Table 2.
Pooled ORs and corresponding 95% CIs of pancreatic cancer according to alcohol consumption in the International Pancreatic Cancer Case–Control Consortium (PanC4).
| Frequency, drinks per daya | Cases, n (%) | Controls, n (%) | ORb (95% CI) |
| Total alcohol | |||
| 0 to <1 | 3587 (64.2) | 7044 (59.5) | 1 (Referent) |
| 1 to <2 | 693 (12.4) | 1580 (13.4) | 1.02 (0.76–1.37) |
| 2 to <3 | 434 (7.8) | 1149 (9.7) | 0.91 (0.73–1.15) |
| 3 to <4 | 230 (4.1) | 637 (5.4) | 0.93 (0.69–1.26) |
| 4 to <5 | 229 (4.1) | 541 (4.6) | 1.26 (0.99–1.61) |
| 5 to <6 | 88 (1.6) | 245 (2.1) | 1.14 (0.86–1.50) |
| 6 to <7 | 80 (1.4) | 171 (1.4) | 1.59 (1.16–2.20) |
| 7 to <8 | 46 (0.8) | 94 (0.8) | 1.30 (0.81–2.09) |
| 8 to <9 | 61 (1.1) | 156 (1.3) | 1.25 (0.74–2.10) |
| ≥9 | 121 (2.2) | 201 (1.7) | 1.60 (1.16–2.22) |
| Missing | 16 (0.3) | 9 (0.1) | – |
| P value for trend | 0.302 | ||
| Total alcoholc | |||
| 0 to <1 | 3108 (63.5) | 6698 (59.0) | 1 (Referent) |
| 1 to <2 | 613 (12.5) | 1525 (13.4) | 1.01 (0.74–1.38) |
| 2 to <3 | 398 (8.1) | 1105 (9.7) | 0.96 (0.76–1.22) |
| 3 to <4 | 194 (4.0) | 621 (5.5) | 0.88 (0.66–1.19) |
| 4 to <5 | 205 (4.2) | 536 (4.7) | 1.20 (0.91–1.59) |
| 5 to <6 | 76 (1.6) | 243 (2.2) | 1.13 (0.81–1.56) |
| 6 to <7 | 75 (1.5) | 170 (1.5) | 1.63 (1.18–2.26) |
| 7 to <8 | 44 (0.9) | 93 (0.8) | 1.34 (0.82–2.19) |
| 8 to <9 | 56 (1.1) | 153 (1.4) | 1.27 (0.69–2.32) |
| ≥9 | 110 (2.3) | 192 (1.7) | 1.86 (1.40–2.47) |
| Missing | 14 (0.3) | 9 (0.1) | – |
| P value for trend | 0.302 | ||
| Wine onlyd | |||
| 0 to <1e | 3317 (90.8) | 5805 (85.2) | 1 (Referent) |
| 1 to <4 | 251 (6.9) | 704 (10.3) | 1.18 (0.91–1.52) |
| ≥4 | 87 (2.4) | 307 (4.5) | 1.46 (1.02–2.08) |
| P value for trend | 0.017 | ||
| Beer onlyd | |||
| 0 to <1e | 3317 (96.7) | 5805 (96.9) | 1 (Referent) |
| 1 to <4 | 78 (2.3) | 143 (2.4) | 0.81 (0.46–1.43) |
| ≥4 | 34 (1.0) | 46 (0.8) | 0.69 (0.41–1.17) |
| P value for trend | 0.105 | ||
| Liquor onlyd | |||
| 0 to <1e | 3317 (97.3) | 5805 (96.9) | 1 (Referent) |
| ≥1 | 93 (2.7) | 188 (3.1) | 1.13 (0.84–1.52) |
Derived from cumulative lifetime consumption for study providing information (see ‘Methods’).
Pooled ORs were computed using random effects models. Study-specific ORs were adjusted for age, sex, race/ethnicity, education, body mass index, history of diabetes, tobacco smoking (in categories, plus a continuous term), and center for multicentric studies.
Excluding proxy interviews (692 cases and 482 controls).
No information on type of alcoholic beverages was available in the NCI study.
Abstainers or occasional drinkers of any kind of alcoholic beverages. OR, odds ratio; CI, confidence interval; NCI, National Cancer Institute.