Coffee drinking and cancer risk: an umbrella review of meta-analyses of observational studies

Zhao, Long-Gang; Li, Zhuo-Ying; Feng, Guo-Shan; Ji, Xiao-Wei; Tan, Yu-Ting; Li, Hong-Lan; Gunter, Marc J.; Xiang, Yong-Bing

doi:10.1186/s12885-020-6561-9

BMC Cancer

Table 2 Evaluation of bias and heterogeneity in 36 associations of coffee intake with cancer incidence

From: Coffee drinking and cancer risk: an umbrella review of meta-analyses of observational studies

Author, year	Cancer type	Unit of comparison	I² (95%CI)	P for Q*	Egger’s P value	Observed^a	Expected^a	P for TES^b
Thomopoulos, 2015 [20]	Acute lymphocytic leukemia	H/L	8 (0, 73)	0.368	0.309	2	1.24	–
Thomopoulos, 2015 [20]	Acute myelogenous leukemia	H/L	63 (2, 86)	0.029	0.9	1	1.16	0.444
Godos, 2017 [4]	Biliary tract cancer	H/L	0 (0, 79)	0.849	0.073	0	0.42	–
Wu, 2015 [24]	Bladder cancer	H/L	42 (11, 63)	0.009	0.014	9	10.42	–
Jiang, 2013 [9]	Breast cancer	H/L	0 (0, 39)	0.485	0.374	4	5.64	–
Gan, 2017 [3]	Colon cancer	H/L	30 (0, 62)	0.124	0.699	3	3.14	–
Li, 2013 [13]	Colorectal cancer	H/L	57 (38, 70)	< 0.001	0.419	11	10.27	–
Lukic, 2018 [8]	Endometrial cancer	H/L	29 (0, 60)	0.132	0.326	8	3.17	0.786
Zheng, 2013 [28]	Esophageal cancer	H/L	42 (0, 69)	0.049	0.038	3	4.88	0.008
Botelho, 2006 [1]	Gastric cancer	H/L	34 (0, 61)	0.055	0.695	3	5.13	–
Malerba, 2013 [15]	Glioma	H/L	0 (0, 75)	0.479	0.135	0	0.85	–
Wijarnpreecha, 2017 [23]	Kidney cancer	H/L	35 (0, 62)	0.052	0.882	2	8.51	–
Ouyang, 2014 [16]	Laryngeal cancer	H/L	76 (53, 88)	< 0.001	0.329	2	2.83	–
Thomopoulos, 2015 [20]	Leukemia	H/L	55 (0, 82)	0.048	0.128	2	2.34	–
Sang, 2013 [17]	Liver cancer	H/L	10 (0, 47)	0.337	0.05	7	4.24	0.118
Galarraga, 2016 [2]	Lung cancer	H/L	88 (84, 92)	< 0.001	0.104	9	9.29	–
Han, 2016 [6]	Lymphoma	H/L	68 (30, 86)	0.004	0.775	2	2.32	–
Yew, 2016 [27]	Melanoma	H/L	46 (0, 73)	0.048	0.824	5	2.7	0.107
Vaseghi, 2016 [22]	Nonmelanoma	H/L	29 (0, 71)	0.22	0.512	1	1.28	–
Li, 2016 [11]	Oral cancer	H/L	53 (16, 74)	0.008	0.685	7	4.89	0.245
Miranda, 2017 [15]	Oral/Pharyngeal cancer	H/L	50 (13, 72)	0.009	0.287	6	4.86	0.537
Steevens, 2007 [18]	Ovarian cancer	H/L	51 (11, 73)	0.012	0.047	3	4.57	–
Turati, 2012 [21]	Pancreatic cancer	H/L	50 (32, 64)	< 0.001	0.876	12	14.86	–
Xia, 2017 [26]	Prostate cancer	H/L	52 (32, 71)	0.001	0.229	4	6.63	–
Gan, 2017 [3]	Rectal cancer	H/L	13 (0, 51)	0.308	0.822	1	3.13	–
Han, 2017 [5]	Thyroid cancer	H/L	0 (0, 71)	0.591	0.746	0	1	–
Li, 2013 [12]	Breast cancer	Per 1 cup	0 (0, 45)	0.796	0.619	0	3.30	–
Gan, 2017 [3]	Colon cancer	Per 1 cup	23 (0, 58)	0.203	0.818	3	3.00	0.999
Gan, 2017 [3]	Colorectal cancer	Per 1 cup	34 (0, 63)	0.081	0.434	3	4.80	–
Huang, 2013 [7]	Endometrial cancer	Per 1 cup	35 (0, 73)	0.161	0.274	3	1.30	0.098
Xie, 2014 [26]	Gastric cancer	Per 1 cup	48 (0, 76)	0.051	0.198	1	2.66	–
Malerba, 2013 [14]	Glioma	Per 1 cup	44 (0, 81)	0.15	0.961	0	0.83	–
Kennedy, 2017 [10]	Liver cancer	Per 1 cup	59 (30, 76)	0.001	0.00005093	13	7.22	0.005
Tang, 2010 [19]	Lung cancer	Per 1 cup	41 (0, 73)	0.092	0.499	3	2.21	0.543
Turati, 2012 [21]	Pancreatic cancer	Per 1 cup	65 (47, 76)	< 0.001	0.326	4	8.67	–
Gan, 2017 [3]	Rectal cancer	Per 1 cup	12 (0, 50)	0.323	0.383	0	3.15	–

Abbreviation: H/L The highest intake vs. lowest intake of coffee
Reference was provided in Additional file 1: Table S3
* P value from Cochran’s Q test for heterogeneity.
^a Observed and expected number of significant studies using effect of largest study (smallest standard error) of each meta-analysis as plausible effect size
^b P value from test for excess significance bias. All statistical tests are two sided

Back to article page

ISSN: 1471-2407

Contact us

Submission enquiries: bmccancer@biomedcentral.com
General enquiries: ORSupport@springernature.com