Table 2.

Performance assessment of the noninvasive indices and models for the prediction of MAFLD in the NHANES cohort

Variables	AUC (95% CI)	SEN (95% CI)	SPE (95% CI)	PPV (95% CI)	NPV (95% CI)	Cutoff value
METS-IR	0.846 (0.837–0.854)	0.818 (0.805–0.830)	0.717 (0.703–0.732)	0.748 (0.735–0.761)	0.793 (0.779–0.807)	40.887
TyG	0.746 (0.735–0.757)	0.727 (0.713–0.741)	0.640 (0.625–0.656)	0.675 (0.661–0.689)	0.695 (0.680–0.711)	8.497
TyG-BMI	0.850 (0.841–0.859)	0.819 (0.807–0.831)	0.727 (0.712–0.741)	0.755 (0.741–0.768)	0.796 (0.782–0.810)	240.222
TyG-WC	0.863 (0.855–0.871)	0.859 (0.848–0.870)	0.702 (0.687–0.717)	0.748 (0.735–0.760)	0.829 (0.816–0.842)	822.332
TyG-WtHR	0.853 (0.845–0.862)	0.810 (0.798–0.823)	0.740 (0.726–0.754)	0.762 (0.749–0.775)	0.791 (0.778–0.805)	5.066
HSI	0.835 (0.826–0.844)	0.799 (0.787–0.812)	0.722 (0.708–0.737)	0.747 (0.734–0.761)	0.778 (0.764–0.792)	37.131
VAI	0.741 (0.730–0.752)	0.711 (0.697–0.726)	0.658 (0.643–0.673)	0.681 (0.667–0.696)	0.689 (0.674–0.705)	1.494
FLI	0.859 (0.851–0.867)	0.839 (0.828–0.851)	0.714 (0.700-0.729)	0.751 (0.738–0.764)	0.812 (0.799–0.826)	49.405
LAP	0.834 (0.825–0.843)	0.787 (0.773-0.800)	0.725 (0.711–0.740)	0.746 (0.733–0.760)	0.768 (0.754–0.782)	45.180
ZJU	0.845 (0.836–0.854)	0.826 (0.814–0.838)	0.711 (0.696–0.726)	0.746 (0.733–0.759)	0.799 (0.786–0.813)	38.522
FSI	0.858 (0.850–0.867)	0.768 (0.755–0.782)	0.775 (0.761–0.789)	0.778 (0.765–0.792)	0.765 (0.751–0.779)	-1.068
K-NAFLD	0.838 (0.829–0.847)	0.771 (0.757–0.784)	0.748 (0.734–0.762)	0.759 (0.745–0.772)	0.760 (0.746–0.774)	-1.023

Abbreviations: AUC Area under the receiver operating characteristic curve, SPE Specificity, SEN Sensitivity, NPV Negative predictive value, PPV Positive predictive value