Table 4.

The performance of predictive models in validation set

Models	Variables	AUROC	Sensitivity	Specificity
Logistic regression	Basic model*	0.789	0.779	0.665
	+ BMI	0.805	0.779	0.679
	+ CCI	0.807	0.710	0.756
	+ PRS	0.792	0.776	0.679
	All except PRS	0.818	0.737	0.747
	Full Model	0.821	0.768	0.722

*Basic model were constructed with age, gender, and income as predictors