MAFLD outperforms NAFLD in identifying metabolic dysfunction in U.S. adolescents: a NHANES 2017-2020 analysis

The newly proposed metabolic-associated fatty liver disease (MAFLD) definition better reflects the metabolic pathogenesis shared with non-alcoholic fatty liver disease (NAFLD), though with incomplete diagnostic overlap; this study systematically compares their prevalence and risk profiles in U.S. adolescents to inform optimal clinical management strategies.

This cross-sectional study analyzed participants with complete survey data from the 2017-2020 cycle of the National Health and Nutrition Examination Survey (NHANES), identifying and comparing 532 MAFLD cases and 527 NAFLD cases through comprehensive prevalence assessment, detailed clinical characterization, regression for variable selection, and multivariate logistic regression analysis of independent risk factor associations.

MAFLD prevalence was 22.8% (95%CI:18.8-26.8) vs. NAFLD's 25.8% (21.5-30.0). MAFLD cases demonstrated significantly greater metabolic severity, evidenced by elevated BMI (31.49 ± 6.63 vs 21.70 ± 4.18 kg/m²), worsened insulin resistance (HOMA-IR 5.44 ± 3.59 vs 2.48 ± 1.44), and more pronounced hepatic injury markers (ALT 21.85 ± 13.04 vs 14.36 ± 7.40 U/L; all p < 0.001). MAFLD and NAFLD prevalence showed strong ethnic disparities, peaking in Mexican Americans (40.6% vs 42.9%) and obese females (79.0% vs 78.1%). LASSO regression selection identified obesity as the primary shared risk factor (MAFLD coefficient = 1.203; NAFLD = 0.844), while revealing MAFLD's more complex metabolic signature through retention of additional variables (8 vs 6 in NAFLD model). Waist circumference showed consistent associations across both classifications (MAFLD OR = 1.10, 95%CI:1.02-1.19; NAFLD OR = 1.09, 95%CI:1.03-1.15), though triglyceride levels only approached significance in NAFLD (OR = 1.60, 95%CI:0.93-2.75, p = 0.086). Multivariate logistic regression analyses confirmed stronger metabolic abnormalities associations in MAFLD (OR = 3.46, 95%CI:1.41-8.53, p = 0.011) versus NAFLD (OR = 1.45, 95%CI:0.63-3.32, p = 0.364), with similar patterns for obesity (MAFLD OR = 5.71, 95%CI:1.08-30.35; NAFLD OR = 2.24, 95%CI:0.62-8.13).

This study demonstrates that while MAFLD and NAFLD share obesity as a core risk factor, MAFLD more specifically identifies adolescents with severe metabolic dysfunction and provides superior risk stratification. The stronger metabolic dysfunction associations and greater model complexity support MAFLD's clinical utility for early intervention in high-risk youth populations.