Table 2. Risk of CRC by vitamin B2 and B6 status.
| Vitamin B2a | Vitamin B6a |
||
|---|---|---|---|
| Tertile 1 (<30.8) | Tertile 2 (30.8–45.6) | Tertile 3 (≥45.6) | |
| Tertile 1 (<9.0) | |||
| Cases/controls (n) | 105/148 | 76/151 | 46/97 |
| OR-BNb | ref | 0.78 | 0.80 |
| OR-CLR (95% CI)c | ref | 0.70 (0.48, 1.02) | 0.67 (0.44, 1.04) |
| Tertile 2 (9.0–15.3) | |||
| Cases/controls (n) | 78/138 | 64/134 | 73/123 |
| OR-BNb | 0.77 | 0.69 | 0.79 |
| OR-CLR (95% CI)c | 0.78 (0.54, 1.14) | 0.66 (0.44, 0.99) | 0.83 (0.56, 1.21) |
| Tertile 3 (≥15.3) | |||
| Cases/controls (n) | 38/110 | 50/110 | 78/175 |
| OR-BNb | 0.53 | 0.70 | 0.61 |
| OR-CLR (95% CI)c | 0.48 (0.31, 0.75) | 0.63 (0.41, 0.97) | 0.62 (0.43, 0.90) |
| Pinteractiond = 0.004 | |||
OR: Odds ratio - CI: Confidence interval - BN: Bayesian Network - CLR: conditional logistic regression – PLP: Pyridoxal 5′ phosphate.
aConcentrations in nmol/l, cut-offs based on the distribution of the controls.
bEstimated using fitted parameters of the estimated BN, additionally adjusted for the matching variables.
cEstimates from a CLR-model adjusted for the matching variables by risk set stratification. Adjusting for other potential confounders had essentially no effect on the estimates.
dCalculated by modeling log-transformed variables as multiplicative interaction terms in a CLR-model.