Concordance of smoking status as per latest UK Biobank record and from latest of any of UK Biobank or linked primary care datasets (ie, the smoking exposure used in the final analyses)
UK Biobank smoking data | Most recently recorded smoking data from any of the three data sources (UK Biobank or linked primary care data) | |||||
Never-smoker | Former smoker | Light smoker | Moderate smoker | Heavy smoker | Missing | |
Never-smoker (n=233 782) | 224 137 (95.87) | 9602 (4.11) | 26 (0.01) | 13 (0.01) | <10 | 0 (0.00) |
Former smoker (n=144 772) | 21 820 (15.07) | 122 485 (84.61) | 213 (0.15) | 158 (0.11) | 96 (0.07) | 0 (0.00) |
Light smoker (n=5954) | 249 (4.18) | 3291 (55.27) | 2269 (38.11) | 137 (2.30) | <10 | 0 (0.00) |
Moderate smoker n=11 906 | 325 (2.73) | 6249 (52.49) | 697 (5.85) | 4490 (37.71) | 145 (1.22) | 0 (0.00) |
Heavy smoker (n=9444) | 189 (2.00) | 4505 (47.70) | 265 (2.81) | 823 (8.71) | 3662 (38.78) | 0 (0.00) |
Missing (n=15 611) | 2232 (14.30) | 9462 (60.61) | 477 (3.06) | 178 (1.14) | 50 (0.32) | 3212 (20.58) |
Figures in italic in parentheses correspond to the row percentage. If there was conflict between the two data sources, the most recent record was used as the exposure definition.