External validation of steatosis index in health check-up population

Abstract

Objectives: Recognizing the presence of fatty liver is the first step in screening high-risk group of steatohepatitis in the general population. Several hepatic steatosis formulae are available with limited validation. We tried to compare the diagnostic performance of prediction models of hepatic steatosis in average risk group using magnetic resonance imaging (MRI). Methods: In total, 1,301 health check-up subjects who underwent hepatic fat quantification using MRI method enrolled. Anthropometric, clinical, and laboratory data were examined and fatty liver diagnosed by MRI. We assessed diagnostic performance by using area under the receiver operating characteristic curve (AUROC). Results: Non-alcoholic fatty liver disease (NALFD) liver fat score showed the highest diagnostic performance with an AUROC of 0.72, followed by Framingham steatosis index (0.70), hepatic steatosis index (HSI, 0.69), ZJU index (0.69), and fatty liver index (FLI, 0.68). There were considerable intermediate zone in three fatty liver prediction models using two cut-offs (FLI (28.9%), HSI (48.9%), and ZJU index (53.6%)). The best cut-off value of NAFLD liver fat score for predicting fatty liver was -1.33. New cut-off point showed a sensitivity of 63.5%, specificity of 73.6%, positive predictive value of 51.0%, and negative predictive value of 82.3%. Diagnostic performance of NAFLD liver fat score to detect steatosis was comparable with ultrasonography. The diagnostic agreement between NAFLD liver fat score and MRI was 72.7%, and 70.9% between ultrasound and MRI for fatty liver. Conclusions: Diagnostic performance of all five estimation formulae for hepatic steatosis was somewhat low, suggesting modest efficacy to detect steatosis in community setting.