Table 4.
ORs and 95% CIs from conditional logistic regression for the association between cysteine related-metabolites (thioproline, cysteine, and cystine) and lethal prostate cancer risk in the ATBC Study
| Model | Thioproline | Cysteine | Cystine | Thioproline, cysteine, and cystine a |
|---|---|---|---|---|
| Crude model adjusting for matching factors | ||||
| Continuous | ||||
| Per s.d. | 0.75 (0.65, 0.86) | 0.80 (0.68, 0.93) | 0.80 (0.70, 0.91) | 0.71 (0.61, 0.83) |
| P | 8.2×10−5 | 0.004 | 0.0008 | 1.1×10−5 |
| Quartile categories | ||||
| First | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Second | 0.93 (0.65, 1.34) | 0.81 (0.56, 1.18) | 0.78 (0.55, 1.11) | 0.77 (0.54, 1.10) |
| Third | 0.68 (0.47, 0.99) | 0.63 (0.42, 0.93) | 0.65 (0.45, 0.93) | 0.67 (0.47, 0.96) |
| Fourth | 0.54 (0.37, 0.79) | 0.57 (0.37, 0.87) | 0.73 (0.51, 1.05) | 0.53 (0.36, 0.78) |
| P for trend b | 0.0005 | 0.006 | 0.048 | 0.0011 |
| Multivariable-adjusted model c | ||||
| Continuous | ||||
| Per s.d. | 0.73 (0.63, 0.85) | 0.79 (0.67, 0.93) | 0.78 (0.68, 0.90) | 0.70 (0.59, 0.82) |
| P | 6.3×10−5 | 0.0038 | 0.0005 | 6.6×10−6 |
| Quartile categories | ||||
| First | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Second | 0.89 (0.61, 1.29) | 0.83 (0.57, 1.21) | 0.80 (0.56, 1.14) | 0.76 (0.53, 1.09) |
| Third | 0.66 (0.45, 0.96) | 0.64 (0.43, 0.96) | 0.63 (0.43, 0.91) | 0.63 (0.44, 0.92) |
| Fourth | 0.51 (0.34, 0.76) | 0.56 (0.36, 0.87) | 0.73 (0.50, 1.06) | 0.50 (0.33, 0.75) |
| P for trend b | 0.0003 | 0.0055 | 0.039 | 0.0006 |
Abbreviations: OR = odds ratio, CI = confidence interval, s.d. = standard deviation, ATBC = Alpha-Tocopherol, Beta-Carotene Cancer Prevention, ICD = International Classification of Diseases.
The cysteine-related metabolite combination (thioproline, cysteine, and cystine) is modeled based on the formula: β1X1 + β2X2 + β3X3, Xj denoting the standardized value from the jth metabolite, and βj denoting the coefficient of the metabolite from regression model. Lethal prostate cancer was defined as cases who died of prostate cancer (ICD-9=185 or ICD-10=C61).
P for trend: the statistical significance of the coefficient of the quartile variable (median value within each quartile).
Model adjusting for matching factors, BMI, smoking, ATBC intervention group, and baseline serum total cholesterol, HDL cholesterol, alpha-tocopherol and retinol. Matching variables included age at blood collection (± 1 year), and date of baseline blood collection (± 30 days).