Skip to main content
. 2015 Oct 17;6(39):42312–42321. doi: 10.18632/oncotarget.5994

Table 2. Meta-analyses results of the association between germline HOXB13 p.Gly84Glu mutation and cancer risk.

Variables No. Sample size Pheterogeneity Analyzing model OR 95% CI P value
Total 25 145,257 <0.001 Random 2.872 2.121, 3.888 <0.001
Cancer type
 Prostate cancer 19 130,795 <0.001 Random 3.248 2.313, 4.560 <0.001
 Breast cancer 3 9,423 0.958 Fixed 1.423 0.774, 2.615 0.256
 Colotrectal cancer 3 5,039 0.205 Fixed 2.458 0.978, 6.177 0.056
Ethnicity
 Caucasians 21 144,007 <0.001 Random 2.673 1.920, 3.720 <0.001
 Mixed decedents 4 12,507 0.362 Fixed 4.164 2.226, 7.790 <0.001
Genotype method
 TaqMan 10 46,126 0.149 Fixed 3.649 2.728, 4.880 <0.001
 Sanger sequencing 2 4,420 0.201 Fixed 3.862 1.110, 13.441 0.034
 MassARRAY iPLEX 4 13,842 0.201 Fixed 2.956 2.337, 3.740 <0.001
 Illumina SNP chips 6 72,039 0.135 Fixed 3.934 2.479, 6.245 <0.001
 Complex methods 3 8,830 0.067 Fixed 1.119 0.784, 1.597 0.537
Source of control
 Hospital based -HB 14 112,986 0.155 Fixed 3.363 2.449, 4.619 <0.001
 Population based -PB 5 17,670 0.481 Fixed 3.196 2.234, 4.573 <0.001
 PB/HB 4 11,494 0.001 Random 2.378 0.814, 6.952 0.113
 Family based -FB 1 443 Fixed 2.015 1.286, 3.156 0.002
 HB/FB 1 2,664 Fixed 4.168 1.235, 14.062 0.021

OR, Odds ratio; 95% CI, 95% confidence interval.