Estudio de asociación de todo el exposoma de hiperuricemia a través de un estudio de cohorte rural y para predecir...

This study aims to systematically evaluate the associations between various exposure factors and hyperuricemia. A total of 22,765 participants derived from the Henan rural cohort, and 9 categories containing 60 different exposures. The exposome-wide association study (ExWAS) approach was used to estimate the associations between various exposure factors and hyperuricemia. An adaptive elastic net (AENET) model was implemented to select significant exposure factors, followed by the application of a gradient boosting machine (GBM) model to establish the prediction model of these variables to hyperuricemia. The importance of the indicators was assessed through Shapley Additive Explanations (SHAP). Additionally, the weighted quantile sum (WQS) method was employed to validate the selection of these variables. In ExWAS analysis, 40 exposures were significantly associated with the risk of hyperuricemia. The AENET model selected ten exposures as predictors. GBM model and SHAP result showed that the top three exposures were Creatinine, Triglycerides, and PM_2.5, which interpreted the model as 0.460, 0.331, and 0.314, respectively. Furthermore, the area under characteristics (AUC) of the model was 0.815 (95% CI: 0.802-0.820). WQS shows the same ranking results. The systematic evaluation of this study provides new insights into the complex environment-related factors of hyperuricemia.