Table 6. Pooled analysis of the combined effects of GSTT1, GSTM1, and GSTP1 genotypes and RCC risk^&.

Genetic model	Number ofstudies	Sample Size		Analysis	I2 (%)	P h	Test of Association
		Case	Control	Model			P	OR(95% CI)
GSTM1(−)/GSTT1(−)
Total	6	1052	1573	R	81	<0.05	0.84	1.06 [0.62, 1.81]
Geographic area
European	4	880	1187	F	37	0.19	0.02	0.79 [0.65, 0.96]
European*	2	723	1008	F	0	0.56	0.03	0.80 [0.66, 0.98]
Asian	1	111	136				<0.01	2.70 [1.61, 4.52]
USA	1	61	250				0.16	1.58 [0.83, 3.01]
Quality Score
High quality	4	895	1394	R	86	<0.05	0.50	1.22 [0.68, 2.18]
Low quality	2	157	179	R	77	0.04	0.43	0.36 [0.03, 4.56]
GSTM1(−)/GSTP1(−)	2	166	377	R	64	0.10	0.23	1.51 [0.78, 2.92]
GSTT1(−)/GSTP1(−)	2	172	354	R	61	0.11	<0.01	2.79 [1.44, 5.42]
GSTM1(−)/GSTT1(−)/GSTP1(−)	2	94	189	R	74	0.05	0.12	2.58 [0.78, 8.53]

Abbreviations: R, Random-effects model; F, Fixed-effects model; P_h, test for heterogeneity. ^& GSTM1 (−): the null genotype; GSTT1 (−): the null genotype; GSTP1 (−): AG or GG; European*: two studies [23], [32] were omitted because of low quality.