Overview of the study populations included in the discovery and replication cohorts
| Variable | Discovery (GWAS stage 1)* (n=1910) | Replication (stage 2)* (n=1170) | Validation† Patients with alcohol misuse (n=1520) | |||||||||||
| Germany-Switzerland-Austria (n=1910) | UK (cohort 1) (n=860) | Germany (cohort 2) (n=238) | Italy (cohort 3) (n=72) | Germany (n=1179) | UK (n=341) | |||||||||
| Cases (n=1066) | Controls (n=844) | ‡P value | Cases (n=70) | Controls (n=790) | ‡P value | Cases (n=42) | Controls (n=196) | P value‡ | Cases (n=36) | Controls (n=36) | ‡P value | Non-cirrhosis controls | Non-cirrhosis controls | |
| Age (yr) | 64.8 (8.5) (100%) | 57.1 (9.7) (100%) | <0.0001 | 60.2 (5.9) (100%) | 56.3 (8.9) (100%) | <0.0001 | 67.1 (9.1) (100%) | 58.5 (9.7) (100%) | >0.05 | 72.7 (8.0) (100%) | 53.3 (8.9) (100%) | <0.0001 | 42.7 (10.4) (100%) | 48.6 (10.5) (100%) |
| Proportion male (n: %) | 968 (90.8) | 624 (73.9) | <0.0001 | 67 (95.7) | 577 (73.0) | <0.0001 | 35 (83.3) | 131 (66.8) | <0.05 | 32 (88.9) | 31 (86.1) | >0.05 | 1148 (97.4) | 263 (77.1) |
| BMI (kg/m2) § | 28.1 (4.8) (69%) | 26.5 (5.3) (91%) | <0.0001 | 29.3 (4.4) (100%) | 27.5 (4.8) (92%) | <0.05 | 24.8 (3.5) (57%) | 26.1 (5.8) (52%) | >0.05 | 27.0 (3.8) (64%) | 27.0 (6.8) (75%) | >0.05 | 25.3 (4.5) (81%) | 24.7 (2.3) (53%) |
| BMI kg/m2; (n: %) <25 | 183 (24.7) | 308 (40.3) | <0.0001 | 13 (18.6) | 227 (31.3) | <0.05 | 13 (54.2) | 52 (51.0) | >0.05 | 6 (26.1) | 13 (48.1) | >0.05 | 505 (52.6) | 94 (52.2) |
| 25–30 | 333 (45.0) | 295 (38.6) | 27 (38.6) | 288 (39.7) | 8 (33.3) | 27 (26.5) | 12 (52.2) | 7 (25.9) | 345 (35.9) | 86 (47.8) | ||||
| >30 | 224 (30.3) | 162 (21.2) | 30 (42.9) | 211 (29.1) | 3 (12.5) | 23 (22.5) | 5 (21.7) | 7 (25.9) | 110 (11.5) | 0 (0) | ||||
| Diabetes type II + (n: %)§ | 337 (47.1) (67%) | 136 (30.1) (54%) | <0.0001 | 20 (28.6) (100%) | 102 (12.9) (100%) | <0.0001 | 14 (36.8) (90%) | 36 (19.7) (93%) | <0.05 | 16 (44.4) (100%) | 5 (13.9) (100%) | <0.05 | 58 (5.7) (86%) | 8 (3.6) (66%) |
*Cases and controls were assigned to groups as detailed in the Methods section.
†Validation cohorts were used in post hoc risk assessment.
‡P values were calculated from Student’s t-test for quantitative variables and as Pearson’s χ2 test for categorical variables.
§Data are reported as mean±SD or as number (%) ; Completeness of phenotypic information for age, BMI and type 2 diabetes status are reported as percentage of subjects with available information below the mean value.
BMI, body mass index; GWAS, Genome-Wide Association Study.