Risk factors for drug-related gestational diabetes mellitus: a real-world pharmacovigilance study based on the FAERS database.

backgroundGestational diabetes mellitus (GDM) is a common pregnancy complication linked to adverse outcomes for both mother and infant. Research on drug-related GDM remains limited.

objectiveThis study aims to explore the risk factors and potential mechanisms of drug-related GDM using the FDA Adverse Event Reporting System (FAERS).

methodsWe analyzed FAERS data from the past, focusing on drugs potentially associated with GDM. Significant drug signals were identified using statistical reporting odds ratio (ROR), proportional reporting ratio (PRR), empirical Bayesian geometric mean (EBGM), bayesian confidence propagation neural network (BCPNN), and univariate logistic regression analysis. We used multivariate logistic regression to analyze independent risk factors, built a multivariate logistic regression model to predict GDM, and used the generalized variance inflation factor (GVIF) to evaluate the multicollinearity of the predictive factors in the model. Furthermore, we performed gene target prediction, functional enrichment analysis, and protein-protein interaction (PPI) analysis to explore the mechanisms involved, and assessed drug-genes interactions through molecular docking simulations.

resultsBy analyzing 1137 cases of potentially drug-related GDM, we identified 12 drugs as potential independent risk factors for GDM. We constructed a multivariate logistic regression model incorporating drug and patient characteristics to predict GDM, with an area under the curve (AUC) of 0.847. GVIF analysis confirmed that there was no multicollinearity among the predictors. Potential target genes of antipsychotic drugs are mainly enriched in cellular response to dopamine, phospholipase C-activating G protein-coupled receptor signaling pathway, and blood circulation. After integrating the target gene group and the insulin receptor signaling pathway gene set, the potential target genes have extensive connections with the insulin receptor signaling pathway, with DRD2 and KCNH2 being hub genes in the PPI network.

conclusionThis study identified 12 drugs that may be independent risk factors for GDM, including quetiapine, aripiprazole, olanzapine, prednisolone, venlafaxine, risperidone, escitalopram, clozapine, mirtazapine, ziprasidone, rosuvastatin, and trazodone. The established predictive model has certain clinical value, and the potential mechanisms by which antipsychotic drugs may lead to GDM were further analyzed. These results provide potential clues for the early identification of GDM and early warning of suspected drugs, and offer a reference for further exploration of its pathogenesis.