Association of Atherogenic Lipid Indices with Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD): A Cross-Sectional Analysis of NHANES

Abstract

Metabolic dysfunction-associated steatotic liver disease (MASLD) is closely coupled to an atherogenic lipid profile, but the relative strength of association of lipid-derived indices has not been established in a nationally representative U.S. population under the 2023 nomenclature. We compared nine lipid measures — triglycerides (TG), HDL-C, LDL-C, total cholesterol, non-HDL-C, TG/HDL-C, LDL-C/HDL-C, TC/HDL-C and the atherogenic index of plasma (AIP) — in the fasting subsample of NHANES 2021–2023. A controlled attenuation parameter ≥ 248 dB/m defined steatosis, and MASLD required steatosis plus at least one cardiometabolic criterion. Analyses used survey-weighted logistic regression, weighted ROC analysis with a design-based bootstrap, calibration assessment and eleven sensitivity analyses. Of 2,453 adults, 1,428 (58.2%) met MASLD criteria. After full adjustment the derived ratios showed the largest effects (AIP OR = 6.64, 95% CI 3.38–13.04). On a per-SD scale, TG/HDL-C (1.71), TG (1.71) and AIP (1.70) were formally indistinguishable (p = 0.87). TG/HDL-C and AIP gave the highest discrimination (AUC = 0.724), exceeding TG alone by a small margin (ΔAUC = 0.013, p = 0.001). Associations persisted after excluding participants on lipid-lowering therapy, cases reclassifiable as MetALD or alcohol-related liver disease, and active viral hepatitis. TG/HDL-C is a low-cost, routinely available marker warranting prospective, externally validated evaluation; the design is cross-sectional and no threshold is recommended clinically.