Table 3. Main characteristics of studies included in the meta-analysis.

First author (year), ref.	Country	Ethnicity	Study design	Genotyping method	Sample size	Genotype (control)			Genotype (case)			MAF	HWE for control		Power (%)
					Control/case	ArgArg	ArgPro	ProPro	ArgArg	ArgPro	ProPro	Control/case	χ2	P-value
Present study (2014)	India	Asian	HB	AS-PCR	70/70	34	20	16	14	33	23	0.37/0.56	10.54	0.001	23
Zhang (2014)³⁰	China	Asian	PB	PCR-RFLP	477/566	130	229	118	133	292	141	0.48/0.50	0.73	0.39	97
Li (2013)²⁸	China	Asian	PB	PCR-RFLP	360/217	125	186	49	73	113	31	0.39/0.40	2.39	0.12	80
Xiao (2010)²⁹	China	Asian	PB	PCR-RFLP	712/522	226	366	120	117	270	135	0.42/0.51	1.88	0.17	99
Hadhri-Guiga (2007)²⁷	Tunisia	Caucasian	PB	PCR-RFLP	83/115	32	45	6	44	48	23	0.34/0.40	3.39	0.06	32
Sousa (2006)²⁶	Portugal	Caucasian	PB	AS-PCR	285/107	178	93	14	62	32	13	0.21/0.27	0.16	0.68	61
Tiwawech (2003)²⁵	Thailand	Asian	PB	PCR-RFLP	148/102	50	70	28	24	52	26	0.42/0.50	0.15	0.69	40
Tsai (2002)²⁴	China	Asian	HB	PCR-RFLP	59/50	25	26	8	20	14	16	0.35/0.46	0.08	0.76	18
Yung (1997)²³	China	Asian	PB	PCR-RFLP	31/20	10	13	8	6	11	3	0.46/0.42	0.77	0.39	10
Brigander (1997)³²	China	Asian	PB	PCR-RFLP	105/73	31	49	25	16	31	26	0.47/0.56	0.42	0.51	29

Abbreviations: AS, allele specific; HB, hospital based; HWE, Hardy–Weinberg equilibrium; MAF, minor allele frequency; PB, population based; RFLP, restriction fragment length polymorphism. The study design based on samples collected from hospitals or random populations, different countries and ethnicities, power of the study, genotyping method and the distribution of the genotype among NPC and controls were listed. HWE was tested using the web-based tools (http://www.oege.org/software/ we-mr-calc.shtml). Power analysis was performed by G power software (version 3.1).⁵²