Table 1. The effects of SNPs on PSA level (ng/mL) in men with PSA3-10ng/mL and a diagnosis of prostate cancer.
| Low Risk (N = 684) | High (N = 226) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SNP | Alleles (O/X) | Allele associated with Increasing PSA (X) | Effect of X allele on PSA (ng/mL) a | 95% CI | p-value | F- statistic b | R2 (%) | Effect of X allele on PSA (ng/mL) a | 95% CI | p-value | F- statistic b | R2 (%) |
| rs10788160 | G/A | A | -0.16 | (-0.36, 0.03) | 0.10 | 3.08 | 0.45 | -0.28 | (-0.76, 0.20) | 0.26 | 0.54 | 0.30 |
| rs11067228 | G/A | A | -0.13 | (-0.30, 0.04) | 0.14 | 1.95 | 0.29 | -0.38 | (-0.81, 0.05) | 0.08 | 3.18 | 1.72 |
| rs17632542 | C/T | T | 0.17 | (-0.21, 0.56) | 0.38 | 0.65 | 0.09 | 0.52 | (-0.29, 1.34) | 0.21 | 0.77 | 0.42 |
| rs2736098 | C/T | T | 0.05 | (-0.14, 0.25) | 0.60 | 0.19 | 0.03 | 0.14 | (-0.34, 0.61) | 0.57 | 0.43 | 0.24 |
CI = confidence interval.
a Additive model, calculated using regression, adjusting for exact age, study centre and 10 principal components to account for confounding by population stratification.
b F-statistic and R-squared (R2) indicate how much of the variability in PSA levels is explained by each SNP. Calculated using regression, unadjusted