Construction of a Risk Identification Model Based on Electrolytes and Related Factors for Nausea and Vomiting in Pregnancy: A Secondary Analysis of Clinical Trial Data

Background: Nausea and vomiting in pregnancy (NVP) is one of the most common symptoms during early gestation. A subset of affected women may progress to severe NVP, which is frequently accompanied by electrolyte disturbances, impaired nutritional status, and adverse pregnancy outcomes. However, the relationships between multiple electrolyte-related factors and the severity of NVP have not been systematically investigated. Objective: To explore the statistical associations between multiple electrolytes, related laboratory indicators, and NVP severity, and to develop a risk identification model integrating electrolyte, nutritional, and psychological factors for the risk stratification of severe NVP. Methods: This study was a secondary analysis of data derived from a multicenter randomized controlled trial and included 351 women with moderate-to-severe NVP during early pregnancy. Baseline demographic characteristics, biochemical parameters (including electrolytes, vitamins, and liver and renal function indicators), and psychological assessment scores from the Self-Rating Anxiety Scale (SAS) and the Self-Rating Depression Scale (SDS) were considered candidate variables. Variable selection was performed using least absolute shrinkage and selection operator (LASSO) regression, followed by multivariable logistic regression to construct two risk identification models: an electrolyte-based model (Model 1) and a multidimensional model integrating psychological and nutritional factors (Model 2). Model performance was evaluated using the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). Results: LASSO regression identified nine variables for inclusion in the final analysis, including Calcium (Ca), Potassium (K), Iron (Fe), Chloride (Cl), Zinc (Zn), Urea, SAS, SDS, and Pyridoxal (PL). Compared with Model 1, Model 2 demonstrated superior discriminatory performance (area under the ROC curve [AUC]: 0.717 vs. 0.655). Both models showed acceptable calibration. DCA indicated that Model 2 provided a slightly greater net clinical benefit than Model 1 across threshold probabilities ranging from 0.10 to 0.40. Conclusion: Electrolytes and related laboratory indicators were statistically associated with the severity of NVP, and integrating psychological and nutritional factors further improved risk identification performance. Because this study was based on cross-sectional baseline data, the findings reflect associations rather than causal relationships. The proposed model may serve as a preliminary tool for the risk stratification of severe NVP; however, its clinical utility requires further validation in independent external cohorts.