A clinical and laboratory-based nomogram for predicting nonalcoholic fatty liver disease in non-diabetic adults: a cross-sectional study

• Published in: Annals of palliative medicine Vol. 11; no. 7; pp. 2349 - 2359
• Main Authors: Qian, Yiting, Sun, Bo, Zhang, Yu, Zhang, Minbo, Jiao, Xiaoxiao, Lai, Linying, Yang, Wenzhuo
• Format: Journal Article
• Language: English
• Published: China 01.07.2022
• Subjects: prediction model; Nonalcoholic fatty liver disease (NAFLD); nomogram; risk prediction
• ISSN: 2224-5820; 2224-5839
• DOI: 10.21037/apm-21-2988

Although the close relationship between nonalcoholic fatty liver disease (NAFLD) and insulin resistance has been clarified and there is a five-fold higher prevalence of NAFLD in patients with diabetes compared to that in patients without diabetes, this is not a reason to focus only on the incidence of NAFLD in people with diabetes because people who are insulin resistant are not necessarily diagnosed with diabetes, which leads to the overlook of NAFLD in non-diabetic population. Actually, we are obligated to pay more attention to the non-diabetic population for early detection and intervention of NAFLD. There is a lack of a convenient tool for predicting NAFLD in non-diabetic adults, and thus we aim to develop and validate a novel clinical nomogram to predict NAFLD among non-diabetic population to save more medical resources and make less missed diagnosis. Researchers initially enrolled 20,944 patients and excluded those with history of drinking, known medication usage, viral hepatitis, known liver disease, missing covariant data, age <18 years, and impaired fasting blood glucose, leaving 14,251 adults participating in the baseline analysis, who were randomly divided in a ratio of 3:1 into a training dataset with 10,689 participants and a validation dataset with 3,562 participants, using the classification and regression training (caret) package in R software v. 4.0.3. Variables for prediction were selected by multivariable logistic regression analysis, the LASSO method, and clinical experience. Based on these, we constructed a prediction model. Performance of this model was validated by the area under the receiver operator characteristic curve, calibration curve, and decision curve analysis. We used 6 variables to construct the prediction model: body mass index (BMI), aspartate aminotransferase (AST), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), hemoglobin A1c (HbA1c), and diastolic blood pressure (DBP). In the training and validation datasets, the AUROC value of this prediction was 0.891 [95% confidence interval (CI): 0.884 to 0.899] and 0.902 (95% CI: 0.890 to 0.914), respectively. The calibration plots and the decision curve analysis (DCA) demonstrated that the accuracy of this model was good, with high clinical practicability. The nomogram could screen non-diabetic adults for NAFLD and may aid clinical decisionmaking.