Validation of Fatty Liver Scoring Systems for the Diagnosis of NAFLD in Adolescents

Higher levels of alcohol and tobacco use put men at risk for liver disease, such as cirrhosis and alcoholic liver disease. Men who have sex with men are at increased risk for viral hepatitis B, and should be screened accordingly.
A team of investigators sought to identify the accuracy of diagnostic tools for the detection of nonalcoholic fatty liver disease in adolescent patients.

The diagnostic performance of fatty liver disease scoring systems among adolescents with nonalcoholic fatty liver disease (NAFLD) was evaluated in a study published in Digestive and Liver Disease. Scores for the Fatty Liver Index (FLI), Hepatic Steatosis Index (HSI), and Zhejiang University Index (ZJU index) were each positively correlated with an ultrasound-confirmed diagnosis of NAFLD. However, suprailiac skinfold thickness (SFT) was found to be a better predictor of NAFLD than any of the 3 scoring systems.

Investigators extracted data from patients in the Raine Study, a prospective cohort of children, adolescents, and adults in Western Australia. Briefly, the Raine Study recruited pregnant individuals between 1989 and 1991, and then continued follow-up on offspring over the life course. The present analysis used clinical data collected when participants were 17 years of age. The 17-year follow-up assessment included an abdominal ultrasound and a detailed questionnaire assessing nutrition and lifestyle. Venous blood samples were taken and analyzed for certain liver parameters. Using these clinical data, the investigators calculated scores for the FLI, HSI, and ZJU. The diagnoses made by these scoring systems were compared with diagnoses obtained via abdominal ultrasound. Area under the receiver operating curve (AUC) was used to assess the discriminative capacity of each scoring system compared with ultrasound.

The study cohort comprised 899 participants, among whom 413 were female. Overall, 11.57% patients had ultrasound-defined NAFLD at the 17-year follow-up visit. NAFLD was more prevalent among women compared with men (14.53% vs 9.05%). The AUC values for the HSI, ZJU, and FLI were 0.755, 0.758, and 0.710, respectively, indicating moderate to fair discrimination between patients with NAFLD and patients without NAFLD. All indices performed better with men compared with women (FLI AUC: 0.816 vs 0.674; HIS: 0.833 vs 0.668; ZJU: 0.829; 0.672). Suprailiac SFT, with an AUC of 0.816, outperformed all 3 indices, although was also found to be more effective in men compared with women (AUC, 0.875 vs 0.732). FLI had the best calibration performance.

In this study, suprailiac SFT was found to be a better predictor of ultrasound-diagnosed NAFLD compared with other fatty liver scoring systems. All metrics performed better among men compared with women. As a study limitation, the investigators noted that the use of ultrasound over biopsy may have led to an underdiagnosis of NAFLD.

“The use of the FLI as a continuous variable is a reasonable alternative for NAFLD diagnosis in epidemiological studies,” the investigators wrote. “However, the higher performance characteristics of the algorithmic scoring systems in males compared with females may have implications for use in population assessments.”


Wan F, Pan F, Ayonrinde OT, et al. Validation of fatty liver disease scoring systems for ultrasound diagnosed non-alcoholic fatty liver disease in adolescents. Published online December 14, 2020. Dig Liver Dis. doi:10.1016/j.dld.2020.11.037