Innovative nomogram integrating bile acid metabolomics for early diagnosis of intrahepatic cholestasis of pregnancy.

BACKGROUND

Intrahepatic cholestasis of pregnancy (ICP) is a severe pregnancy-specific liver disorder that increases the risk of adverse perinatal outcomes, including preterm birth and stillbirth. Early diagnosis of mild ICP remains challenging, as traditional biomarkers such as total bile acids (TBA) offer limited sensitivity and specificity during the early disease phase. Integrating metabolomic bile acid profiling with clinical indicators may enhance early diagnostic accuracy.

OBJECTIVE

To develop and validate an innovative nomogram-based predictive model that integrates bile acid metabolomics and clinical indicators for early diagnosis of mild ICP.

METHODS

This prospective cohort study included 128 pregnant individuals diagnosed with ICP and matched healthy controls recruited at Zhejiang Provincial People's Hospital (January 2019-June 2024). Bile acid profiles were quantified using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Clinical and biochemical variables were collected. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used to select candidate variables, followed by multivariate logistic regression to construct a predictive model. A nomogram was generated. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) with an internal training/validation cohort split (85:15).

RESULTS

Ten independent predictors were identified and incorporated into the nomogram, including TBA subtypes and liver function markers. The model demonstrated excellent discrimination, with an AUC of 0.920 (95% CI: 0.873-0.959) in the training cohort and 0.940 (95% CI: 0.847-1.000) in the validation cohort, outperforming TBA alone (AUC = 0.832). Calibration curves confirmed strong agreement between predicted and observed outcomes. DCA indicated superior clinical utility at low-to-moderate risk thresholds (0.0-0.8). The model enables individualized risk-prediction and supports early clinical intervention for mild ICP.

CONCLUSIONS

This novel nomogram-based model integrating bile acid metabolomics and clinical indicators significantly improves the early diagnostic accuracy of mild ICP over conventional methods. The approach offers high translational potential for clinical decision-making and maternal-fetal risk management. Future multicenter validation is warranted to confirm generalizability and clinical applicability.