Table 3. Summary risk estimates of the association between coffee consumption and risk of liver cancer.

Studies groups	References number	No. of cases/No. of total participants	RR (95% CI)	Heterogeneity
				I2(%)	P
All studies	3,24–32	3,389/2,248,500	0.55(0.44–0.67)	38.0	0.081
Study gender
Male	3,24–27,30	1,547/656,146	0.58(0.40–0.83)	59.5	0.022
Female	3,24–27	604/879,632	0.57(0.42–0.79)	0.0	0.588
Study region
Asia	24–26,28,29	1,295/299,426	0.50(0.38–0.66)	0.0	0.763
Europe	27,30,31	523/574,159	0.37(0.25–0.54)	22.6	0.275
North America	3,32	1,571/1,374,915	0.75(0.59–0.95)	0.0	0.543
Adjustment for tea
Yes	24,29,30,32	1,341/342,768	0.53(0.39–0.72)	0.0	0.557
no	3,25–28,31	2,041/1,905,732	0.55(0.41–0.74)	53.0	0.030
Adjustment for history of liver diseases
Yes	25–28	727/204,060	0.48(0.36–0.63)	0.0	0.665
No	3,24, 29–32	2,662/2,044,440	0.58(0.43–0.78)	53.5	0.035
Adjustment for history of diabetes
Yes	24,27–32	1,894/890,534	0.43(0.33–0.56)	9.9	0.353
No	3,25,26	1,495/1,357,966	0.69(0.56–0.84)	0.0	0.453
Adjustment for BMI
Yes	3,27–32	2,680/2,224,975	0.56(0.42–0.74)	51.5	0.036
No	24–26	709/235,525	0.50(0.36–0.70)	0.0	0.527
Adjustment for education
Yes	24,27,29–32	1,670/889,899	0.43(0.32–0.58)	24.6	0.249
No	3,25,26,28	1,719/1,358,601	0.67(0.54–0.82)	0.0	0.430

No., Number.