A simple clinical model using waist circumference and cholesterol predicts liver fibrosis in MAFLD

Abstract

Metabolic dysfunction-associated fatty liver disease (MAFLD) involves chronic low-grade inflammation; however, the utility of inflammatory markers for fibrosis risk stratification remains unclear. We aimed to investigate the relationships between inflammatory cytokines, steatosis, and fibrosis and to develop a simple predictive model for liver stiffness. We enrolled 49 patients with MAFLD who underwent anthropometric measurements, laboratory testing (including Interleukin-6 (IL-6) and High-sensitivity of C-reactive protein (hs-CRP), and vibration-controlled transient elastography (FibroScan) for steatosis ( controlled attenuation parameter, CAP) and fibrosis (liver stiffness measurement, LSM). Correlations between inflammatory markers, metabolic parameters, and liver outcomes were analyzed. Logistic regression identified fibrosis predictors (LSM ≥ 6.0 kPa, indicating any degree of fibrosis), and predictive performance was compared with that of traditional indices. Internal validation was performed using bootstrap resampling with 5000 iterations (bias-corrected accelerated method). A sensitivity analysis further adjusted for age, sex, diabetes, and use of glucose-lowering and lipid-lowering medications.(exploratory, not bootstrapped). IL-6 levels were positively correlated with CAP ( R  = 0.390, P  = 0.007), triglycerides, and uric acid levels, and negatively correlated with high density lipoprotein (HDL) levels. hs-CRP showed strong correlations with LSM ( R  = 0.639, P  < 0.001), body mass index (BMI), and liver enzymes. Waist circumference (odds ratio [OR] = 1.21, P  = 0.004) and cholesterol levels (OR = 4.80, P  = 0.011) independently predicted early fibrosis in the parsimonious model. After adjusting for age, sex, diabetes, and medication use, waist circumference (OR = 1.226, 95% CI: 1.055–1.425, P  = 0.008) and cholesterol (OR = 4.677, 95% CI: 1.249–17.509, P  = 0.022) remained significant, with AUC 0.897, sensitivity 73.7%, specificity 86.2%, and accuracy 81.3%. Bootstrap internal validation confirmed model stability, with bootstrap-corrected ORs of 1.230 (95% CI: 1.070–1.499) for waist circumference and 5.366 (95% CI: 1.376–32.187) for cholesterol, with AUC 0.866, sensitivity 73.7%, specificity 89.7%, and accuracy 83.3%. The combined model yielded an area under the curve (AUC) of 0.862 (95% CI: 0.744–0.980), which compared favorably with APRI (0.722), FIB-4 (0.537), and AST/ALT ratio (0.340). Central obesity and lipid dysfunction are key determinants of MAFLD-related fibrosis. A simple model using waist circumference and cholesterol provides robust, internally validated risk stratification for early fibrosis, independent of common metabolic confounders. External validation in larger cohorts is required before clinical implementation.