Table 3.

AUCs of discrimination models calculated using bootstrap validation and 10-fold cross-validation in the replication population.

	Bootstrap (n = 1000)			Cross-validation (10-fold)
	AUC	Std error	95% CI	AUC	Std error	95% CI
Model 1	0.742	0.023	0.683-0.828	0.744	0.024	0.690-0.811
Model 2	0.902	0.018	0.881-0.943	0.916	0.018	0.883-0.949
Model 3	0.918	0.015	0.890-0.949	0.923	0.014	0.900-0.954

Model 1: clinical traits (age, sex, BMI, HBP, HLDL-cholesterol, nonHDL-cholesterol) included; model 2: glycan score included; model 3: clinical traits and glycan score included.