Table 6.
Covariates | Univariate analysis | Multivariate analysis | ||
---|---|---|---|---|
p | OR | 95 % CI | p | |
First modela | ||||
Age | < 0.001 | 1.04 | 1.02 - 1.05 | < 0.001 |
Gender | < 0.001 | 3.58 | 2.10 - 6.09 | < 0.001 |
rs744166 | 0.007 | 1.64 | 1.16 - 2.31 | 0.005 |
EPIYA-C number | < 0.001 | 2.28 | 1.41 - 3.68 | 0.001 |
Second modelb | ||||
Age | < 0.001 | 1.05 | 1.05 - 1.06 | < 0.001 |
Gender | < 0.001 | 3.23 | 2.02 - 5.17 | < 0.001 |
rs744166 plus EPIYA-C number | < 0.001 | 3.01 | 2.29 - 3.97 | < 0.001 |
The Hosmer-Lemeshow tests showed good fit for models a(10 steps, 8 degrees of freedom; p = 0.42) b(10 steps, 8 degrees of freedom; p = 0.25), OR, odds ratio; 95 % CI: confidence interval. We constructed two models studying the loci, according to the amount of G rs744166 allele, treated as categorical variables having 3 values with 0 meaning the AA genotype, 1 the carrier of one G allele and 2 the GG genotype. In the second model the variable was obtained by the sum of the amount of STAT3 G allele and the number of EPIYA-C segments