Nomogram for predicting 5-year metabolic dysfunction-associated steatotic liver disease risk: retrospective cohort study

2024

Endocrine Connections, Vol 13, Iss 8, Pp 1-10 (2024)

Objective: To create a nomogram-based model to estimate the Chinese population's 5-year risk of metabolic dysfunction-associated steatotic liver disease (MASLD). Methods: We randomly divided 7582 participants into two groups in a 7:3 ratio: one group was assigned to work with the training set, which consisted of 5307 cases, and the other group was assigned to validate the model using 2275 cases. The least absolute shrinkage and selection operator model was employed to ascertain the variables with the highest correlation among all potential variables. A logistic model was constructed by incorporating these selected variables, which were subsequently visualized using a nomogram. The discriminatory ability, calibration, and clinical utility of the model were assessed using the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). Results: During the 5-year follow-up, 1034 (13.64%) total participants were newly diagnosed with MASLD. Using eight variables (gender, body mass index, waist, hemoglobin, alanine aminotransferase, uric acid, triglycerides, and high-density lipoprotein), we built a 5-year MASLD risk prediction model. The nomogram showed an area under the ROC of 0.795 (95% CI: 0.779–0.811) in the training set and 0.785 (95% CI: 0.760–0.810) in the validation set. The calibration curves revealed a 5-year period of agreement between the observed and predicted MASLD risks. DCA curves illustrated the practicality of this nomogram over threshold probability profiles ranging from 5% to 50%. Conclusion: We created and tested a nomogram to forecast the risk of MASLD prevalence over the next 5 years.

article

English

Bioscientifica, 2024.

5-year risk, metabolic dysfunction-associated steatotic liver disease, nomogram, predictive model, retrospective cohort study, Diseases of the endocrine glands. Clinical endocrinology, RC648-665