. 2003 Oct 15;9(10):2186–2193. doi: 10.3748/wjg.v9.i10.2186

Table 4.

Multiple regression models of predictors on levels of oxidative DNA damage biomarkers

Predictors in the model	Double staining percentage (%)		8-oxoGP		8-oxoGI		hOGG₁P		hOGG₁I
Predictors in the model	Beta^b	P	Beta	P	Beta	P	Beta	P	Beta	P
8-oxoGP							0.463		0.000^c
8-oxoGI	0.521	0.000^c							131.226	0.000^c
hOGG₁P			0.922	0.000^c
hOGG₁I	0.411	0.000^c			6.591E-02	0.008^c
Age	0.090	0.201	-0.139	0.108	-0.075	0.426	2.532E-02	0.001^c	3.069	0.018^c
Gender	-0.027	0.693	-0.081	0.336	-0.106	0.253	0.058	0.474	0.104	0.262
BMI	-0.046	0.496	-0.055	0.511	-1.048E-02	0.050^c	-0.053	0.516	0.037	0.696
hOGG₁ 326SNP	0.056	0.427	0.103	0.029^c	2.370E-03	0.000^c	-0.112	0.166	-15.684	0.006^c
Constant	0.480	0.000	6.969E-02	0.705	1.601	0.000	0.401	0.001	-133.512	0.001
R²(F)^d	0.667 (79.227)^e		0.470 (35.033)^e		0.377 (15.729)^e		0.509 (40.875)^e		0.386 (16.369) ^e

A: hOGG₁ 326SNP: Cys/Cys = 1; Cys/Ser = 2; Ser/Ser = 3; B: Beta: standardized coefficient; C: Variables entered final equation (backward); D: R square of model (F value of ANOVA); E: ANOVA P value < 0.01.