Table 2.
Association of smoking and alcohol consumption with CRC risk by single molecular pathological subtypes.
| Ever regular smoking | Alcohol: g/ethanol/daya | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Never | Ever | Former | Current | ≤24.6 | >24.6 | |||||
| N (%) | N (%) | OR (95% CI)b | N (%) | OR (95% CI)b | N (%) | OR (95% CI)b | N (%) | N (%) | OR (95% CI)b | |
| Controls | 1257 (50.8) | 1217 (49.2) | 1 | 945 (38.3) | 1 | 268 (10.9) | 1 | 1996 (80.8) | 473 (19.2) | 1 |
| MSI | 111 (47.0) | 125 (53.0) | 1.60 (1.19–2.16) | 79 (33.6) | 1.26 (0.90–1.76) | 45 (19.1) | 2.79 (1.86–4.18) | 196 (82.7) | 41 (17.3) | 1.07 (0.72–1.58) |
| MSS | 916 (46.8) | 1041 (53.2) | 1.24 (1.08–1.43) | 756 (38.6) | 1.19 (1.03–1.39) | 285 (14.6) | 1.41 (1.14–1.75) | 1486 (76.1) | 466 (23.9) | 1.32 (1.11–1.57) |
| p-heterogeneity | 0.071 | 0.584 | 0.001 | 0.385 | ||||||
| BRAF-mut | 91 (51.4) | 86 (48.6) | 1.73 (1.22–2.46) | 63 (35.6) | 1.56 (1.07–2.28) | 23 (13.0) | 2.40 (1.41–4.07) | 152 (86.4) | 24 (13.6) | 1.09 (0.61–1.65) |
| BRAF-wt | 955 (46.1) | 1118 (53.9) | 1.26 (1.10–1.44) | 793 (38.3) | 1.17 (1.01–1.36) | 324 (15.6) | 1.52 (1.24–1.88) | 1574 (76) | 496 (24) | 1.30 (1.10–1.54) |
| p-heterogeneity | 0.036 | 0.075 | 0.074 | 0.373 | ||||||
| KRAS-mut | 350 (47.9) | 381 (52.1) | 1.19 (0.99–1.43) | 285 (39.0) | 1.17 (0.96–1.43) | 96 (13.1) | 1.26 (0.95–1.68) | 568 (77.9) | 161 (22.1) | 1.19 (0.94–1.49) |
| KRAS-wt | 690 (45.6) | 824 (54.4) | 1.34 (1.15–1.55) | 577 (38.1) | 1.22 (1.04–1.44) | 246 (16.3) | 1.70 (1.36–2.13) | 1147 (75.9) | 364 (24.1) | 1.30 (1.09–1.56) |
| p-heterogeneity | 0.226 | 0.641 | 0.039 | 0.456 | ||||||
| CIMP-high | 196 (49.2) | 202 (50.8) | 1.47 (1.15–1.87) | 147 (36.9) | 1.32 (1.02–1.71) | 55 (13.8) | 2.01 (1.40–2.88) | 323 (81) | 76 (19) | 1.22 (0.90–1.66) |
| CIMP-low/neg | 930 (46.1) | 1088 (53.9) | 1.25 (1.09–1.44) | 771 (38.2) | 1.18 (1.01–1.37) | 316 (15.7) | 1.50 (1.22–1.85) | 1529 (76) | 484 (24) | 1.28 (1.08–1.52) |
| p-heterogeneity | 0.128 | 0.261 | 0.101 | 0.868 | ||||||
aMultinomial logistic regression model adjusted for: Sex, age, BMI, education level, history of colorectal cancer in first-degree relative, previous endoscopy, diabetes, ever NSAIDs regular use and average lifetime alcohol consumption/ever regular smoking. Ever/former/current smoking compared to never smoking.
bHigh alcohol consumption was defined as the fourth quartile of the average daily lifetime gram ethanol consumption among alcohol drinkers (>24.6 g) and was compared in analyses to low/never consumption (≤24.6 g).