Validation of fatty liver disease scoring systems for ultrasound diagnosed non-alcoholic fatty liver disease in adolescents

Background and aims: The incidence of non-alcoholic fatty liver disease (NAFLD) is increasing in young populations. However, there are inadequate data regarding diagnosis of NAFLD. We aimed to validate three scoring systems against a previous standard of suprailiac skinfold thickness for diagnosing NAFLD in population-based adolescents.

Methods: Seventeen-year-old adolescents (n = 899), participating in the Raine Study, attended a cross-sectional follow-up. NAFLD was diagnosed using liver ultrasound. Scores for Fatty liver index (FLI), Hepatic Steatosis Index (HSI) and Zhejiang University index (ZJU index) were calculated. Diagnostic accuracy of these diagnostic tests was evaluated through discrimination and calibration.

Results: NAFLD was diagnosed 9% in males and 15% in females. The three scoring systems demonstrated better discrimination performance for NAFLD in males (AUC was FLI:0.82, HSI: 0.83 and ZJU index: 0.83) compared to females (AUC was FLI: 0.67, HSI: 0.67 and ZJU index: 0.67). Suprailiac skinfold performed better than the scoring systems (overall AUC: 0.82; male AUC:0.88; female AUC:0.73). FLI had best calibration performance.

Conclusion: Suprailiac skinfold thickness was a better predictor of ultrasound-diagnosed NAFLD than the three diagnostic scoring systems investigated. The higher performance characteristics of the algorithmic scoring systems in males compared with females may have implications for use in population assessments.