Table 3.

The statistical parameters of stepwise logistic regression analysis model.

Step	BG-EPVS				WM-EPVS
	Effect(s) entry	P for effect selection tests	P for final model fitting	Goodness of fit (PPearson)	Effect(s) entry	P for effect selection tests	P for final model fitting	Goodness of fit (PPearson)
Model 1
1	age	<0.001	<0.001	0.638	UA	0.003	0.001	0.441
2	UA	<0.001			alcohol	0.046
3	HBP	0.008			—	—
Model 2
1	age	<0.001	<0.001	0.525	UA	0.003	0.001	0.493
2	HDL	<0.001			BUN	0.033
3	UA	0.001			—	—
4	HBP	0.009			—	—
5	BUN	0.010			—	—
Model 3
1	Fazekas	<0.001	<0.001	0.436	Fazekas	<0.001	<0.001	0.447
2	age	<0.001			UA	0.002
3	HDL	<0.001			BUN	0.047
4	UA	0.004			—	—
5	HBP	0.011			—	—
6	BUN	0.009			—	—
7	HbA1c	0.040			—	—

The consequential significance level for entered and removal variables was 0.05 and 0.1, respectively.

Model1: Age, sex, smoking, alcohol, hypertension, diabetes, CAD and BMI were added.

Model2: model 1 + HDL, LDL, TC, TG, HbA1c, Glu, BUN were added.

Model3: model 2 + Fazekas scale score were added.