Table 2.
Associations of dimensions of alcohol drinking behavior with PSA and prostate cancer risk
| Ratio o means or rela ratiosf geometric, odds ratios tive risk | p1 | ||
|---|---|---|---|
| PSA levels – controls only2 (12,727 controls) | |||
| Binge drinking3 | 0.96 | (0.92; 1.00) | – |
| Drinking most days | 0.96 | (0.93; 0.98) | – |
| Total prostate cancer4 (2,386 cases) | |||
| Binge drinking3 | 0.93 | (0.81; 1.08) | – |
| Drinking most days | 0.95 | (0.87; 1.04) | – |
| Localized stage prostate cancer5 (1,914 cases) | |||
| Binge drinking3 | 0.94 | (0.81; 1.10) | – |
| Drinking most days | 0.97 | (0.88; 1.07) | – |
| Advanced stage prostate cancer5 (206 cases) | |||
| Binge drinking3 | 0.93 | (0.60; 1.45) | 0.955 |
| Drinking most days | 0.77 | (0.58; 1.04) | 0.156 |
| Low-grade prostate cancer5 (1,603 cases) | |||
| Binge drinking3 | 0.87 | (0.73; 1.03) | – |
| Drinking most days | 0.91 | (0.81; 1.01) | – |
| High-grade prostate cancer5 (775 cases) | |||
| Binge drinking3 | 1.09 | (0.87; 1.37) | 0.097 |
| Drinking most days | 1.02 | (0.88; 1.19) | 0.191 |
Analysis of 2,386 cases and 12, 727 controls.
Test for difference in the effect estimates for localized versus advanced and low-grade versus high-grade prostate cancer.
Ratios of geometric means and 95% confidence intervals from linear regression, adjusted for the design variables on which controls and cases were matched–additionally adjusted for age as continuous variable.
Regularly consuming 5+ units/occasion.
Odds ratios and 95% confidence intervals from conditional logistic regression–strata defined by age, center and date of recruitment, and clinic visit–additionally adjusted for age as continuous variable.
Relative risk ratios and 95% confidence intervals from logistic regression, adjusted for the design variables on which controls and cases were matched—additionally adjusted for age as continuous variable. For definitions of localized/advanced stage and low/high-grade, see methods section. Abbreviations PSA, prostate-specific antigen.