Work

Gestational Diabetes Mellitus and Birth Weight: A Bidirectional Mendelian Randomization and Cohort Study in China

Year: 2026

Type: article

Abstract: Background: Gestational diabetes mellitus (GDM) is regarded as a driver of fetal macrosomia, yet observational estimates of its effect on birth weight (BW) are vulnerable to confounding. Whether GDM causally shapes fetal growth, and through which shared biology, remains unresolved. Methods: We performed a bidirectional two-sample Mendelian randomization (MR) using FinnGen R11 GDM (16,802 cases) and the EGG Consortium birth-weight GWAS (n=286,868), with inverse-variance weighted (IVW, primary), MR-Egger, weighted median, weighted mode, RAPS estimators, MR-PRESSO and Bayesian colocalization. Multi-omics evidence (GSE92772, ST001948, INTERVAL cis-pQTLs) was integrated as hypothesis-generating evidence. In a cohort of 24,448 singleton pregnancies, linear and logistic regression, bootstrap mediation, and double/debiased machine learning (DML) were performed for verification. Results: The primary IVW estimate for GDM to BW was null (β=-0.004, P=0.889) under extreme heterogeneity, and robust estimators showed nominally positive signals. Seven of 21 instruments flagged as outliers overlapping six colocalizing loci. Therefore, it is impossible to precluding a reliable causal interpretation. Reverse MR (BW to GDM) was exploratory (five instruments, all outliers). Colocalization identified six loci ( CCND2 , MTNR1B , GPSM1 , ADCY5 , CDKAL1 , TCF7L2 ) with evidence of a shared causal variant (PP.H4≥ 0.5). All established islet beta-cell/insulin-secretion genes suggested shared upstream biology. In the cohort, GDM was associated with higher rates of macrosomia (OR 1.21) and lower birth weight (LBW, OR 1.45). The association of macrosomia was largely attenuated after adjustment for maternal BMI and lipids (OR 1.01, P=0.913). Using bidirectional mediation analysis, five candidate mediators (BMI, TG, HDL-C, LDL-C, and TG/HDL-C ratio) were identified as significant mediators in both directions. Moreover, hypertensive disorders showed the largest associations with LBW (PE β=-0.204; GH β=-0.067) in DML analyses. Conclusions: Evidence for a consistent causal effect of GDM on birth weight remains inconclusive. The observational association of macrosomia was largely explained by maternal metabolic status. The shared insulin-secretory biology more plausibly reflects genetic pleiotropy than a direct causal pathway. Hypertensive disorders showed stronger associations with birth weight reduction than GDM, supporting integrated management of glycemia, BMI, lipids and blood pressure. Keywords: gestational diabetes mellitus, birth weight, macrosomia, low birth weight, mendelian randomization, multi-omics, cohort study, causal inference

Source: Diabetes Metabolic Syndrome and Obesity

DOI: 10.2147/dmso.s653174

Authors: Xia Li, Yue Peng, Yan Wang, Bin Yu

Institution: Nanjing Medical University

Language: English

Topic: Gestational Diabetes Research and Management

Subfield: Obstetrics and Gynecology

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $2,836