Development and validation of a new nomogram to screen for MAFLD

Area of Science:

• Hepatology
• Metabolic Diseases
• Predictive Modeling

Background:

• Metabolic dysfunction-associated fatty liver disease (MAFLD) presents a global health and economic challenge.
• Early identification and management of MAFLD patients are crucial for mitigating disease burden.
• Existing predictive models often lack clinical utility due to complex variable requirements.

Purpose of the Study:

• To develop a novel predictive model for MAFLD.
• To enhance clinical utility by utilizing readily available simple clinical and laboratory measures.
• To compare the performance of the new model against existing MAFLD predictive tools.

Main Methods:

• A retrospective cross-sectional study utilizing NHANES 2017-2020.3 data from 7300 participants.
• Least Absolute Shrinkage and Selection Operator (LASSO) regression to identify significant MAFLD predictors.
• Development and validation of the MAFLD Prediction Nomogram (MPN) and comparison with six existing models using AUC, NRI, IDI, calibration, and DCA.

Main Results:

• Nine key predictors identified: age, race, arm circumference, waist circumference, BMI, ALT/AST ratio, TyG index, hypertension, and diabetes.
• The MPN demonstrated significantly superior diagnostic accuracy compared to six existing models in both training (AUC 0.868) and validation (AUC 0.863) cohorts.
• The MPN exhibited a higher net benefit, indicating improved clinical utility.

Conclusions:

• The developed nonimaging-assisted nomogram effectively predicts MAFLD using demographics, laboratory factors, anthropometrics, and comorbidities.
• The MAFLD Prediction Nomogram (MPN) outperforms six existing models in predicting MAFLD.
• This model facilitates rapid assessment of MAFLD risk in the general population.